IJCIIS September 2010 Vol. 1 No. 7

8/8/2019 IJCIIS September 2010 Vol. 1 No. 7

http://slidepdf.com/reader/full/ijciis-september-2010-vol-1-no-7 1/79

International Journal of

Computational Intelligence and

Information Security

ISSN: 1837-7823

September 2010

Vol. 1 No. 7

© IJCIIS Publication



International Journal of Computational Intelligence and Information Security, Vol. 1 No. 7, September 2010

2

IJCIIS Editor and Publisher

P Kulkarni

Publisher’s Address:

5 Belmar Crescent, Canadian

Victoria, Australia

Phone: +61 3 5330 3647

E-mail Address: [email protected]

Publishing Date: September 30, 2010

Members of IJCIIS Editorial Board

Prof. A Govardhan, Jawaharlal Nehru Technological University, India

Dr. Awadhesh Kumar Sharma, Madan Mohan Malviya Engineering College, India

Prof. Ayyaswamy Kathirvel, BS Abdur Rehman University, India

Prof. Deepankar Sharma, D. J. College of Engineering and Technology, India

Dr. D. R. Prince Williams, Sohar College of Applied Sciences, Oman

Prof. Durgesh Kumar Mishra, Acropolis Institute of Technology and Research, India

Dr. Imen Grida Ben Yahia, Telecom SudParis, France

Dr. Himanshu Aggarwal, Punjabi University, India

Dr. Jagdish Lal Raheja, Central Electronics Engineering Research Institute, India

Prof. Natarajan Meghanathan, Jackson State University, USA

Dr. Oluwaseyitanfunmi Osunade, University of Ibadan, Nigeria

Dr. Ousmane Thiare, Gaston Berger University, Senegal

Dr. K. D. Verma, S. V. College of Postgraduate Studies and Research, India

Prof. M. Thiyagarajan, Sastra University, India

Dr. Manjaiah D. H. Mangalore University, India

Dr.N.Ch.Sriman Narayana Iyengar, VIT University ,India

Prof. Nirmalendu Bikas Sinha, College of Engineering and Management, Kolaghat, India

Dr. Rajesh Kumar, National University of Singapore, Singapore

Dr. Raman Maini, University College of Engineering, Punjabi University, India

Dr. Shahram Jamali, University of Mohaghegh Ardabili, Iran

Dr. Shishir Kumar, Jaypee University of Engineering and Technology, India

Prof. Sriman Narayana Iyengar, VIT University, India

Dr. Sujisunadaram Sundaram, Anna University, India

Dr. Sukumar Senthilkumar, National Institute of Technology, India

Prof. V. Umakanta Sastry, Sreenidhi Institute of Science and Technology, India

Dr. Venkatesh Prasad, Lingaya's University, India

Journal Website: https://sites.google.com/site/ijciisresearch/




3

Contents

1. Modelling and Simulation of Closed Loop Speed Controlled PFC Half Bridge

Converter fed PMBLDC motor ( pages 4-13)

2. An Approach to Compress & Secure Image Communication (pages 14-19)

3. Cloud Computing Approaches for Educational Institutions (pages 20-28)

4. Preprocessing of Digital Mammogram for Image Analysis (pages 29-41)

5. Comparative Analysis of Performance of Series FACTS Devices Using PSO Based

Optimal Power Flow Solutions (pages 42-52)

6. Secure and Unique Biometric Template Using Post Quantum Cryptosystem (pages 53-

61)

7. Statcom In Eight Bus System For Power Quality Enhancement (pages 62-68)

8. Improved Modification of single stage Ac-Ac converter for Induction Heating

Application (pages 69-77)




4

Modelling and Simulation of Closed Loop Speed Controlled PFC Half

Bridge Converter fed PMBLDC motor

C.Umayal and S.Rama Reddy

Research Scholar,

Electrical and Electronics Engg Dept,

Anna University, Chennai, India

[email protected]

Professor,

Electrical and Electronics Engg Dept ,

Jerusalem College of Engineering, Chennai,

[email protected]

Abstract

Digital Simulation of a Power Factor Correction (PFC) half bridge converter based adjustable speed voltagecontrolled VSI fed PMBLDC motor is presented in this paper. A single-phase AC-DC converter topology basedon the half bridge converter is employed for PFC which ensures near unity power factor over wide speed range.The proposed speed control scheme has the concept of DC link voltage control proportional to the desired speedof the PMBLDC motor. The speed is regulated by a PI controller. The PFC converter based PMBLDCM drive isdesigned, modeled and simulated using MATLAB-SimuLink environment. This drive also ensures highaccuracy, robust operation from near zero to high speed.

Keywords: Boost rectifier, low conduction losses, power factor correction (PFC), Hall position sensors,

permanent magnet brushless DC motor, BLDC motor, closed loop speed control, PI controller.




5

1. Introduction

Research on PFC circuits for high power applications has increased. Three level boost type PFC convertersis an attractive solution for high power density and high efficiency 3 phase pre regulators. Single stage PFC circuitshave presented a serious challenge recently to increase the output power capability with optimized componentratings. PFC using half bridge converters is the recent trend. In these rectifiers only half of the output voltage is

applied across the switches thus reducing the stress on them to a greater extent.Recent developments in power electronics, micro electronics and modern control technologies have greatlyinfluenced the wide spread use of permanent magnet motors. The major classification of Permanent Magnet motorsare permanent magnet synchronous motor (PMSM) and Permanent Magnet Brushless DC motors (PMBLDCM).While PMSM has sinusoidal back-emf waveform the BLDC motor has trapezoidal back-emf waveform.

The BLDC motors have a good performance, high efficiency, low maintenance, high power density, lowinertia. Hence it finds widespread use in variety of applications in motion control. The classical PI controller is awide used controller. Comparing with conventional DC motors, BLDC motors do not have brushes for commutation.Instead they are electronically commutated. BLDC motors have many advantages over brushed DC motors andinduction motors, like better speed-torque characteristics, high dynamic response, high efficiency, noiselessoperation and wide speed ranges. Torque to weight ratio is higher enabling it to be used in applications where spaceand weight are critical factors.

A new generation of microcontrollers and advanced electronics has overcome the challenge of implementing required control functions, making the BLDC motor more practical for a wide range of uses [1], [2],

[3].

2. Power Factor Correction Converters

In the field of inverter appliance, the application of AC/DC/AC converter is more and more popular. Thereare serious contaminations from harmonic currents at the mains side, making the products not pass IEC61000-3-2and IEC61000-3-12 (harmonics standards) successfully. In order to mitigate the harmonic current pollution, most of the household inverters are equipped with power factor corrector as front AC/DC converter, and the input powerfactor approaches one. But the conventional active PFC has to employ uncontrolled rectifier and costly boostinductor, and these power components result in power loss, low efficiency and high cost. Nevertheless, the bridgelessPFC (BLPFC) is characteristic of small number of power switches, making room for low power loss [4].Additionally, in the conventional active PFC the power switches are in on and off state in a whole mains period,

enduring high voltage and current stresses, producing a lot of switching loss and conduction loss and limiting theefficiency.

A voltage source inverter can run the BLDC motor by applying three phase square wave voltages to the statorwinding of the motor .A variable frequency square wave voltage can be applied to the motor by controlling theswitching frequency of the power semiconductor switches. The square wave voltage will induce low frequencyharmonic torque pulsation in the machine. Also variable voltage control with variable frequency operation is notpossible with square wave inverters. Even updated pulse-width modulation (PWM) techniques used to controlmodern static converters such as machine drives, power factor compensators do not produce perfect waveforms,which strongly depend on the semiconductors switching frequency. Voltage or current converters, as they generatediscrete output waveforms, force the use of machines with special isolation, and in some applications largeinductances connected in series with the respective load. Also, it is well known that distorted voltages and currentwaveforms produce additional power losses, and high frequency noise that can affect not only the power load butalso the associated controllers. All these unwanted operating characteristics associated with PWM converters could

be overcome with improved bridgeless PFC boost converters.

2.1. Principle of the bridgeless topology

The basic topology of the bridgeless PFC boost rectifier [5] is shown in Fig. 1. Compared to theconventional PFC boost rectifier, shown in Fig. 2, one diode is eliminated from the line-current path, so that the linecurrent simultaneously flows through only two semiconductors resulting in reduced conduction losses [6]. Thebridgeless PFC topology removes the input rectifier conduction losses and is able to achieve higher efficiency.However, the bridgeless PFC boost rectifier in Fig. 1 has significantly larger common-mode noise than theconventional PFC boost rectifier [7].




6

g

m

d

s

SB2

g

m

d

s

SB1

RL

LB

D2D1

CAC

Figure 1: Bridgeless Boost Converter

Based on the analysis above, the bridgeless PFC circuit can simplify the circuit topology and improve theefficiency as well.

3. Types of control techniques of PMBLDC motor

Though various control techniques are discussed in [8].Basically two methods are available for controllingPMBLDC motor. They are sensor control and sensorless control.

To control the machine the present position of the rotor is required to determine the next commutationinterval. One is by controlling the DC bus rail voltage and the next one is by PWM method. Some designs utilizeboth to provide high torque at high load and high efficiency at low load. Such hybrid design also allows the controlof harmonic current [9].

In control methods using sensors, mechanical position sensors, such as a hall sensor, shaft encoder orresolver have been utilized in order to provide rotor position information. Hall Position sensors or simply Hallsensors are widely used and popular. The rotor position information is used to generate precise firing commands forpower converter. This ensures drive stability and fast dynamic response. The speed feedback is derived from theposition sensor output signals. Between the two commutation signals the angle variation is constant as the HallEffect Sensors are fixed relative to the motor, thus reducing speed sensing to a simple division. Usually speed and

position of a permanent magnet brushless direct current motor rotor is controlled in a conventional cascade structure.The inner current control loops runs at a larger width than the outer speed loop to achieve an effective cascadecontrol [10].

Various sensorless methods for BLDC motors are analyzed in [11-18]. [11] Proposes a speed control of brushless drive employing PWM technique using digital signal processor. A PSO based optimization of PIDcontroller for a linear BLDC motor is given in [12], Direct torque control and indirect flux control of BLDC motorwith non sinusoidal back emf method controls the torque directly and stator flux amplitude indirectly using d-axiscurrent to achieve a low-frequency torque ripple-free control with maximum efficiency[13-14]. [15] Proposes anovel architecture using a FPGA-based system. Fixed gain PI speed controller has the limitations of being suitablefor a limited operating range around the operating point and having overshoot. To eliminate this problem a fuzzybased gain scheduled PI speed controller is proposed in [16].A new module structure of PLL speed controller isproposed by [17].A fixed structure controller (PI or PID) using time constrained output feedback is given in [18].The above literatures does not deal with PFC in closed loop controlled PMBLDC. This work proposes PFC at the

input of PMBLDC drive.

4. Mathematical Model of the PMBLDC motor

Modeling and simulation play an important role in the design of power electronics system. The classicdesign approach begins with an overall performance investigation of the system, under various circumstancesthrough mathematical modeling.




7

The circuit model of PMBLDC motor is shown in Fig 2.

Figure 2:Motor Circuit Model

The voltage equations of the BLDC motor are as follows:( )

( ) ar a a a aa a ab b ac c

d d V R i L i L i L i

dt dt

λ θ = + + + +

( )( ) br

b b b ba a bb b bc c


dt dt

λ θ = + + + +

( )( ) cr

c c c ca a cb b cc c


dt dt

λ θ = + + + +

In balanced system the voltage equation becomes

0 0

0 0 (1)

0 0

a a a ba ca a a

b b ba b cb b b

c c ca cb c c c

V R i L L L i ed

V R i L L L i edt

V R i L L L i e

= + + − − − − − −

The mathematical model for this motor is described in Equation (1) with the assumption that the magnet has highsensitivity and rotor induced currents can be neglected [3]. It is also assumed that the stator resistances of all thewindings are equal. Therefore the rotor reluctance do not change with angle. Now

a b c

ab bc ca

L L L L

L L L M

= = =

= = =

Assuming constant self and mutual inductance, the voltage equation becomes

0 0 0 0

0 0 0 0 (2)

0 0 0 0

a a a a

b b b b

c c c c

V R i L M i ed

V R i L M i edt

V R i L M i e

−

= + − + − − − − − −

In state space form the equation is arranged as

1 1a a a a

b b b b

c c c c

i i e vd R

i i e vdt L L Li i e v

= − − +

The electromagnetic torque is given as ( ) / e a a b b c c r T e i e i e i ω = + +

The equation of motion is given as ( ) / r e L r

d T T B J

dt ω ω = − − _ _ _ _ _ (3)




8

5. BLDC Motor Speed Control

Block diagram of drive system is shown in Fig.3.In servo applications position feedback is used in theposition feedback loop. Velocity feedback can be derived from the position data. This eliminates a separate velocitytransducer for the speed control loop. A BLDC motor is driven by voltage strokes coupled by rotor position. Therotor position is measured using Hall sensors. By varying the voltage across the motor, we can control the speed of

the motor. When using PWM outputs to control the six switches of the three-phase bridge, variation of the motorvoltage can be obtained by varying the duty cycle of the PWM signal.

Figure 3: Block Diagram of Drive System

6. PMBLDC Motor fed from a Voltage source inverter with PFC Full Bridge Converter

Schematic diagram of a three level voltage source inverter fed PMBLDC motor with PFC full bridgeconverter is shown in Fig.4..This is a closed loop control circuit using 3 Hall Sensors. MOSFETs are used asswitching devices here. To control the speed of the motor the output frequency of the inverter is varied.

To maintain the flux constant the applied voltage is varied in linear proportion to the frequency. TheMATLAB simulation is carried out and the results are presented.

For very slow, medium, fast and accurate speed response, quick recovery of the set speed is important

keeping insensitiveness to the parameter variations. In order to achieve high performance, many conventional controlschemes are employed. At present the conventional PI controller handles these control issues. Moreoverconventional PI controller is very sensitive to step change of command speed, parameter variation and loaddisturbances.

With high frequency switching, the PMBLDC motor rotates at a higher speed. But without the strongmagnetic field at stator, the rotor fails to catch up the switching frequency because of weak pull force. Speed of BLDC motor is indirectly determined by the applied voltage magnitude. Current in the winding is increased byincreasing the voltage. This produces stronger magnetic pull to align the rotor’s magnetic field faster with theinduced stator magnetic field. The rotational speed or the alignment is proportional to the voltage applied to theterminals. The torque pulsation is very high as the step size is reduced.

When using PWM outputs to control the six switches of the three-phase bridge, variation of the motorvoltage can be achieved easily by changing the duty cycle of the PWM signal. In this method the speed is controlledin a closed loop by measuring the actual speed of the motor. The error in the set speed and actual speed is calculated.A Proportional plus Integral (P.I) controller is used to amplify the speed error and dynamically adjust the PWM dutycycle.

7. Simulation Results

The technical specifications of the drive system are as follows C= 2200 microfarad.TON= 5.88 µsecs. TOFF=5.88µsecs.T= 11.76 µsecs. Stator Resistance is 2.875 ohms, Stator Inductance is 8.5e-3mH, Motor inertia is 0.8e-3J.With the help of the designed circuit parameters, the MATLAB simulation is done and results are presented here.Speed is set at 1800 rpm and the load torque disturbances are applied at time t=0.6 sec. The speed regulations areobtained at this speed and the simulation results are shown. The waveforms of input voltage and current are shown in




9

Fig. 5.The waveforms of the phase voltage and currents are shown in Figs 6 and 7 respectively. The waveforms of back EMF are shown in Fig.8. From Fig.5 it can be seen that the power factor is 0.98. The stator currentwaveforms are shown in Fig 7. They are quasi sinusoidal in shape and are displaced by 120°.

Figure 4.Closed Loop model of PMBLDC Motor

Figure 5: Input voltage and current (PF=0.98)

Figure 6: Phase Voltage supplied to the stator windings

Figure 7:Three phase inverter stator current




10

Figure 8: Back emf

Figure 9: Load Torque disturbance applied at t= 0.6 sec

Figure10: Rotor speed in rpm

8. PMBLDC Motor fed from a Voltage source inverter with PFC Half Bridge Converter

Simulink model of PMBLDC motor with PFC half bridge converter is shown in Fig 11. A boost converter isused at the input to improve the power factor. AC input voltage and current waveforms are shown in Fig 12. Thewaveforms of back emf are shown in Fig 13. Step change in load torque is shown in Fig 14.




11

Figure11: Closed Loop Speed Control of PMBLDC Motor with PFC Half Bridge Converter

Figure 12: Input voltage and current

Figure 13: Back EMF

Figure 14: Load torque disturbance applied at t=0.6 sec




12

Figure 15: Rotor speed in rpm.

Figures 9 and 15 show the load torque disturbance applied at time t=0.6 sec for a set speed of 1800 rpm.From Figs 10 and 15, it can be seen that the closed loop system brings the speed to the normal value. From thefigures 5 and 12, it can be seen that the power factor is improved by using half bridge PFC converter.

9. Conclusion

Closed loop controlled VSI fed PMBLDC motor with PFC full bridge and half bridge converters aremodeled and simulated. Feedback signals from the PMBLDC motor representing speed and position are utilized to

get the driving signals for the inverter switches. The simulated results shown are on par with the theoreticalpredictions. The power factor is corrected by using PFC converter. PFC converter fed PMBLDC drive is a viablealternative since it has improved power factor.

References

[1] Tay Siang Hui, K.P. Basu and V.Subbiah Permanent Magnet Brushless Motor Control Techniques, NationalPower and Energy Conference (PECon) 2003 Proceedings, Bangi, Malysia

[2] P.Thirusakthimurugan, P.Dananjayan,’A New Control Scheme for the Speed Control of PMBLDC Motor Drive’

1-4244-0342-1/06/$20.00 ©2006 IEEE[3] Nicola Bianchi,Silverio Bolognani,Ji-Hoon Jang, Seung-Ki Sul,” Comparison of PM Motor structures and

sensorless Control Techniques for zero-speed Rotor position detection” IEEE transactions on Power Electronics,Vol 22, No.6, Nov 2006.

[4] Woo-Young Choi, Jung-Min Kon, Eung-Ho Kim, Jong-Jae Lee and Bong-Hwan Kwon, ”Bridgeless BoostRectifier with Low Conduction Losses and Reduced Diode Reverse-Recovery Problems” IEEE Transactions onIndustrial Electronics, vol 54, No.2, pp.769-780,April 2007.

[5] J. C. Salmon, “Circuit topologies for PWM boost rectifiers operated from 1-phase and 3-phase ac supplies andusing either single or split dc rail voltage outputs,” in Proc. IEEE Applied Power Electronics Conf.,

[6] Laszlo Huber,Yungtaek Jang, Milan M. Jovanovic, ”Performance Evaluation of bridgeless PFC BoostRectifiers”, IEEE Trans ,Power Electronics.,vol.23,No.3,May 2008

[7] B. Lu, R. Brown, and M. Soldano, “Bridgeless PFC implementation using one cycle control technique,” IEEEApplied Power Electronics (APEC) Conf. Proc., pp. 812-817, Mar. 2005 Mar. 1995, pp. 473–479.

[8] R.Krishnan, Electric Motor Drives Modeling, Analysis,and Control, Prentice-Hall Internationa Inc., New Jersey,2001.

[9] New Approach to Rotor Position Detection and Precision Speed Control of the BLDC Motor Yong-Ho YoonTae-Won Lee Sang-Hun Park Byoung-Kuk Lee Chung- 1-4244-0136-4/06/$20.00 '2006 IEEE

[10]Ling KV, WU Bingfang HE Minghua and Zhang Yu, “ A Model predictive controller for multirate cascadesystem”, Proc.of the American Control Conference,ACC 2004,USA, pp.1575-1579.2004[11] G.Madhusudhanrao,B.V.SankerRam,B.Sampath Kumar,K.Vijay Kumar,” Speed Control of BLDC Motor using

DSP”, International Journal of Engineering Science and Technology Vol.2(3), 2010.[12] Yingfa Wang, Changliang Xia, Zhiqiang Li, Peng Song,” Sensorless Control for BLDC motor using support

vector machine based on PSO[13] Salih Baris Ozturk, Hamid A.Toliyat, “Sensorless Direst Torque and Indirect Flux Control of Brushless DC

Motor with Non-Sinusoidal back-emf,978-1-4244-1766-7/08 ©2008 IEEE[14] Salih Baris Ozturk, William C.Alexander,Hamid A.Toliyat,”Direct Torque Control of four-switch brushless DC

motor with non-sinusoidal back emf, IEEE transactions on power electronics, Vol 25,No-2 Feb 2010




13

[15] Kuang-Yao Cheng ,” Novel Architecture of a mixed-mode sensorless control IC for BLDC motors with widespeed ranges”, 978-1-422-2812-0/09 ©2009 IEEE.

[16] M.S.Srinivas, K.R.Rajagopal,” Fuzzy Logic based gain scheduled PI speed controller for PMBLDC motor, 978-1-4244-4859-3/09 © 2009.

[17] Ting-Yu-Chang, Ching-Tsai-Pan,Emily Fang,” A novel high performance variable speed PMBLDC motor drivesystem”, 978-1-4244-4813-5/10 ©2010 IEEE.

[18] Shinn-Ming Sue, Kun-Lin Wu,” A voltage controlled brushless DC motor over extended speed range”, 978-i-4244-1709-3/08 ©2008 IEEE

About the Authors

Umayal Chandrahasan has obtained her ME degree from Anna University, TamilNadu, India, in the year 2005. She is presently doing her research in the area of BLDCmotors. She has 9 years of teaching experience and 8 years of Industrial experience.She is doing her research on PMBLDC motor drives

Rama Reddy Sathi obtained the ME degree from Anna University, Tamil Nadu, India,in 1989. He has pursued research in the area of resonant converters in 1995. He has 2years of industrial experience and 18 years of teaching experience. He has securedA.M.I.E institution Gold medal for obtaining higher marks. He has secured AIMO bestproject award. He is a life member of IEE, IETE, ISTE, SSI, and SPE. He has worked inTata Consulting Engineers, Bangalore and Anna University. He has authored textbookson power electronics and electronic circuits. He has published 20 technical papers inNational and International Conference proceedings / journals in the area of powerelectronics and FACTs.




14

An Approach to Compress & Secure Image Communication

Ramveer Singh1and Deo Brat Ojha

2

1 Mr. Ramveer Singh, R. K .G. Institute of Technology, Gzb. U.P.(India)& Research Scholar Singhania University, jhunjhunu, Rajsthan, INDIA.

e-mail: [email protected]. .2 Dr. Deo Brat Ojha, Prof., Deptt. Of mathematics, R. K. G. Institute of Technology, Gzb.,U.P.(India),

e-mail: [email protected]

Abstract

In our day to day competitive and insecure busy life, requires more secure communication. Like medical field(Telemedicine) uses the transmission of images or videos nearly up to complete efficiency for saving human life,secrecy of communication between secret agents and their relative government, to maintain the confidentiality inmilitary operations, etc. In our approach, we introduced a new scheme to transmit a image over infringeablecommunication environment. Our approach is combination of cryptography and compression. Cryptography

provides secure transmission, whereas compression increases the capacity of communication channel.

Keywords: Image, Compression, Encryption, Decryption, Secure Communication




15

1. Introduction

The amount of image has increased rapidly on the Internet. Image Two different approach of technologieshave been developed for this purpose. The first approach is based on content protection through encryption [1], [2].In this approach, proper decryption of data requires a key. The second approach bases the protection on digitalwatermarking or data hiding, aimed at secretly embedding a message into the data. In the current era the transmission

of Image over internet is so much challenging over the internet. In this manner, the better way to transmit the imageover internet is encryption. Using the cryptography we secure the image as well as also better utilise thecommunication channel through compression technique.

Cryptography is a branch of applied mathematics that aims to add security in the ciphers of any kind of messages. Cryptography algorithms use encryption keys, which are the elements that turn a general encryptionalgorithm into a specific method of encryption. The data integrity aims to verify the validity of data contained in agiven document. [3]

In this current article, we introduced a manner of image transmission over highly traffic channel. Thesection 2, describes the main tools of our newly introduced approach. It also shows that the SEQUITUR and EHDES(Enhanced Data Encryption Standard). SEQUITUR is a dictionary based lossless compression technique andEHDES is a symmetric key encryption technique. Using the SEQUITUR, we can get a compress form of image afterthat we apply the EHDES, for the security of image. The combination of both makes this approach very much usableand secure for medical era. In section 3, it shows the complete working style and behaviour of combination of SEQUITUR and EHDES.

2 Preliminaries

2.1 Enhanced Data Encryption Standard (EHDES)In Enhanced Data Encryption Standard (EHDES) [4, 5,], we use the block ciphering of data and a symmetric key. Astraditional Data Encryption Standard (DES), we also break our data into 64-Bit blocks and use a symmetric key of 56-Bit. EHDES having three phases: 1. Key Generation. 2. Encryption on Input Data. 3. Decryption on InputCipher.

Figure 1: Encryption and Decryption process of EHDES.

2.1.1 Key GenerationIn this phase of EHDES, We moderate the initial 56 Bit key using Random Number Generator (RNG) for everyblock of message (M1, M2, M3 ...Mn). The new generated 56 Bit keys (Knew1, Knew2, Knew3................ Knew n) from initial keyK is used for encryption and decryption for each block of data. For new keys, we generate a random number andimplement a function F on generated random number (NRNG) and the initial key K.

M1, M2, M3......

...Mn

Data or

Message (M)

Encryption Process Decryption Process

EHDES EHDES

56 Bit

Symmetric

Key ‘k’

56 Bit

Symmetric

Key ‘k’

Data or

Message (M)

M1, M2, M3......

...MnCipher text or

Encrypted

Message (C)




16

Figure 2: Process of new generated key (Knew i) of EHDES.

2.1.2. Encryption on Input Data.As we know Data Encryption Standard (DES) is based on block cipher scheme. Message breaks in 64 Bit n blocks of plain text.

M = {M1, M2, M3,...................., Mn}Now, we encrypt our message {M1, M2, M3,...................., Mn} blocks by each new generated key Knew1, Knew2, Knew3................ Knew n.

2.1.3. Decryption on Input CipherDecryption is the reverse process of encryption. For decryption, we also used the same key which is used inencryption. On the receiver side, the user also generate the same new key Knew i for each block of cipher and generateplain text through decryption process of data encryption standard.

2.2 Compression:Previous well known data compression techniques includes the standard algorithm for example, Huffman

coding is an entropy encoding algorithm used for lossless data compression. The term refers to the use of a variable-length code table for encoding a source symbol (such as a character in a file) where the variable-length code tablehas been derived in a particular way based on the estimated probability of occurrence for each possible value of thesource symbol. It was developed by David A. Huffman. There are various data compression algorithm like Run-length encoding, Burrows-Wheeler transform, Dynamic Markov Compression, entropy encoding: Huffman coding

,Adaptive Huffman coding, Shannon-Fano coding, arithmetic coding etc., which has been used for data compressionsuccessfully internationally.In addition, to avoid sending files of the enormous size, a compression scheme can be employed what is

known as lossless compression on secrete message to increase the amount of hiding secrete data, a scheme thatallows the software to exactly reconstruct the original message [6]. The transmission of numerical images often needs an important number of bits. This number is again moreconsequent when it concerns medical images. If we want to transmit these images by network, reducing the imagesize is important. The goal of the compression is to decrease this initial weight. This reduction strongly depends of the used compression method, as well as of the intrinsic nature of the image. Therefore the problem is the following:1. To compress without lossy, but with low factor compression. If you want to transmit only one image, it issatisfactory. But in the medical area these are often sequences that the doctor waits to emit a diagnostic.2. To compress with losses with the risk to lose information. The question that puts then is what the relevantinformation is’s to preserve and those that can be neglected without altering the quality of the diagnosis or theanalysis. The human visual system is one of the means of appreciation, although subjective and being able to varyfrom an individual to another. However, this system is still important to judge the possible causes of degradation andthe quality of the compression [7].

2.2.1 The SEQUITUR Algorithm [8]The SEQUITUR algorithm represents a finite sequence _ as a context free grammar whose language is the singletonset {σ}. It reads symbols one-by-one from the input sequence and restructures the rules of the grammar to maintainthe following invariants:(A) no pair of adjacent symbols appear more than once in the grammar, and

F

56 Bit Initial Key K

Random No.

Generated by RNG

NRNG

Generated 56 Bit keys Knew1, Knew2, Knew3........... Knew n




17

(B) every rule (except the rule defining the start symbol) is used more than once. To intuitively understand thealgorithm, we briefly describe how it works on a sequence 123123. As usual, we use capital letters to denote non-terminal symbols. After reading the first four symbols of the sequence 123123, the grammar consists of the singleproduction rule S 1, 2, 3, 1 where S is the start symbol. On reading the fifth symbol, it becomes S 1, 2, 3, 1, 2Since the adjacent symbols 1, 2 appear twice in this rule (violating the first invariant), SEQUITUR introduces a non-terminal A to get

S A, 3,A A1, 2

Note that here the rule defining non-terminal A is used twice. Finally, on reading the last symbol of the sequence123123 the above grammar becomes

S A, 3, A, 3 A 1, 2

This grammar needs to be restructured since the symbols A, 3 appear twice. SEQUITUR introduces another non-terminal to solve the problem. We get the rules

S B,BB A 3A 1 2

However, now the rule defining non-terminal A is used only once. So, this rule is eliminated to produce the finalresult.

S B, B B 1, 2, 3

Note that the above grammar accepts only the sequence 123123.

3. Our SchemeA complete compression and encryption process includes the following phases:

Phase 1: Generating blocks:

In RGB space the image is split up into red, blue and green images. The image is then divided into blocks of pixels and accordingly the image of pixels will contain blocks. Where, , .

Phase 2: DCT: All values are level shifted by subtracting 128 from each value. The Forward Discrete CosineTransform of the block is then computed. The mathematical formula for calculating the DCT is:

Where,

Where

Phase 3: Quantization: Quantization is the step where the most of the compression takes place. DCT really does notcompress the image, as it is almost lossless. Quantization makes use of the fact that, the high frequency componentsare less important than the low frequency components. The Quantization output is

The matrix could be anything, but the JPEG committee suggests some matrices which work well withimage compression.




18

Phase 4: Compression using SEQUITUR: After quantization, the scheme uses a filter to pass only the string of non-zero coefficients. By the end of this

process we will have a list of non-zero tokens for each block preceded by their count.DCT based image compression using blocks of size 8x8 is considered. After this, the quantization of DCTcoefficients of image blocks is carried out. The SEQUITER compression is then applied to the quantized DCTcoefficients.The compression achieved in this approach is evaluated based on the overall compression ratio (CR) which isdefined as:

C.R. =

Phase 5: Encryption using EHDES:

In encryption phase, EHDES take output of compression phase as a message block and a new generated keyimplement encryption process as per traditional DES.

In this process, New key is also make 16 different key for every round of EHDES using shifting property as pertraditional DES. For every block of message M, new key makes a new key block for every round of DES toimplement in the encryption process.

Decryption Process is the inverse step of encryption process. In decryption, we also use the same key whichis used in encryption.

{mi} and {Ci},where 1≤ i ≤ n.

Cipher Text and Plain Text

Figure 3: Block Diagram of Proposed Scheme.

4. Security Analysis

We verified that the compression ratio of Sequitur outperforms Gzip as well as Compress. On the other hand,

however, the compression and decompression are very slow compared to Gzip and Compress, because Sequiturutilizes the arithmetic coding that is time consuming, and the program might not be fully optimized. From our viewpoint of compressed pattern matching, compression time is not a serious matter, while the decompression time iscritical. In the original program of Sequitur, decompression routine borrows the same data structures, such as doublylinked list, that are unnecessary for decompression only. Thus we simply rewrote the decompression routine using astandard array.Cryptographic scheme strength is often described by the bit length of encryption key. The more bits in the key, theharder it is to decrypt data simply by all possible key. DES uses 56 bit, Cracking 56- bit algorithm with a single keysearch might take around a week on a very powerful computer.

Divide Source

Image into n

Blocks

Phase 1

EHDES

Phase 5

DCT

Transform

Phase 2

Quantization

Phase 3

SQUITER

Phase 4

Source

Image

Output – A compressed

and encrypted object




19

Now,At time t, the generated key is,At time t + 1, the generated key is ,And At time t + n, the generated key is

Here,≠

It might be possible that are equal if and only if the generated no. NRNG at time t, t + 1, t + nare same.

5. Conclusion

Our scheme certainly provides the required high level security measures and increase the capacity of communicationchannel up to the extent of satisfactory level like for saving human life in telemedicine, military communication andcommunication between secret agents of governments as well as the personal use of secure communication.

References

[1] G. Lo-varco,W. Puech, and M. Dumas. “Dct-based watermarking method using error correction codes”, InICAPR’03, International Conference on Advances inPattern Recognition, Calcutta, India, pages 347–350, 2003.

[2] R. Norcen, M. Podesser, A. Pommer, H.P. Schmidt, and A. Uhl. “Confidential storage and transmission of medical image data”, Computers in Biology and Medicine, 33:277–292, 2003.

[3] Diego F. de Carvalho, Rafael Chies, Andre P. Freire, Luciana A. F. Martimiano and Rudinei Goularte, “VideoSteganography for Confidential Documents: Integrity, Privacy and Version Control” , University of Sao Paulo –

ICMC, Sao Carlos, SP, Brazil, State University of Maringa, Computing Department, Maringa, PR, Brazil.[4] Ramveer Singh , Awakash Mishra and D.B.Ojha “An Instinctive Approach for Secure Communication –

Enhanced Data Encryption Standard (EHDES)” International journal of computer science and Information

technology, , Vol. 1 (4) , 2010, 264-267[5] D.B. Ojha, Ramveer Singh, Ajay Sharma, Awakash Mishra and Swati Garg “An Innovative Approach to

Enhance the Security of Data Encryption Scheme” International Journal of Computer Theory and Engineering,Vol. 2,No. 3, June, 2010,1793-8201

[6] Nameer N. EL-Emam, “Hiding a Large Amount of Data with High Security Using Steganography Algorithm” Applied Computer Science Department, Faculty of Information Technology, Philadelphia University, Jordan

[7] Borie J., Puech W., and Dumas M., “Crypto-Compression System for Secure Transfer of Medical Images”, 2nd International Conference on Advances in Medical Signal and Information Processing (MEDSIP 2004),

September 2004.

[8] N.Walkinshaw, S.Afshan, P.McMinn ”Using Compression Algorithms to Support the Comprehension of Program Traces” Proceedings of the International Workshop on Dynamic Analysis (WODA 2010) Trento, Italy,July 2010.

Ramveer Singh, Bachelor of Engineering from Dr. B.R. Ambedkar University, Agra (U.P.), INDIA in 2003. Masterof Technology from V.M.R.F. Deemed University, Salem (T.N.), INDIA in 2007. Persuing Ph.D from SinghaniaUniversity, Jhunjhunu, Rajsthan, INDIA. The major field of study is Cryptography and network security. He hasmore than eight year experience in teaching and research as ASSOCIATE PROFESSOR. He is working at RajKumar Goel Institute of Technology, Ghaziabad (U.P.), INDIA. The current research area is Cryptography and

Network security. Mr. Singh is the life-time member of Computer Society of India and Computer Science TeacherAssociation.

Dr. Deo Brat Ojha, Ph.D from Department of Applied Mathematics, Institute of Technology, Banaras HinduUniversity, Varansi (U.P.), INDIA in 2004. His research field is Optimization Techniques, Functional Analysis &Cryptography. He has more than Six year teaching & more than eight year research experience. . He is working as aProfessor at Raj Kumar Goel Institute of Technology, Ghaziabad (U.P.), INDIA. He is the author/co-author of morethan 50 publications in International/National journals and conferences.




20

Cloud Computing Approaches for Educational Institutions

N.Mallikharjuna Rao1, V.Satyendra Kumar

2, Sudhakar

3, P.Seetharam

4

1. Associate Professor, Annamacharya P.G college of Computer Studies, Rajampete-mail: [email protected]

2. Assistant Professor, Annamacharya Institute of Technology and Sciences, Rajampet,e-mail: [email protected]. Assistant Professor, Annamacharya P.G college of Computer Studies, Rajampet,

e-mail: [email protected]. Systems Engineer, Annamacharya Institute of Technology and Sciences, Rajampet,

email:[email protected]

Abstract

Cloud computing is a deployment model leveraged by IT in order to reduce infrastructure costs and scalabilityconcerns. Cloud computing is not about the application itself; it is about how the application is deployed as how it isdelivered Cloud ware is an extension of cloud computing but they do not enable business to leverage cloud

computing. Cloud computing is rapidly increasing and attracting popularity. Companies such as Red Hat, Microsoft,Amazon, Google, and IBM are increasingly funding cloud computing infrastructure and research, making it important for students to gain the necessary skills to work with cloud-based resources. Cloud Computing refers toboth the applications delivered as services over the Internet and the hardware and systems software in thedatacenters that provide those services. The services themselves have long been referred to as Software as a Service(SaaS). When a Cloud is made available in a pay-as-you-go manner to the general public, we call it a Public Cloud;the service being sold is Utility Computing. In this paper, we are proposing cloud computing for to use ineducational institutions for improving quality in education. Cloud computing is a simple idea, but it can have ahuge impact on business.

Keywords: Cloud Computing, IaaS, SaaS, HaaS




21

I. Introduction

Cloud Computing has many antecedents and equally as many attempts to define it. The players in the large world of clouds are Software as Service providers, outsourcing and hosting providers, network and IT infrastructureproviders and, above all, the companies whose names are closely linked with the Internet's commercial boom. But,all these services in combination outline the complete package known as Cloud Computing – depending on thesource with the appropriate focus. That which long ago established itself in the private environment of the Internet is

now, noticeably, coming to the attention of businesses too. Not only developers and startups but also largecompanies with international activities recognize that there is more to Cloud Computing than just marketing hype.Cloud Computing offers the opportunity to access IT resources and services with appreciable convenience andspeed. Behind this primarily, is a solution that provides users with services that can be drawn upon on demand andinvoiced as and when used. Suppliers of cloud services, in turn, benefit as their IT resources are used more fully andeventually achieve additional economies of scale. Additionally, Intel is already taking advantage of external cloudcomputing technologies. IBM has much opportunistic software a service (SaaS) implementations. Our preliminaryexperiences with Infrastructure as a Service (IaaS) suggest that it may be suitable for rapid development and somebatch applications.

Many applications are not suitable for hosting in external clouds at present. Good candidates may be applicationsthat have low security exposure and are not mission-critical or competitive differentiators for the corporation. Astrategy of growing the cloud from the inside out delivers many of the benefits of cloud computing and positions usto utilize external clouds over time. IBM expect to selectively migrate services to external clouds as supplier

offerings mature, enterprise adoption barriers are overcome, and opportunities arise for improved flexibility andagility as well as lower costs. A strategy of growing the cloud from the inside out delivers many of the benefits of cloud computing and positions us to utilize external clouds over time.

In this paper, we are proposing to use cloud computing in educational institutions especially in engineering and

management fields, those who are not able to invest large amount of money on softwares for improving the

student ability and quality in advanced software, to meet the requirements of business industry and Information

Technology sector. In Section-II we have discussed background of cloud computing, Section-III presented

several approaches of cloud computing techniques and at the end concluded about the advantages with cloud

computing technology.

II Background

Cloud computing is about how an application or service is deployed and delivered. It's really about the behavior of the entire infrastructure; how the cloud delivers an application,

Dynamism: This is the ability of the application delivery infrastructure to expand and contract automatically basedon capacity needs. Note that this does not require virtualization technology, though many Providers are usingvirtualization to build this capability. There are several approaches for implementing dynamism in architecture.

Abstraction: Do you need to care about the underlying infrastructure when developing an application fordeployment in the cloud. If you have to care about the operating system or any piece of the infrastructure, it's not

abstracted enough to be cloud computing.

Resource Sharing: The architecture must be compute and network resources of the cloud infrastructure are sharableamong applications. This ties back to dynamism and the ability to expand and contract as needed. If an application'smethod of scaling is to simply add more servers on which it is deployed rather than be able to consume resources onother servers as needed, the infrastructure is not capable of resource sharing.

Provides A Platform: Cloud computing is essentially a deployment model. If it provides a platform on which you

can develop and deploy an application and meets the other two criterion of cloud computing. 2.1 Compute resources

CPU time

Cores & clock cycles per

Floating point processing vs. Integer processing

RAM (MBytes used)




22

Access speed / latency time

Data storage (GBytes used)

I/O throughput (MBytes per second Transactions (I/O operations per second), Resilience (e.g. RAID level)

, Bandwidth (GBytes transferred)

2.2 Latency & resilience

According to the IEEE Computer Society Cloud Computing is: "A paradigm in which information is permanentlystored in servers on the Internet and cached temporarily on clients that include desktops, entertainment centers,table computers, notebooks, wall computers, handhelds, etc."Cloud computing means and open market for computing resources; utility computing applied to multiple grids.A compute cloud is a grid spanning multiple administrative domains with applications able to move betweendomains in response to cost and SLA requirements. Cloud is about scale and the computing resource market. The role of the open-source community especially for Cloud Computing Modern Clouds mostly consists of alreadyexisting, well-known and often open-source components. For every aspect of a Cloud environment a different set of utilities is used e.g. Puppet for automated configuration management. This paper explores whether cloudcomputing services are suitable for highperformance computing (HPC) workloads. In contrast, web serviceworkloads that often have little intra-cluster communication are the primary users of cur-rent cloud computing services. However, cloud nodes are typically con-gured to run user-provided software so that cloud computing nodescan just as easily run scientific applications. The ability to quickly create and scale-up a custom compute cluster is aboon to individual scientists whose computing needs can be sporadic. Cloud computing services can also be used toextend existing clusters for larger problem sizes. Although cloud providers currently place small bounds ondynamically allocated resources, trends point to-ward increasing bounds on these resources over time.

Cloud computing is often associated with a couple of acronyms:

• SaaS – Software as a Service; and,

• HaaS – Hardware as a Service.

Successful SaaS examples include: Salesforce.com and Google’s Google Reader and Google Docs which are

useful for engineering and management students for sharing the useful information to develop useful projects.

Successful HaaS examples include: Amazon’s Elastic Compute Cloud provides compute capacity in the cloud.

III Several Approaches for Cloud Computing

The value of the Cloud Computing industry is expected to reach $100 Billion in five years so it’s easy to understandwhy the big IT vendors like Microsoft, Google, and Amazon are rapidly ramping up cloud computing platforms and

services.

3.1 Tribulations with the Predictable Solution

We need to invest in infrastructure to implement the mechanisms necessary for continued integration. Also, we needto set up machines for carrying out the tasks like analyzing the source code, running the tests et al. naturally, thisleads to up- front capital expenditure as well as operational expenditure. As the modern software systems are gettingsizable, tasks like analyzing the source code and the large test base is implying lengthy test execution cycles! Undersuch circumstances, it would become difficult to ensure that the feedback is rapid. This would get in the way of ensuring the required agility and would also hamper the time to market. The problem can be summed up in just aline; given that doing this stuff takes Time, Money and Resources, how do we ensure that we create value instead of

incurring costs? Cloud Computing is an innovation we can apply to bring economy to our solution in terms of time,resources and of course, the money!

3.2 Infrastructure as a Service.

IPs manages a large set of computing resources, such as storing and processing capacity. Through virtualization,they are able to split, assign and dynamically resize these re-sources to build ad-hoc systems as demanded bycustomers, the SPs. They deploy the software stacks that run their services. This is the Infrastructure as a Service (IaaS) scenario.




23

3.3 Platform as a Service.

Cloud systems can offer an additional abstraction level: instead of supplying a virtualized infrastructure, they canprovide the software platform where systems run on. The sizing of the hardware resources demanded by theexecution of the services is made in a transparent manner. This is denoted as Platform as a Service (PaaS). A well-known example is the Google Apps Engine.

3.4 Software as a Service.

Finally, there are services of potential interest to a wide variety of users hosted in Cloud systems. This is analternative to run applications locally. An example of this is online alternatives of typical office applications such asword processors. This scenario is called Software as a Service (SaaS).

IV Physical Components of a Cloud Computing

Cloud computing is a paradigm shift in the way scalable applications are architected and delivered. Since decades,enterprises have spent time and resources building an infrastructure that could provide them a competitiveadvantage. In most cases this approach resulted in:

Large tracts of unused computing capacity

Dedicated resources for server maintenance

Risk mitigation & energy utilization High cost for build, acquire, own & maintain (Total cost of ownership)

With cloud computing, excess computing capacity can be put to use and be profitably sold to consumers. This transformation of computing and IT infrastructure into a utility, which could be available to all, is the basis of cloudcomputing. It forces competition based on innovation rather than computing resources. There are different coloredclouds present in the computing space today which could be classified into the following components:

4.1 Infrastructure

Cloud Infrastructure is the concept of providing “Hardware as a service” i.e. shared or reusable hardware for aspecific time of service. Examples include virtualization, grid computing, and pervert utilization. This service helpsreduce maintenance and usability costs, considering the need for infrastructure management & upgrade.

4.2 Storage Cloud Storage is the concept of separating data from processing and storing in a remote place. Cloud Storage alsoincludes database services. Examples are Google’s Big Table, Amazon’s SimpleDB etc. software needs. This couldbe a single service platform or a solution stack. Examples include Web application frameworks, Web hosting as we

shown in figure 1.

4.3 Application

A Cloud Application is an offering of software architecture that eliminates the need to install, run and maintain anapplication at the user’s desktop/device. A cloud application eliminates the cost/resources required to maintainand/or support applications.




24

4.4 Services

A Cloud Service is an independent piece of software which can be used in conjunction with other services toachieve an interoperable machine- to-machine interaction over the network. Examples include Amazon’s SimpleQueue Service, Google maps, Amazon’s flexible payment service etc.

4.5 Client

Cloud Client is a requester software or hardware device which tries to utilize cloud computing services over thenetwork. The client device could be a Web browser, PC, laptop or mobile etc.

4.6 Characteristics of Cloud Service

Standardized IT-based capability and it has computed, network or software-based capabilities. Standard offering

defined by the services provider with little or no flexibility for customization outside the offering,

4.6.1 Accessible via Internet protocol from any computer

Modern Internet-type protocols over IP, such as HTTP, Representational State Transfer (REST), or Simple Object

Access Protocol (SOAP) that are part of any modern operating system. Always available, and scales automaticallyto adjust to demand

4.6.2 Resilient and highly available

Service provider offers massive capacity, such that any given customer can get as much capacity as they need at agiven moment- and give it back when not needed.

4.6.3 Pay-per-use

Free or pay-per-use, usually without long-term contracts, setup charges, or exit fees. The service is paid for one of three ways these are:

Advertising, usually for consumers

Subscription,

Billed by availability per unit of time, such as a month

Transaction, billed for actual usage, such as minutes of compute time, gigabytes of network bandwidth, or gigabytes of storage.

4.6.4 Web or programmatic-based control interfaces

Cloud-oriented Web sites with human interfaces host the customers data provide interactions with others, and offer arich internet application interface, such as Face book or Microsoft Virtual Earth 3D.

Figure 1: Components of Cloud Computing




25

V Advantages and Benefits from a cloud Computing

With the different Cloud enabler technologies like utility computing, Grid Computing, RTI, web infrastructure and others maturing, the different services would be cloud enabled.

Infrastructure service providers are taking advantage of the paradigm and offering Cloud Services. Cloudcomputing is considered an extension to SOA and SaaS.

Information services, entertainment-oriented services such as video on demand, simple business servicessuch as customer authentication or identity management and contextual services such as location ormapping services are positioned well to become cloud-delivered.

Other services, such as corporate processes (for example, billing, deduction management and mortgagecalculation) and transactional services (for example, fiscal transactions), would take longer to reach thecloud and the mainstream.

5.1 Motivating Forces for Change

There are many forces at play that are creating a ground swell for the change that is enabling the growth of CloudComputing. Some forces are technology advancements which have been evolving over the past decade, whileothers include market and economic forces that are more recent and can be more potent in specific vertical markets.

The technology developments by themselves are not significant but the culmination of all them has profound effects

across industries. The following list describes some of the reasons for the change and also provides insights on howyou can adjust to and thrive in this dynamic environment.

1. Telecommunications – With the .COM boom of the late 90s, there were fiber optics and high bandwidth data infrastructure put in place for global connections. This allowed access to fast speed voice and data connectivity. The.COM bust during the start of the millennium affected many of the companies that established this infrastructure byforcing them out of business. This offered the fiber optics and equipment for whole sale or at a market discount. Thiscreated cost effective opportunities for remote computing in many environments and in particular for thedevelopment of Cloud Computing.

2. Open Source – The decentralized approach to software development challenged the institutionalized form of software development. It allowed for individuals to contribute program code updates to form sophisticated operating

systems such as Linux that runs on most web servers and many computer electronic devices. Open source softwarealso played a significant role in web server technologies that form the core components of Cloud Computing. Inaddition to the Linux operating system that was first installed on commodity hardware, apache web servers providedweb services. The server side software on these web servers along with middle ware including client side XML based components are all open source forming the foundation for many Cloud Computing applications. As per the

AICTE norms and standards all the technical institutes are required to purchase costly software’s like MATLAB,

Prowess and SPSS and some SAP applications. Through open source facility in cloud computing we will get

easily for rent and use what ever software as well as hardware need for educational institutions.

3. Social Network – What was once viewed as a social experiment that college kids did with Face book or a group of eccentric academics tinkering with their Wikipedia has matured into a powerful business tool. The same principlethat drove the social networks to connect, share and learn has become a powerful tool for collaboration, marketingand other business applications. Organizations develop new features to their products, get ratings, and connect totheir customers with greater efficiencies compared to the traditional print and broadcast media. The success of Cloud Computing leverages on the implementation of social networks. Cloud Computing shares some of the samecore technologies and therefore works symbiotically with social networking.

4. Cloud Computing Services – As the technologies and processes mature and large companies such as Amazon,Google and IBM operate large operations using Cloud Computing, they are starting to put together offerings toexternal organizations and individuals. These pioneering organizations have worked out many of the challenges of implementing Cloud Computing from the IT hardware infrastructure to the software and development tools needed.More and more services are beginning to be offered to businesses and individuals to implement new CloudComputing services based on the existing infrastructure so you do not have to re-invent the wheel. This can be seenat all levels of the traditional software market including consumer based individuals extending to large enterprisesolutions.




26

5. Economic Down Turn – The economic down turn has put a strain on IT budgets and refrained capital fromimplementing large systems. Companies still need to deliver services to their customers and IT departments’ demandsolutions for their customers. However, the credit crunch has halted many projects that require a huge setup cost.Rather than purchasing or committing to large systems, businesses can only spend a small amount on critical tasks

to operate their business

6. Outsourcing IT – With the economic environment, companies turn to outsourcing as a way of cutting cost to

maintain competitiveness. Various different parts of IT are outsourced. Support call centers can provide cost savingswhen supporting their customers. In order to cut cost within their own organization, the ability to outsource theimplementation of software systems is another layer of services which can be outsourced. Educational institutionsare also the use of service of IT outsourcing for improving the industry norms and standards. This frees the resources of an organization to focus on core business. Cloud Computing will be the enabler and the fuel to this type of software as a service (SaaS).

5.2 Cloud Attributes and Taxonomy

Because this is an emerging and somewhat confusing area, we have created definitions that provide us with acommon basis for discussion and developing our strategy. We define cloud computing as a computing paradigmwhere services and data reside in shared resources in scalable data centers, and those services and data are accessibleby any authenticated device over the Internet.

We have also identified some key attributes that distinguish cloud computing from conventional computing. Cloudcomputing offerings are:

Abstracted and offered as a service.

Built on a massively scalable infrastructure.

Easily purchased and billed by consumption.

Shared and multi-tenant.

Based on dynamic, elastic, flexibly configurable resources.

Accessible over the Internet by any device.

Today, we have identified three main categories of external service that fall within our broad cloud computingdefinition. Software as a service (SaaS). Software deployed as a hosted service and accessed over the Internet.

Platform as a service (PaaS): Platforms that can be used to deploy applications provided by customers or partners of the PaaS provider. Infrastructure as a service (IaaS): Computing infrastructure, such as servers, storage, and network,delivered as a cloud service, typically through virtualization. It is also possible to build an internal IT environmentwith cloud computing characteristics. We call this an internal cloud, to differentiate it from the external cloudsprovided by suppliers.

5.3 Software as a Service Benefits

This outsourcing approach has enabled us to take advantage of suppliers’ expertise. The amount of data regularlytransmitted between Intel and SaaS providers has been a huge challenge, causing difficulties during initialdeployments and upgrades. Testing solutions has also provided challenges, demonstrating the need for fulldocumentation and up-front clarification with suppliers about roles, responsibilities, and process. Overall, SaaS has been successful in our environment and met Intel’s expectations for the intended use of the services.

• Save money by not having to purchase servers or other software to support use• Focus Budgets on competitive advantage rather than infrastructure

• Monthly obligation rather than up front capital cost

• Reduced need to predict scale of demand and infrastructure investment up front as available capacitymatches demand

• Multi-Tenant efficiency

• Flexibility and scalability




27

Software’s available are:

Platforms: Windows, Linux, Solaris Oracle, Sybase, PostGres Core Java, JSP and JDBC BV, VB.NET Java Script, VB Script, HTML, XML, XSLT,

Visual Studio.Net, Visio ANT, WinCVS Rational Rose, Enterprise Architect

5.4 Infrastructure as a Service

Intel uses IaaS for certain niche applications. For example, some of the content on Intel’s Web site is hosted by acloud service provider. This allows us to take advantage of the supplier’s worldwide infrastructure rather than facingthe expense and difficulty of building similar infrastructure ourselves.

We also gained experience with IaaS when we build a globally distributed Web-monitoring application. Intel needed a service that would allow us to look at visitors’ experiences accessing the Intel Web site from differentregions of the globe.

We were initially concerned about the possibility of highly variable and unreliable performance. We found that the Cloud Computing was highly available and it provides all type hardware and infrastructural facilities to the all end-users. We were pleased to find that the time on the cloud computing was kept reasonably accurate.

Our experience suggests that once security and manageability concerns are addressed, current commercial IaaSimplementations may be good for rapid prototyping and compute-intensive batch jobs. After IaaS services provethemselves for these applications, they could be considered for more demanding applications with stringent response-time requirements.

“Cloud Computing is more an evolution than a revolution.”

Conclusion Cloud computing is style of computing in which dynamically scalable and often virtualized resources are providedas a service over the internet. Users need not have knowledge of, expertise in, or control over the technologyinfrastructure in the “Cloud” that supports them. The more complex software becomes, ironically, the more simple

the user machines need to be to run them. This is because the goal is to simplify the usage and managementrequirements so that the user can access sophisticated software from any computer with an internet access. The stepsneeded to implement a successful solution with this Cloud Computing environment requires changes in howsoftware is implemented both on the client computer and also on the server. The foundation for changes on theclient machine has already taken shape due to ecommerce and social technologies. The challenge remains for

Figure 2: Infrastructure as Service Components




28

complex solutions such as SaaS to adjust and adapt on the server side to function as a service within CloudComputing. In this paper supports well to use cloud computing SaaS and IaaS are in educational institutions for

improving the educational quality.

References

[1] Google app engine web site. Web Resource, Sept 2008.

[2] Amazon simple storage service. Web Page http://www.amazon.com/gp/browse.html?node=16427261.[3] Mark Baker, Amy Apon, Clayton Ferner, and Jeff Brown. Emerging grid standards. Computer, (4):43–50, April2005.

[4] Miguel L. Bote-Lorenzo, Yannis A. Dimitriadis, and Eduardo G´omez-S´anchez. Grid characteristics and uses: agrid defining. Pages 291–298, February 2004.

[5] Roy Bragg. Cloud computing: When computers really rule. Tech News World, July 2008. Electronic Magazine,available at http://www.technewsworld.com/story/63954.html .

[6] Rajkumar Buyya, Chee Shin Yeo, and Srikumar Venugopal. Market-oriented cloud computing: Vision, hype, and reality for delivering it services as computing utilities. CoRR,(abs/0808.3558), 2008.

[7] Brian de Haaff. Cloud computing - the jargon is back! Cloud Computing Journal, August 2008. ElectronicMagazine, article available at http://cloudcomputing.sys-con.com/node/613070

[8] Kemal A. Delic and Martin Anthony Walker. Emergence of the academic computing clouds. ACM Ubiquity,(31), 2008.

[9] Flexi scale web site. http://www.flexiscale.com, last visited: August 2008.




29

Preprocessing of Digital Mammogram for Image Analysis

Prof. Samir Kumar Bandyopadhyay and Indra Kanta MaitraDept. of Computer Sc. & Engg, University of Calcutta

92 A.P.C. Road, Kolkata – 700009, India

Email: [email protected], [email protected]

AbstractMammography is at present one of the available method for early detection of abnormalities, which is related to

human breast cancer. Raw digital mammograms are medical images that are difficult to interpret, thus a preparationprocess is needed in order to improve the image quality and make the segmentation results more accurate for furtherresearch. Three distinct steps have been proposed for preprocessing of digital mammogram to make it ready toautomatic detection of abnormalities in breast. The first step involves contrast enhancement. The Contrast LimitedAdaptive Histogram Equalization (CLAHE) technique is used in the process with 8X8 tiles, Clip Limit of 0.01,histogram bins of 64 and distribution is Rayleigh. The second step identifies the region of interest (ROI) in medio-lateral oblique (MLO) view of mammogram to determine the pectoral muscle. The third or final step is to suppressthe pectoral muscle using proposed modified seeded region growing (SRG) algorithm. The pectoral musclesuppression, we have obtained 84% of near accurate result including accurate results on selected 50 numbersmammograms of MIAS Database of different size, shapes and types. The said techniques facilitate for further imageanalysis without destroying required identification features to detect the abnormalities of the digital mammogram. Tosummarize, the results obtained by the method show that it is a robust approach but it can be improved in terms of accuracy.

Keywords: Mammogram, Contrast Limited Adaptive Histogram Equalization (CLAHE), Region of Interest (ROI),seeded region growing (SRG), Noise, Edge-Shadowing Effect and Pectoral Muscle




30

1. IntroductionCancer is a group of diseases that cause cells in the body to change and grow out of control. Most types of

cancer cells eventually form a lump or masses called a tumor, and are named after the part of the body where thetumor originates. Breast cancer begins in breast tissue, which is made up of glands for milk production, calledlobules, and the ducts that connect lobules to the nipple. The remainder of the breast is made up of fatty, connective,and lymphatic tissue [1].

Breast cancer is a leading cause of cancer deaths among women. For women in US and other developedcountries, it is the most frequently diagnosed cancer. About 2100 new cases of breast cancer and 800 deaths areregistered each year in Norway. In India, a death rate of one in eight women has been reported due to breast cancer[2].

Efficient detection is the most effective way to reduce mortality, and currently a screening programme based onmammography is considered one the best and popular method for detection of breast cancer. Mammography is alow-dose x-ray procedure that allows visualization of the internal structure of the breast. Mammography is highlyaccurate, but like most medical tests, it is not perfect. On average, mammography will detect about 80%-90% of thebreast cancers in women without symptoms. Testing is somewhat more accurate in postmenopausal than inpremenopausal women [3]. The small percentage of breast cancers that are not identified by mammography may bemissed for just as mammography uses x-ray machines designed especially to image the breasts. An increasingnumber of countries have started mass screening programmes that have resulted in a large increase in the number of mammograms requiring interpretation [4]. In the interpretation process radiologists carefully search each image for

any visual sign of abnormality. However, abnormalities are often embedded in and camouflaged by varying densitiesof breast tissue structures. Estimates indicate that between 10 and 30 per cent of breast radiologists miss cancersduring routine screening [4,5]. In order to improve the accuracy of interpretation, a variety of screening techniqueshave been developed.

Recent studies showed that the interpretation of the mammogram by the radiologists give high rates of falsepositive cases indeed the images provided by different patients have different dynamics of intensity and present aweak contrast. Moreover the size of the significant details can be very small. Several research works have tried todevelop computer aided diagnosis tools. They could help the radiologists in the interpretation of the mammogramsand could be useful for an accurate diagnosis [6,7,8].

Digital mammography is a technique for recording x-ray images in computer code instead of on x-ray film, aswith conventional mammography. The first digital mammography [9] system received U.S. Food and DrugAdministration (FDA) approval in 2000. Digital mammography may have some advantages over conventionalmammography. The images can be stored and retrieved electronically. Despite these benefits, studies have not yetshown that digital mammography is more effective in finding cancer than conventional mammography [10].

Imaging techniques play an important role in helping perform digital mammogram, especially of abnormal areasthat cannot be felt but can be seen on a conventional mammogram [11]. Before any image-processing algorithm of mammogram preprocessing steps are very important in order to limit the search for abnormalities without undueinfluence from background of the mammogram. These steps are needed only on digitized screen film mammography(SFM) images because digital mammography devices perform this step automatically during the image storingprocess. Breast segmentation consists of breast border contour extraction, pectoral muscle extraction, nippleidentification etc. On images obtained directly from the digital mammography devices segmentation process is muchmore easier. Previous work from many authors used mammography image databases including this paper, especiallyMiniMIAS [12] and DDSM [13], both comprised of scanned and digitized SFM images. In this paper we haveproposed three steps of preprocessing of raw digital mammogram for future automatic elaborate investigation of abnormalities using some image processing technique.

2. Image SegmentationThe paper is based on the image segmentation method, which refers to the major step in image processing, the

inputs are images and, outputs are the attributes extracted from those images. Segmentation divides image into itsconstituent regions or objects. The level to which segmentation is carried out depends upon the problem being solvedi.e. segmentation should stop when the objects of interest in an application have been isolated. Image segmentationrefers to the decomposition of a scene into its components. Image segmentation techniques can be broadly classifiedas into five main classes threshold based, Cluster based, Edge based, Region based, Watershed based segmentation[14].




31

Segmentation plays an important role in image analysis. The goal of segmentation is to isolate the regions of interest depending on the problem and its characters. Many applications of image analysis need to obtain the regionsof interest before the analysis can start. Therefore, the need of an efficient segmentation method has always beenthere. A gray level image consists of two main features, namely region and edge.

Segmentation algorithms for gray images are generally based on of two basic properties of image intensityvalues, discontinuity and similarity. In the first category, the approach is to partition an image based on abrupt

changes in intensity, such as edges in an image. The principle approaches in the second category are based onpartitioning image into regions that similar according to a set of predefined criteria. Thresholding, region growingand region splitting and merging are examples of the methods in this category.

3. Region GrowingFor the segmentation of intensity images like digital mammogram, there are four main approaches [15], [16],

namely, threshold techniques, boundary-based methods, region-based methods, and hybrid techniques whichcombine boundary and region criteria.

Threshold techniques [17] are based on the postulate that all pixels whose value (gray level, color value, orother) lies within a certain range belong to one class. Such methods neglect all of the spatial information of the imageand do not cope well with noise or blurring at boundaries. Boundary-based methods [18] use the postulate that thepixel values change rapidly at the boundary between two regions. The complement of the boundary-based approachis to work with the regions [19]. Region-based methods rely on the postulate that neighboring pixels within the one

region have similar value. This leads to the class of algorithms known as region growing of which the “split andmerge” technique [20] is probably the best known. The general procedure is to compare one pixel to its neighbor(s).If a criterion of homogeneity is satisfied, the pixel is said to belong to the same class as one or more of its neighbors.The fourth type is the hybrid techniques, which combine boundary and region criteria. This class includesmorphological watershed segmentation [21] and variable-order surface fitting [15]. The watershed method isgenerally applied to the gradient of the image.

We mainly use a method known as “seeded region growing” (SRG), which is closer to that of the watershed[22] with some necessary change in proposed technique which is based on the conventional region-growing postulateof similarity of pixels within a regions. For Seeded Region Growing (SRG), seed or a set of seed can beautomatically or manually selected. Their automated selection can be based on finding pixels that are of interest, e.g.the brightest pixel in an image can serve as a seed pixel. They can also be determined from the peaks found in animage histogram. On the other hand, seeds can also be selected manually for every object present in the image. Themethod is employed to segment an image into different regions using a set of seeds. Each seeded region is aconnected component comprising of one or more points and is represented by a set S. The set of immediateneighbours bordering the pixel is calculated. The neighbours are then examined and if they intersect any region fromset S, then a measure δ (difference between a pixel and the intersected region) is computed. If the neighboursintersect more than one region, then the set is taken as that region for which difference measure δ is maximum. Thenew state of regions for the set then constitutes input to the next iteration. This process continues until the entireimage pixels have been assimilated into regions. Hence for each iteration the pixel that is most similar to a regionthat it borders is appended to that region. The SRG algorithm is inherently dependent on the order of processingimage pixels. The method has the advantage that it is fairly robust, quick, and parameter free except for itsdependency on the order of pixel processing.

4. Previous Works on SRG AlgorithmMehnert & Jackway (1997) improved the above seeded region-growing algorithm by making it independent of

the pixel order of processing and making it more parallel. Their study presents a novel technique for ImprovedSeeded Region Growing (ISRG). ISRG algorithm retains the advantages of Seeded Region Growing (SRG) such asfast execution, robust segmentation and no parameters to tune. The algorithm is also pixel order independent. If morethan one pixel in the neighbourhood has same minimum similarity measure value, then all of them are processed inparallel. No pixel can be labelled and no region can be updated until all other pixels with the same priority have beenexamined. If a pixel cannot be labelled, because it is equally likely belong to two or more adjacent regions, then it islabelled as ‘tied’ and takes no part in the region growing process. After all of the pixels in the image have beenlabelled, the pixels labelled ‘tied’ are independently re-examined to see whether or not the ties can be resolved. Toresolve the ties an additional assignment criterion is imposed, such as assigning a tied pixel to the largest




32

neighbouring region or to the neighbouring region with the largest mean. ISRG algorithm produces consistentsegmentation because it is not dependent on the order of pixel processing. Parallel processing ensures that the pixelswith the same priority are processed in the same manner simultaneously [23].

Beaulieu & Goldberg (1989) proposed a hierarchical stepwise optimisation algorithm for region merging, whichis based on stepwise optimisation and produces a hierarchical decomposition of the image. The algorithm starts withan initial image partition into a number of regions. At each iteration, two segments are merged provided they

minimise a criteria of merging a segment to another. In this stepwise optimisation, the algorithm searches the wholeimage context before merging two segments and finds the optimal pair of segments. This means that the most similarsegments are merged first. The algorithm gradually merges the segments and produces a sequence of partitions. Thesequences of partitions reflect the hierarchical structure of the image [24].

Gambotto (1993) proposed an algorithm that combines the region growing and edge detection methods forsegmenting the images. The method is iterative and uses both of these approaches in parallel. The algorithm startswith an initial set of seeds located inside the true boundary of the region. The pixels that are adjacent to the regionare iteratively merged with it if they satisfy a similarity criterion. A second criterion uses the average gradient overthe region boundary to stop the growth. The last stage refines the segmentation. The analysis is based on cooperationbetween the region growing algorithm and the contour detection algorithm. Since, adding segments to a region, somepixels that belong to the next region and to the previous region, may be misclassified performs the growing process.A nearest neighbour rule is then used to locally reclassify them [25].

Hojjatoleslami & Kittler (2002) proposed a new region growing approach for image segmentation, which uses

gradient information to specify the boundary of a region. The method has the capability of finding the boundary of arelatively bright/dark region in a textured background. The method relies on a measure of contrast of the region,which represents the variation of the region grey-level as a function of its evolving boundary during segmentation.This helps to identify the best external boundary of the region. The application of a reverse test using a gradientmeasure then yields the highest gradient boundary for the region being grown. The unique feature of the approach isthat in each step at most one candidate pixel will exhibit the required properties to join the region. The growingprocess is directional so that the pixels join the grown region according to a ranking list and the discontinuitymeasurements are tested pixel by pixel. The algorithm is also insensitive to a reasonable amount of noise. The mainadvantage of the algorithm is that no a priori knowledge is needed about the regions [26].

5. Proposed MethodDigital Mammograms are medical images that are difficult to interpret, thus a preparation phase is needed in

order to improve the image quality and make the segmentation results more accurate. The main objective of thisprocess is to improve the quality of the image to make it ready for further processing by removing the irrelevant andunwanted parts in the background of the mammogram. Here we have proposed three distinct phases of preprocessingbefore actual algorithm of automatic analysis of digital mammogram can be conducted. The first phase is to contrastenhancement of the digital mammogram, second phase is to identify region of interest and the third or final phase isto eliminate pectoral tissues from the mammogram.

5.1. Enhancement of Digital MammogramThe contrast enhancement phase is done using the Contrast Limited Adaptive Histogram Equalization (CLAHE)

technique, which is a special case of the histogram equalization technique [27] that functions adaptively on the imageto be enhanced. The pixel's intensity is thus transformed to a value within the display range proportional to the pixelintensity's rank in the local intensity histogram. CLAHE [28] is a refinement of Adaptive Histogram Equalization(AHE) where the enhancement calculation is modified by imposing a user-defined maximum, i.e. clip level, to height

of the local histogram and thus on the maximum contrast enhancement factor. The enhancement is there by reducedin very uniform areas of the image, which prevent over enhancement of noise and reduces the edge-shadowing effectof unlimited AHE [29].

The CLAHE method seeks to reduce the noise and edge-shadowing effect produced in homogeneous areas andwas originally developed for medical imaging [30]. This method has been used for enhancement to remove the noiseand reduces the edge-shadowing effect in the pre-processing of digital mammogram [31].

The CLAHE operates on small regions in the image called tiles rather than the entire image. Each tile’s contrastis enhanced, so that the histogram of the output region approximately matches the uniform distribution or Rayleighdistribution or exponential distribution. Distribution is the desired histogram shape for the image tiles. The




33

neighbouring tiles are then combined using bilinear interpolation to eliminate artificially induced boundaries. Thecontrast, especially in homogeneous areas, can be limited to avoid amplifying any noise and reduce edge-shadowingeffect that might be present in the image; the CLAHE technique is described below [32]:

Step 1: Mammogram was divided into a number of non-overlapping contextual regions of equal sizes, experimentally set to be 8x8, which corresponds to approximately 64pixels.

Step 2: The histogram of each contextual region was calculated.Step 3: A clip limit, for clipping histograms, was set (t=0.001). The clip limit was athreshold parameter by which the contrast of the image could be effectively altered;a higher clip limit increased mammogram contrast.

Step 4: Each histogram was redistributed in such a way that its height did not exceed theclip limit.

Step 5: All histograms were modified by the transformation function

is the probability density function of the input mammogram image grayscale value j, n is the total number of pixels in the input mammogram image and nj is the input pixel number of grayscale value j.

Step 6: The neighboring tiles were combined using bilinear interpolation and themammogram image grayscale values were altered according to the modifiedhistograms.

In our experiment, we defined tiles size i.e. the rectangular contextual regions to 8X8, which is chosen frombest result from trial. Contrast factor that prevents over-saturation of the image specifically in homogeneous areas isrestricted to 0.01 here to get the optimized output. The number of Bins for the histogram building for contrastenhancing transformation is restricted to 64 and the distribution of histogram is 'Rayleigh' i.e. Bell-shaped for ourexperimentation. The range is not specified in the experiment to get the full range of output image.

5.2. Detection and Define ROIMammograms show a projection of the breast that can be made from different angles. The two most common

projections are medio-lateral oblique and cranio-caudal. The advantage of the medio-lateral oblique projection is thatalmost the whole breast is visible, often including lymph nodes. The main disadvantage is part of the pectoral musclewill be shown in upper part of the image, which is superimposed over a portion of the breast. The cranio-caudal viewis taken from above, resulting in an image that sometimes does not show the area close to the chest wall. In ourresearch work we consider the earlier one for its advantage but pectoral muscle detection is one more difficult task inthe breast segmentation process. Reason for detecting pectoral muscle is to remove. Suppression can help in someauto detect procedures such as finding bilateral asymmetry etc.

It is important to detecting the pectoral muscle and defines the region of interest (ROI) for further analysis. Thisoperation is important in medio-lateral oblique (MLO), where the pectoral muscle, slightly brighter compared to therest of the breast tissue, can appear in the mammogram. To detect the same it is very important to determine whetherthe mammogram of the left or the right breast is viewed. We consider right breast first. At first, a straight line (AB) isplotted between the left background of the mammogram and starting of actual part of breast image. In second step, isto determine middle point (C) at the top margin of the mammogram and plot a straight line (CD) from the middlepoint (C) to lower left corner point of the mammogram. The line CD cross the line AB at point E resulting an

inverted right angle triangle (ACE) that is our region of interest (ROI) to detect the pectoral tissues frommammogram, the process is depicted in figure 1 and figure 2. Except the ROI rest part of mammogram is convertedto background color to make said ROI prominent. Finally to reduce the computational complexity inverted rightangle triangle (ACE) is cropped from the original mammogram shown in figure 3 for further processing.




34

Figure 1: The Inverted Right Angle Triangle (ACE)

Figure 2: Detected ROI from Mammogram

Figure 3: Isolated Part of ROI from Mammogram for Further Processing

Now the question will arise that in all the cases the said triangle will cover the entire region of pectoral muscle.It gives absolute success ratio on our set of fifty digital mammography images from twenty-five different pairs of mammogram of different shapes and size. In case of left breast we will consider the mirror image to run the saidprocess.

5.3. Suppression of Pectoral Muscle from ROIWe have used Seeded Region Growing (SRG) algorithm to suppress the pectoral muscle from predefined ROI.

‘Region growing’ is a procedure that groups pixels or sub regions into larger regions based on predefined criteria.The basic approach is to start with sets of “seed” points and from these grow regions by appending to each seed thoseneighboring pixels that have properties similar to the seed. Selection of the seed depends on the nature of theproblem.

This method takes a set of seeds as input along with the image. The seeds mark each of the objects to besegmented. The pixel with the smallest difference measured this way is allocated to the respective region. Thisprocess continues until all pixels are allocated to a region. Another problem in seeded region growing is the




35

formulation of a stopping rule. Basically, growing a region should stop when no more pixels satisfy the criteria forinclusion in that region.

In our proposed method we have used the basic rules of SRG algorithm but introducing some new ideas to makeit more efficient, problem specific and less time consuming. Instead of a seed, a set of seeds has been considered.The left top to right bottom diagonal has been considered to select the seeds, specifically up to the cross point of right top to left bottom diagonal in the cropped image. The double arrow red line in figure 4 is indicating the line of

consideration.

Figure 4: The Double Arrow Red Line of Consideration to Collect the Set of Seeds

The Cartesian slope-intercept equation for the line may be chosen for traversing the line of consideration withend points (x1, y1) and (x2, y2) in mammogram.

Where m representing the slop of the line and b as the y intercept. If we specify the two end points (x1, y1) and(x2, y2) in the mammogram image, we can determine value of the slope and y intercept as following:

For any given x interval δx along a line, we can calculate corresponding y interval δy from the followingequation:

Similarly, we can obtain the x interval δx corresponding to a specified δy as

All pixels along the line of consideration is read one after another from left top to right bottom and calculatesthe Minimum, Maximum and Average Intensity value of the pixel.

Figure 5 : Histogram of Figure 3 showing the Area of Pectoral Muscle Region by Bouble Arrow Black Line




36

Now a selection of pixel for region growing has been introduced. Read the inverted right angle triangle areafrom left top corner. The selection method is subtracting average intensity from that pixel intensity and divided bythe difference between maximum intensity and average intensity. If the pixel intensity is greater than 0 and less thanequals to 1, pixel will be merged to the region growing and the intensity value will be 0 else it will be remainunchanged.

Where the I(x,y) is the intensity of a pixel where IAvg and IMax is average intensity and maximum intensitycalculated earlier from the set of seeds. The region growing will be restricted with in the boundary of the invertedright angle triangle area of the cropped image. Using the method pectoral muscle part of the mammogram will beeliminated. Superimposing the ROI to the original mammogram will show the breast image excepting the pectoralmuscle past, which will be helpful for further investigation of mammogram properly.

6. Experimental Results AnalysisThe mammogram images used in this experiment are taken from the mini mammography database of MIAS

[33]. The original MIAS Database (digitized at 50 micron pixel edge) has been reduced to 200-micron pixel edge andclipped/padded so that every image is 1024 pixels x 1024 pixels. All images are held as 8-bit gray level scale imageswith 256 different gray levels (0-255) and physically in portable gray map (.pgm) format. The list is arranged in pairsof mammograms, where each pair represents the left and right breast of a single patient. In our experiment we haveconsider all types of breast tissues i.e. Fatty, Fatty-glandular, Dense-glandular and the abnormalities likecalcification, well-defined or circumscribed masses, speculated masses and other ill-defined masses. Fifty, medio-lateral oblique (MLO) view of bilateral pairs of mammogram images is used as a test case.

Main objective of the proposed method is restricted to prepare raw mammogram for further analysis to identifythe abnormalities. In the proposed method, uniform contrast enhancement prevents the over enhancement of noiseand reduces the edge-shadowing effect and on the other hand, method of region of interest identification along withmodified region growing algorithm detects of pectoral muscle and suppress the same. The presence of noise, edge-shadowing effect and pectoral muscle can affect results of lesion detection algorithms, so, it is recommended to haveit removed from the image before.

First we used the contrast enhancement algorithm, which produces excellent result by reducing the noise and the

edge-shadowing effect. The raw mammogram and the output of contrast enhancement phase are sited in the figure 6and 8 respectively along with the histogram in figure 7 and 9 justifying the result of proposed technique.

Figure 5: The Original Mammogram Image

Figure 6: The Histogram of Original Mammogram Image




37

Figure 7: The Mammogram Image after Contrast Enhancement

Figure 8: The Histogram of Contrast Enhanced Mammogram Image

In the second phase is to establish the region of interest (ROI) from the contrast-enhanced mammogram to

extract the pectoral muscle. We have tested over 50 mammogram images of different shapes and types and obtained a90% of “near accurate” results, which include the “accurate” results. Following figure 10,11 and 12 are showing theresults on different shapes and types of breast.

Figure 9: Mammogram Showing the Indentified Region of Interest (ROI)





38


The inverted right angle triangle (ACE) of region of interest (ROI) is cropped from the mammogram figure 10,11 and 12 are showing in figure 13 respectively.

Figure 12: Cropped Inverted Right Angle Traingle or the ROI of figure 10, 11 and 12 from left to right

In the final phase, modified seeded region growing algorithm executed on the cropped mammogram images tosuppress the pectoral muscle. Figure 14 showing superimposing the cropped image over the mammogram after

suppression of pectoral muscle.

Figure 13: The Mammogram After Suppression of Pectoral Muscle




39

Figure 14: The Final Output, the Superimposed Original Mammogram after Suppression of Pectoral Muscle

The pectoral muscle suppression, we have obtained 84% of near accurate result including accurate results.Results of pectoral muscle suppression are divided in three groups and showing in Table 1. Total percentage for"Good", "Acceptable" and "Unacceptable" results are shown at the end of each column.

Table 1: Pectoral Muscle Suppression Results

Result Good Acceptable UnacceptableTotal No

(50)23 19 8

Percentage 46% 38% 16%84% 16%

Figure 15: Results of Pectoral Muscle Suppression in Cropped Images

We notice that some of results are a little bit over or under segmented. The behavior of the method shows anover-segmentation of the breast in cases with dense tissue, where the contrast between the muscle and the tissue isunclear. In those cases, our method rejects the muscle detection and provides the region obtained without suppressingthe muscle as a final result. A possible solution could be to impose shape restrictions to the growing process. Tosummarize, the results obtained by the method show that it is a robust approach but it can be improved in terms of accuracy. Even so, we accept this method because it provides encouraging results.

7. ConclusionsA set process of preprocessing has been presented with contrast enhancement, pectoral muscle detection and

suppression. The results obtained over MiniMIAS database have shown a general good behavior. Using thispreprocessing segmentation processes reduce noise and edge-shadowing effect, accurately detect region of interest(ROI) for pectoral muscle, suppress the pectoral muscle successfully. So, the processed mammogram can be used forthe automated abnormalities detection of human breast like calcification, circumscribed masses, speculated massesand other ill-defined masses speculated, circumscribed lesions, asymmetry analysis etc. Further work may beconducted to development to smoothing of the pectoral muscle segmentation, can impose shape restrictions to theregion growing process and that gives better results. This method has the potential for further development becauseof its simplicity and encouraging results that will motivate online or real-time breast cancer diagnosis system further.

References[1] Breast Cancer Facts & Figures, 2009-2010, American Cancer Society, Inc.




40

[2] Norum J. “Breast cancer screening by mammography in Norway. Is it cost-effective?” Ann Oncol 1999, 10:197-203.

[3] Michaelson J, Satija S, Moore R, et al. “The pattern of breast cancer screening utilization and its consequences”Cancer . Jan 1 2002; 94(1): 37-43.

[4] Baines CJ, McFarlane DV, Miller AB. “The role of the reference radiologist: Estimates of interobserveragreement and potential delay in cancer detection in the national screening study”, Invest Radiol 1990, 25: 971-

6.[5] Wallis MG, Walsh MT, Lee JR. “A review of false negative mammography in a symptomatic population”, Clin

Radiol 1991, 44: 13-5.[6] Sterns EE, “Relation between clinical and mammographic diagnosis of breast problems and the cancer/ biopsy

rate,” Can. J. Surg., vol. 39, n°. 2, p 128-132, 1996.[7] Highnam R and Brady M, “Mammographic Image Analysis”, Kluwer Academic Publishers, 1999. ISBN: 0-

7923- 5620-9.[8] Kekre HB, Sarode Tanuja K and Gharge Saylee M, “Tumor Detection in Mammography Images using Vector

Quantization Technique”, International Journal of Intelligent Information Technology Application, 2009,2(5):237-242

[9] FDA Web site, http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfMQSA/mqsa.cfm[10] RadiologyInfo.org developed jointly by Radiological Society of North America and American College of

Radiology

[11] National Cancer Institute (NCI) Web site, http://www.cancernet.gov[12] Suckling J., Parker J., Dance D.R., Astley S., Hutt I., Boggis C.R.M., Ricketts I., Stamatakis E., Cernaez N.,

Kok S.L., Taylor P., Betal D., Savage J., “The Mammographic Image Analysis Society Digital MammogramDatabase”, Proceedings of the 2nd International Workshop on Digital Mammography, York, England, 10–12July 1994, Elsevier Science, Amsterdam, The Netherlands, pp. 375-378.

[13] Heath M., Bowyer K., Kopans D., Moore R., Kegelmeyer P. Jr., “The Digital Database for ScreeningMammography”, Proceedings of the 5th International Workshop on Digital Mammography, Toronto, Canada,11-14 June 2000, Medical Physics Publishing, 2001, pp. 212-218.

[14] Sanmeet Bawa, A thesis on “Edge Based Region Growing”, Department of Electronics and communicationEngimeering, Thapar Institute of Engineering & Technology (Deemed University), India, June – 2006

[15] P. J. Besl and R. C. Jain, “Segmentation through variable-order surface fitting,” IEEE Trans. Pattern Anal.

Machine Intell., vol. PAMI-IO, pp.167-192, 1988[16] R. M. Haralick and L. G. Shapiro, “Image segmentation techniques,” Comput. Vis. Graph. Image Process, vol.

29, pp. 100-132, 1985.[17] P. K. Sahoo, S. Soltani, and A. K. C. Wong, “A survey of thresholding techniques,” Comput. Vis.. Graph.

Image Process, vol. 41, pp. 233-260, 1988.[18] L. S. Davis, “A survey of edge detection techniques,” Compur. Graph. Image Process, vol. 4, pp. 248-270,

1975.[19] S. W. Zucker, “Region growing: Childhood and adolescence,” Comput. Graph. Image Process, vol. 5, pp. 382-

399, 1976.[20] S. L. Horowitz and T. Pavlidis, “Picture segmentation by a directed split-and-merge procedure,” Proc. 2nd Int.

Joint Conf Pattern Recognit , 1974, pp, 424-433.[21] F. Meyer and S. Beucher, “Morphological segmentation,” J. Vis. Commun. Image Represent , vol. 1, pp. 2146,

1990[22] Rolf Adams, et al, "Seeded Region Growing", IEEE Transactions on Pattern Analysis and Machine

Intelligence, VOL. 16, NO. 6, JUNE 1994, pp. 641-647.[23] Mehnert and Jackway P, “An improved seeded region growing algorithm”, Pattern Recog/ Letters. 18, 1997,

pp.1065-1071[24] Beaulieu JM and Goldberg M, A Hierarchy Research article in "Picture segmentation: a stepwise optimisation

approach", IEEE Trans. Pattern Analysis & Machine Intellig. 11 (2), 1989, pp.150-163.[25] Gambotto JP, "A new approach to combining region growing and edge detection", Pattern Recog. Letters. 14,

1993, pp. 869-875.[26] Hojjatoleslami SA and Kittler J, "Region growing: a new approach", Dept. Electric.Electronics Engg., Univ.

Surrey, Guildford, UK, GU25XH, 2002.[27] Gonzalez, R.C., Woods, R.E. (1992), Digital image processing.




41

[28] Pizer SM, “Psychovisual issues in the display of medical images, in Hoehne KH (ed), Pictoral Informationsystems in Medicine”, Berlin, Germany, Springer-Verlag, 1985, PP 211-234.

[29] Pisano ED, et al, “Contrast Limited Adaptive Histogram Equalization Image Processing to Improve theDetection of Simulated Spiculations in Dense Mammograms”, Journal of Digital Imaging, Publisher SpringerNew York, Issue Volume 11, Number 4, 1998, pp 193-200.

[30] Wanga X, Wong BS, Guan TC, “Image enhancement for radiography inspection”, International Conference on

Experimental Mechanics, 2004, pp 462-468.[31] Ball JE, “Digital mammogram spiculated mass detection and spicule segmentation using level sets”,Proceedings of the 29th Annual International Conference of the IEEE EMBS, 2007, pp. 4979-4984.

[32] Antonis Daskalakis, et al, “An Efficient CLAHE-Based, Spot-Adaptive, Image Segmentation Technique forImproving Microarray Genes' Quantification”, 2nd International Conference on Experiments/Process/System

Modelling/Simulation & Optimization, Athens, 4-7 July, 2007.[33] J Suckling et al (1994) “The Mammographic Image Analysis Society Digital Mammogram Database Exerpta

Medica”. International Congress Series 1069 pp375-378.




42

Comparative Analysis of Performance of Series FACTS Devices Using PSO

Based Optimal Power Flow Solutions

K.Padma1

and K.Vaisakh2

Department of Electrical Engineering, AU College of Engineering

Andhra University, Visakhapatnam-530003,AP, IndiaE-mail: [email protected]

AbstractThis paper incorporates three FACTS devices such as PST, TCSC and SSSC in optimal power flow solutions to

enhance the performance of the power systems. The particle swarm optimization is used for solving the optimalpower flow problem for steady-state studies. The effectiveness of the proposed approach was tested on IEEE 14-buswith series FACTS devices. Results show that the proposed algorithm gives better solutions to enhance the systemperformance with SSSC compared to other devices.

Keywords: Power system operation, series FACTS devices, particle swarm optimization, optimal power flowsolution




43

1. Introduction

Complexity of operating modern power systems is continually increasing because of larger power transfer overlonger distances, greater interdependence among interconnected systems, more complicate coordination andinteraction among various system controllers and less power reserves. These demands have forced systems to beoperated closer to their security limits, because instability has become a major threat for system operation, as

evidenced by the recent state of blackouts [1]. Voltage Stability is becoming an increasing source of concern insecure operation of present day power systems. Hence it is necessary to consider the voltage stability aspects insolving the optimal reactive power control problems.

To meet the increasing power demand with existing transmission lines, the introduction of FACTS devicesbecomes an alternative. FACTS can improve the stability of network, and reduce the flows in heavily loaded lines bycontrolling their parameters including series impedance, current, and voltage and phase angle. Especially, FACTSdevices can enable a line to carry its flow close to rating capacity and consequently can improve the power systemsecurity in contingency [2-4].

In a power system, the FACTS devices may be used to achieve several goals. In steady-state, for a meshednetwork, they can permit to operate transmission lines close to their thermal limits and to reduce the loop flows. Inthis respect, they act by supplying or absorbing reactive power, increasing or reducing voltage and controlling seriesimpedance or phase angle [5-6]. Different types of devices have been developed such as series controllers, shuntcontrollers, and combined series-shunt controllers. Inside a category, several FACTS devices exist and each one has

its own properties and may be used in specific contexts. The choice of the appropriate device is important since itdepends on the goals to be reached.

Recently, the success achieved by evolutionary algorithms in the solution of complex problems and theimprovement made in computation such as parallel computation have stimulated the development of new algorithmslike Particle Swarm Optimization (PSO) present great convergence characteristics and capability of determiningglobal optima[7-13].

This paper examines the effect of series FACTS devices on the power system performance using PSO basedOPF solutions. The effectiveness of the proposed method was tested on IEEE 14-bus system and comparison wasmade on the performance of the three FACTS devices.

2. Voltage Stability Index (L-index) Computation

The voltage stability L-index is a good voltage stability indicator with its value change between zero (no load) and

one (voltage collapse) [14]. Moreover, it can be used as a quantitative measure to estimate the voltage stabilitymargin against the operating point. For a given system operating condition, using the load flow (state estimation)results, the voltage stability L -index is computed as [14],

j L = ∑

=

−g

i j

i

jiV

V F

1

1 ng j ,...,1+= (1)

All the terms within the sigma on the RHS of equation (1) are complex quantities. The values of ji F arebtained from the network Y -bus matrix.

For stability, the index j L must not be violated (maximum limit=1) for any of the nodes j. Hence, theglobal indicator j

L describing the stability of the complete subsystem is given by maximum of jL for all j (loadbuses). An j L -index value away from 1 and close to 0 indicates an improved system security. The advantage of this

j L -index lies in the simplicity of the numerical calculation and expressiveness of the results.

3. Series FACTS controllers

FACTS controllers are able to change, in a fast and effective way, the network parameters in order toachieve better system performance. FACTS controllers, such as phase shifter, shunt, or series compensation and themost recent developed converter-based power electronic controllers, make it possible to control circuit impedance,voltage angle, and power flow for optimal operation performance of power systems, facilitate the development of competitive electric energy markets, stimulate the unbundling the power generation from transmission and mandateopen access to transmission services, etc. The benefit brought about by FACTS includes improvement of systembehavior and enhancement of system reliability. However, their main function is to control power flows..




44

There are various types of series FACTS devices available for this purpose, namely Phase Shift Transformer(PST), Thyristor-Controlled Series Capacitor (TCSC), and Static Synchronous Series Compensator (SSSC). Each of these FACTS devices, however, has its own characteristics and limitations.

3.1 Phase Shift Transformer (PST)

The Phase shift Transformer circuit diagram can be represented by Figure 1. Due to the installation of phaseshifter, the system will have lots of benefits such as overload release, system loss reduction and generationadjustment reduction. All these benefits may be selected as objective functions for OPF with PST. However, theprimary purpose of installing phase shifter is to remove line over load.

Figure 1 Circuit diagram of phase shifter

3.2 Thyristor Controlled Series Compensator (TCSC)

One important FACTS controller is the TCSC which allows rapid and continuous changes of thetransmission line impedance. TCSC controls the active power transmitted by varying the effective line reactance byconnecting a variable reactance in series with line and is shown in Figure 2. TCSC is mainly used for improving theactive power flow across the transmission line.

Figure 2 Circuit diagram of TCSC

3.3 Static Synchronous Series Compensator (SSSC)

A SSSC usually consists of a coupling transformer, an inverter and a capacitor. The SSSC is series connected with atransmission line through the coupling transformer.

It is assumed here that the transmission line is series connected via the SSSC bus j. The active and reactivepower flows of the SSSC branch i-j entering the bus j are equal to the sending end active and reactive power flows of the transmission line, respectively. In principle, the SSSC can generate and insert a series voltage, which can beregulated to change the impedance (more precisely reactance) of the transmission line. In this way, the power flow of

the transmission line or the voltage of the bus, which the SSSC is connected with, can be controlled.






46

controllers can be formulated as follows:

Minimize F = ∑=

++ NG

i

iGiiGii C PbPa1

2 )(( (11)

The minimization problem is subjected following equality and inequality constraints4.1 Constraints

The OPF problem has two categories of constraints: Equality Constraints: These are the sets of nonlinear power flow equations that govern the power system, i.e,

0)cos(1

=+−−− ∑=

jiijij j

n

j

i DiGi Y V V PP δ δ θ (12)

0)sin(1

=+−+− ∑=

jiijij j

n

j

i DiGi Y V V QQ δ δ θ (13)

where GiP and GiQ are the real and reactive power outputs injected at bus i respectively, the load demand at the

same bus is represented by DiP and DiQ , and elements of the bus admittance matrix are represented by ijY and

ijθ .

Inequality Constraints: These are the set of constraints that represent the system operational and security limits likethe bounds on the following:1) generators real and reactive power outputs

N iPPP GiGiGi ,,1,maxminK=≤≤ (14)

N iQQQ GiGiGi ,,1,maxminK=≤≤ (15)

2) voltage magnitudes at each bus in the network

NLiV V V iii ,,1,maxminK=≤≤ (16)

3) transformer tap settings

NT iT T T iii ,,1,maxminK=≤≤ (17)

4) reactive power injections due to capacitor banks

CSiQQQ CiCiCi ,,1,maxmin

K

=≤≤ (18)5) transmission lines loading

nliSS ii ,,1,maxK=≤ (19)

6) voltage stability index

NLi Lj Lj ii ,,1,maxK=≤ (20)

FACTS devices constraints:i) PST constraints

maxminPiPiPi α α α ≤≤ Phase angle constraint of PST (21)

where Piα = Phase shift angle of PST at line i

maxmin

, PiPi α α = Lower and upper phase shift angle limits of PST at line i ii) TCSC constraints: Reactance constraint of TCSC

TCSC iTCSC iTCSC iTCSC ni X X X ,...,2,1maxmin=≤≤ (22)

where TCSCi X = Reactance of TCSC at line i minTCSCi X = Minimum reactance of TCSC at line i maxTCSCi X = Maximum reactance of TCSC at line i




47

TCSC n = number of TCSC’s

iii) SSSC constraints: Series voltage source magnitude and angle limitsmaxmin

sesese V V V ≤≤ ; (23)

maxminsesese θ θ θ ≤≤ (24)

The equality constraints are satisfied by running the power flow program. The generator bus terminal

voltages ( giV ), transformer tap settings ( k t ) and the reactive power generation of capacitor bank ( CiQ ) are the

control variables and they are self-restricted by the representation itself. The active power generation at the slack bus

( gsP ), load bus voltages ( LiV ) and reactive power generation ( giQ ), voltage stability ( j L )-index are state variables

which are restricted through penalty function approach.

The installation of shunt reactive power sources involves the investment cost. The location of FACTSdevices and its size also involves the investment cost. These issues are beyond the scope of this thesis, and are notconsidered in the solution of optimal power flow problems during minimization of different objective functions.

5. Overview of Particle Swarm Optimization

Basically, the PSO was developed through simulation of birds flocking in two-dimensional space [15]. Theposition of each bird (called agent) is represented by a point in the X–Y coordinates, and the velocity is similarlydefined. Bird flocking is assumed to optimize a certain objective function. Each agent knows its best value so far(pbest) and its current position. This information is an analogy of personal experience of an agent. Moreover, eachagent knows the best value so far in the group (gbest) among pbests of all agents. This information is an analogy of an agent knowing how other agents around it have performed. Each agent tries to modify its position using theconcept of velocity.

The velocity of each agent can be updated by the following equation:

)(*)(* 22111 k

i

k

ii

k

i

k

i sgbest rand cs pbest rand cwvv −+−+=+

(25)

where k iv is velocity of agent i at iteration k,

w is weighting function,c1 and c2 are weighting factors,

rand1 and rand2 are random numbers between 0 and 1,k is is current position of agent i at iteration k,

pbesti is the pbest of agent i, and gbest is the best value so far in the group among the pbests of all agents.The following weighting function is usually used in (25):

iter iter wwww *)) /()(( maxminmaxmax −−= (26)

where maxw is the final weight, minw is the initial weight, itermax is the maximum iteration number, and iter is the

current iteration number. Using the previous equations, a certain velocity, which gradually brings the agents close topbest and gbest, can be calculated. The current position (search point in the solution space) can be modified by thefollowing equation:

11 +++=

k

i

k

i

k

i vss (27)

6. Overall Computational Procedure for solving the problem

The implementation steps of the proposed PSO based algorithm can be written as follows;

Step 1: Input the system data for load flow analysisStep 2: Select a FACTS device and its location in the systemStep 3: At the generation Gen =0; set the simulation parameters of PSO parameters and randomly initialize k

individuals within respective limits and save them in the archive.




48

Step 4: For each individual in the archive, run power flow under the selected network contingency to determine loadbus voltages, angles, load bus voltage stability indices, generator reactive power outputs and calculate linepower flows.

Step 5: Evaluate the penalty functionsStep 6 : Evaluate the objective function values and the corresponding fitness values for each individual.Step 7 : Find the generation local best xlocal and global best xglobal and store them.Step 8: Increase the generation counter Gen = Gen+1.Step 9: Apply the PSO operators to generate new k individualsStep 10: For each new individual in the archive, run power flow to determine load bus voltages, angles, load bus

voltage stability indices, generator reactive power outputs and calculate line power flows.Step 11: Evaluate the penalty functionsStep 12: Evaluate the objective function values and the corresponding fitness values for each new individual.Step 13: Apply the selection operator of PSO and update the individuals.Step 14: Update the generation local best xlocal and global best xglobal and store them.Step 15: If one of stopping criterion have not been met, repeat steps 4-14. Else go to stop 16Step 16 : Print the results

7. Simulation results

The proposed PSO algorithm to solve optimal power flow problems incorporating series FACTS devices for

enhancement of system performance i tested on standard IEEE 14-bus test system. The proposed algorithms areimplemented using MATLAB 7.4 running on Pentium IV, 2.66GHz, and 512MB RAM personal computer. The PSOparameters used for the simulation are summarized in Table 1.

Table 1

Optimal parameter settings for PSOParameter PSO

Population sizeNumber of iterations

Cognitive constant, c1Social constant, c2Inertia weight, W

20150

22

0.3-0.95

The network and load data for this system is taken from [16]. To test the ability of the proposed PSOalgorithm for solving optimal power flow problem with three series FACTS device. One objective function isconsidered for the minimization using the proposed PSO algorithm. In order to show the affect of power flow controlcapability of the FACTS devices in proposed PSO OPF algorithm, four sub case studies are carried out on the IEEE14-bus system.

Case (a): power system normal operation (without FACTS devices installation),Case (b): one PST is installed. PST is installed at line connected between buses 12 and 13 with line real and

reactive power settings of PST, Pmk = 0.025125, Qmk = 0.00 and –п /4 ≤ pi ≤ п /4 for two contingency cases.Case (c): one TCSC is installed. TCSC is installed at line connected between buses 12 and 13 with XTCSC

value of -0.015 and 0 ≤ XTCSCi ≤ 60% of line reactance.Case (d): one SSSC installed. SSSC installed at line connected between buses 12 and 13 with line real and

reactive power settings of SSSC, Pmk= 0.025125 and Qmk = 0.0145.The first case is the normal operation of network without using any FACTS device. In second, third and

forth cases just installation of one device has been considered. Each device is placed in optimal location obtainedfrom the literature and trail and error method.The evolution of objective function during optimization by the proposed method is shown in Figure 4 under

selected series FACTS device. The optimal settings of control variables and FACTS device parameters duringminimization of objective function is given in Tables 2 under the selected series FACTS device respectively. Fromthe Table 2, it is noted that PSO algorithm is able to enhance the system performance while maintaining all controlvariables and reactive power outputs within their limits. Also from the Table 2 it is obvious that SSSC exhibits bestperformance compared to other devices.

. The line loadings, bus voltage profiles, bus voltage angles, and voltage stability indices with and withoutFACTS controllers are shown in Figures 5-8 under the each FACTS device. The Figures 5-8 revel that the proposed




49

PSO methodology incorporating FACTS devices is cable of maintaining better line loadings, load bus voltageprofiles, bus voltage angles and voltage stability indices.

8000

8200

8400

8600

8800

9000

9200

9400

1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145

Cost($/h)

N u m b e r o f i t e r a t i o n s

W/O FACTS

With PST

With TCSC

With SSSC

Figure 4 Convergence of cost of generation without and with FACTS device for IEEE 14-bus system

Table 2

Optimal settings of control variables for IEEE14-bus systemLimits(p.u) With FACTS

ControlVariables Min Max

WithoutFACTS

PST TCSC SSSCPG1 PG2 PG3 PG4 PG5

0.00.00.00.00.0

3.3241.4001.0001.0001.000

1.94470.36470.29190.00000.0830

1.96590.36510.28400.00000.0693

2.04260.35320.14670.00000.1430

1.97430.36890.28200.00000.0549

VG1 VG2

VG3 VG4 VG5

0.950.95

0.950.950.95

1.101.10

1.101.101.10

1.05571.0292

1.00460.99610.9974

1.06201.0403

1.01421.04711.0602

1.06501.0389

1.01261.01911.0018

1.09131.0658

1.04221.04181.0403

Tap - 1Tap - 2Tap - 3

0.90.90.9

1.11.11.1

1.01520.94881.0539

0.94691.05430.9442

1.00310.98251.0097

1.01690.96400.9792

QC6 QC8 QC14

0.00.00.0

0.100.100.10

0.06390.03570.0556

0.00920.02250.0412

0.07180.05910.0447

0.00140.10000.0579

Cost ($/h)Ploss (p.u.)LjmaxCPU time (s)

8087.2000.09420.0872

20.0470

8080.0000.09430.0774

20.2030

8061.6000.09550.081822.3910

8059.7000.09010.0749

24.3800




50

0

10

20

30

40

50

60

70

80

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Line Number

% M V A L o a d i n g

W/O FACTS

With PST

With TCSC

With SSSC

Figure 5 Line loadings of IEEE 14-bus system without and with FACTS devices

0.9

0.92

0.94

0.96

0.98

1

1.02

1.04

1.06

1.08

1.1

1.12

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Bus Number

V o l t a g e ( p . u

. )

W/O FACTS

With PST

With TCSC

With SSSC

Figure 6 Bus voltage profiles of IEEE 14-bus system without and with FACTS devices

-16

-14

-12

-10

-8

-6

-4

-2

0

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Bus Number

A n g l e ( d e g . )

W/O FACTS

With PST

With TCSC

With SSSC

Figure 7 Bus voltage angles of IEEE 14-bus system without and with FACTS devices




51

0

0.01

0.02

0.030.04

0.05

0.06

0.07

0.08

0.09

0.1

6 7 8 9 10 11 12 13 14

Load Bus Number

V o l t a g e s

t a b i l i t y L - i n d e x

W/O FACTS

With PST

With TCSC

With SSSC

Figure 8 Voltage stability indices of IEEE 14-bus system without and with FACTS devices

8. ConclusionsThis paper has presented an OPF model incorporating series FACTS controllers such as PST, TCSC and

SSSC using the PSO algorithm for enhancement of system performance. This model is able to solve power networksof any size and converges with minimum number of iterations and independent of initial conditions. The IEEE 14-bus systems have been used to demonstrate the proposed methods over a wide range of power flow variations in thetransmission system. The results from the two tested systems showed that the proposed integrated OPF with StaticSynchronous Series Compensator scheme is very effective compared to other devices in improving the security of the power system.

References

[1] M.Noroozian, L.Angquist, M.Ghandhari, G.Anderson, "Improving Power System Dynamics by Series-connected FACTS Devices", IEEE Trans. on Power Delivery, Vol.12, No.4, October 1997.

[2] M.Noroozian, L.Angquist, M.Ghandhari, G.Anderson, "Use of UPFC for Optimal Power Flow Control", IEEE Trans. on Power Delivery, Vol.12, No.4, October 1997.

[3] Roy Billinton, Mahmud Fotuhi-Firuzabad, Sherif Omar Faried, Saleh Aboreshaid, "Impact of UnifiedPower Flow Controllers om Power System Reliability", IEEE Trans. on Power Systems Vol.15, No.1,

February 2000.[4] James A. Momoh, Jizhong Z. Zhu, Garfiled D. Boswell, Stephen Hoffman, "Power System Security

Enhancement by OPF with Phase Shifter", IEEE Trans. on Power Systems, Vol.16, No.2, May 2001.[5] N. G. Hingorani, L. Gyugyi, Understanding FACTS: Concepts and Technology of Flexible AC

Transmission Systems, IEEE Press, New- York, 2000.[6] D. Povh and al, Load Flow Control in High Voltage Power Systems Using FACTS Controllers, CIGRÉ

Task Force 38.01.06, Jan. 1996.[7] Gnanadas R., Venkatesh P. & Narayana Prasad Padhy, “Evolutionary Programming Based Optimal Power

Flow For Units With Non-Smooth Fuel Cost Functions”, Electric Power Components and Systems, Vol.33,2005, pp. 1245-1250.

[8] Gnanadass R., Venkatesh P., Palanivelu T. G. & Manivannan K., “Evolutionary Programming Solution Of Economic Load Dispatch With Combined Cycle Co-Generation Effect”, Institute Of Engineers Journal-EL ,Vol. 85, September 2004, pp. 124-128.

[9] Hong-TzerYang, Pai-chuan Yang & Ching-Lein Huang, “Evolutionary programming based economicdispatch for units with non-smooth fuel cost functions”, IEEE Transactions on Power Systems, vol. 11, No.1, February 1996, pp. 112-118.

[10] Jayabarathi T., Jayaprakash K., Jeyakumar D. N. & Raghunathan T., “Evolutionary ProgrammingTechniques for Different Kinds of Economic Dispatch Problems”, Electric Power Systems Research, Vol.73, 2005, pp. 169-176.

[11] Sinha N., Chakravarthi R. & Chattopadhyay P. K., “Improved Fast Evolutionary Program for EconomicLoad Dispatch with Non-Smooth Cost Curves”, Institute Of Engineers Journal-EL, Vol. 84, September2004, pp. 110-114.




52

[12] Somasundaram P., Kuppuswamy K. & Kumidini Devi R.P., “Evolutionary Programming Based SecurityConstrained Power Flow”, Electric Power Systems Reasearch, Vol. 72, July 2004, pp. 137-145.

[13] Somasundaram P., Kuppuswamy K. & Kumudini Devi R.P., “Economic Dispatch With ProhibitedOperating Zones Using Fast Computation Evolutionary Programming Algorithm”, Electric Power SystemsResearch, vol. 70, 2004, pp. 245-252.

[14] P. Kessel, H. Glavitch, “Estimating the voltage stability of a power system,” IEEE Trans. Power Delivery,1986, PWRD-1(3), pp. 346-354.

[15] Abido MA. “Optimal power flow using particle swarm optimization,” Electric Power Energy Syst 2002;24(7): 563-71.

[16] IEEE 14-bus system (1996), (Online) Available at //www.ee.washington.edu

Biographies

K. Padma received the B.Tech degree in electrical and electronics engineering from SVUniversity, Tirupathi, India in 2005, M.E degree from Andhra University, Visakhapatnam,India in 2010.She is currently working as an Assistant Professor in the department of electricalengineering, AU College of engineering, Visakhapatnam, A.P, India. Her research interestincludes power system operation and control, power system analysis, power systemoptimization, soft computing applications and FACTS.

Dr.K.Vaisakh received the B.E degree in electrical engineering from Osmania University,Hyderabad, India in 1994, M.Tech degree from JNT University, Hyderabad, India in 1999,and Ph.D. degree in electrical engineering from the Indian Institute of Science, Bangalore,India in the year 2005.Currently, he is working as professor in the department of electrical engineering, AUCollege of engineering, Andhra University, Visakhapatnam, AP, India. His researchinterests include optimal operation of power system, voltage stability, FACTS, powerelectronic drives and power system dynamics.




53

Secure and Unique Biometric Template Using Post Quantum Cryptosystem

Ajay Sharma1

and Deo Brat Ojha2

1 Research Scholar Singhania University, Jhunjhunu, Rajesthan,India


2Professor, Department of Mathematics,Rajkumar Goel Institute of Technology, Ghaziabad, U.P., India


Abstract In this paper we enhance the accuracy and security of biometric template using fuzzy commitment scheme with postquantum cryptosystem as cryptographic function in it. Here it is possible to generate many different secure biometrictemplate for the same system and also unique biometric templates for multiple systems from the same biometric trait;it is just a matter of using a different set of error vector. It is also easy to cancel a secure template by simply deletingthe compromised template and generating a new one by using different error vector.

Keywords: Cryptography, Fuzzy Commitment Scheme, Biometric System, Template, algorithmic Noise, Enrollmentphase




54

1. IntroductionCryptography is considered to be one of the fundamental building blocks to protect the biometric data with thegrowing use of biometric recognition system. Biometric provides a person with a distinct characteristic that is alwaysprevalent. It is a technique of authentication of a person’s individuality from one or more behavioral or physiologicalfeature[3]. The use of biometrics (e.g., fingerprints, irises, faces) for recognizing individuals is becomingincreasingly popular and many applications are already available. Although these applications can be fundamentally

different, they can still be grouped into one of two categories: verification and identification [4][5][6].A well-known difficulty has been how to cope with the 10 to 20% of error bits within an biometric data and derivean error-free template. It is fundamentally impossible to avoid noise during biometric data acquisition, because “lifemeans change“. For example, faces age and iris patterns are not perfectly invariant to a contraction of a pupil. Morenoise is introduced by changes in the environmental conditions, which is again an unavoidable circumstance. Finallynoise often finds its way into the sensor, during transmission or in the data processing process (“algorithmic noise“).The latter noise sources can be reduced or even removed by improved engineering .To solve this problem, fuzzycommitment scheme play an important role. Fuzzy commitment scheme is a tool for handling the noise in templateof a biometric recognition system.. Juels and Wattenberg’s fuzzy commitment scheme [2] has been introduced tohandle the difference occurring between two captured of biometric data, using error correcting code.The various approach here been proposed to protect the stored template ,some are hardware based which is usedstand alone biometric system-on-devices. Some are software based which is relay on feature transformation andbiometric cryptosystems. Here on biometric cryptosystem common encryption technique, such as AES(AdvanceEncryption standard) or RSA can not be used because of interclass variation in the biometric template[4,5].This paper, itself define an application of a fuzzy commitment scheme with McEliece’s cipher [8,9]. The main ideais here the biometric matching problem is transformed into an error correcting issue. We carefully studied the errorpatterns within biometric data, and devised a two-layer error correction technique that combines Hamming code andGoppa code. The error-correcting methods remove noise in the template [7]. Along with accuracy, Someenhancement in the privacy of biometric cryptosystem, common encryption technique, such as AES or RSA can’t beused, so the auxiliary data can be masked using homomorphic encryption that allows certain arithmetic operation inthe encryption domain[24].

2. Preliminaries

2.1 Biometric SystemA generic biometric system consists of five components: Sensor, feature extractor, template database, matcher, anddecision module

In general , a biometric based recognition system consists of two phase In the enrollment phase, the biometrictemplate b are processed from a user U and stored or registered in the database. The second phase is theverification phase; In verification system captures a new biometric sample b′ from U and compare it to theregistered or reference data via a matching function. Let µ be the biometric measure of U and τ is a recognition

threshold, b′ will be accepted if ( , )b b µ τ ′ ≤ , else rejected. Mainly two kinds of errors are associated to thisscheme: False Reject (FR), when a matching user, i.e. a legitimate user, is rejected; False Acceptance (FA), when anon-matching one, e.g. an impostor, is accepted. Note that, when the threshold increases, the FR’s rate (FRR)decreases while the FA’s rate (FAR) grows, and conversely [11].

2.2 Definition

A metric space is a set C with a distance function dist : ),0[ ∞=→×+ RC C , which obeys the usual

properties(symmetric, triangle inequalities, zero distance between equal points)[12].

2.3 Definition

Let nC }1,0{ be a code set which consists of a set of code words ic of length n. The distance metric between any

two code words ic and jc in C is defined by ∑=

∈−=n

r

ji jr ir ji C ccccccdist 1

, ),(

This is known as Hamming distance [13].




55

2.4 Definition

An error correction function f for a code C is defined as

}}{overminimum,theis),( / {)( i ji ji cC ccdist cc f −= . Here, ( )i j c f c = is called the nearest neighbor

of ic [14].

2.5 DefinitionThe measurement of nearness between two code words c and c′ is defined by nccdist cc / ),(),(nearness ′=′ ,

it is obvious that 1)c(c, nearness0 ≤′≤ [13].

2.6 Definition

The fuzzy membership function for a codeword c′ to be equal to a given c is defined as[13]

otherwise z

1zz)c,nearness(cif 0)( 0

=

<≤=′=′cFUZZ

2.7 Fuzzy Commitment Scheme with McEliece scheme[9]Protocols are essentially a set of rules associated with a process or a scheme defining the process. Commitmentprotocols were first introduced by Blum [1] . Moreover in the conventional commitment schemes, opening key arerequired to enable the sender to prove the commitment. However there could be many instances where thetransmission involves noise or minor errors arising purely because of the factors over which neither sender nor thereceiver have any control , which creates uncertainties. Fuzzy commitment scheme was first introduced by Juels andMartin [2]. The new property “fuzziness” in the open phase to allow, acceptance of the commitment using corruptedopening key that is close to the original one in appropriate metric or distance. Fuzzy commitment scheme is based onhash function [2] which causes them to share two shortcomings:1. The hash functions used should be strongly collision free. However, this property can only be empirically checked.It actually turns out that some schemes are inadvertently based on weakly collision-free hash functions.2. Hash functions alone cannot offer non-repudiability.Here we use the speed of McEliece and its randomness to enhance the fuzzy commitment scheme by using code basecryptosystem which is base on Goppa Code.The scheme consists of three phase: first setup phase, second commitment phase and third opening/verifying phase.

Setup up phase: At time 0t , it is agreed between all that

XORCK ≅ f ≅ nearest neighbour in { ( )}h m .

.20.00 = Z

A Id = Identifier

It is assumed that McEliece public key( AP ) is duly certified and public. It can be described by its k n× generator

matrix G. With the aid of a regular k k × matrix S and an n n× permutation matrix P, a new generator matrix G’ isconstructed that hides the structure of G:G’ = S . G . P

The public key consists of G’and the matrices S and P together with g(x) are the private key( AS ).

Here, the root cause for using A Id as we stated in introduction section that This cryptosystem can not be used for

authentication because the encryption is not one to one and total algorithm is truly asymmetric.

Commitment phase: At time 1t

1. Alice chooses message m in the form of bitstring to which she wish to commit.2. Alice generates a secret pseudo q-bit random vector r.

3. Alice has a identifier A Id of p-bit random vector.

4. Alice concatenate her identifier A Id with secret pseudo q-bit random vector r which give us a vector R= A

Id r .




56

Here ( ) A

h m mP= where ( ) (2 )nh m GF ⊆ ,

Encryption: AC mP e= ⊕ , where ( )e g R= , here g is an invertible function which maps R in to an n-bit error

vector of weight α .

According to the algorithms )lg( 1ecommita into string c i.e. her commitment

lg( , ( ), )c commita XOR h m C = , then after Alice sends c to Bob, which Bob will receive as ( ) f

t c , where f

t is

the transmission function which includes noise .

Open Phase: Alice sends the procedure for revealing the hidden commitment at time 2t and Bob use this,So Alice

discloses the procedure ( )h m and C to Bob to open the commitment.

)lg( 2eopena : Bob constructs c′ using lgcommita , message )(mt and opening key

i.e lg( , ( ( )), ( )) f f c commita XOR t h m t C ′ = and checks whether the result is same as the received commitment

( ) f t c .

Fuzzy decision making

0( ( ( ), ( )) )

f If nearness t c f c Z ′ ≤

Then A is bound to act as in m Else he is free not to act as m .

Then after acceptance ,Bob decrypt the massage as first m can be recovered by using the decryption algorithm in theoriginal scheme. In the meantime, the value ( )g R can also be obtained. Then the receiver computes

1( ( )) R g g R−= , where 1g − is the inverse of g . finally Bob calculates 1( )( ) f c SGP −′ and finally get the

message. Here Bob get the A Id from the R to know the authenticity of the sender.

3. Related workOur work is inspired from a number of authors who combine well known technique from the area of error correctingcode and cryptography to achieve a improve type of cryptographic primitive [1,2,8,13,14]. Further numerous worksthat suggest combination of biometrics and cryptography. A more detailed of related research work on this field canbe found in [15, 16, 17 ].

4. Proposed System ArchitectureIn general, the identity theft problem is drastically exacerbated for the biometric systems. The proposed architectureof biometric system will have enhanced the security and accuracy with respect to traditional system by combineusage of code base cryptosystem and error correcting code.In the enrollment stage (figure1 (a)) of a typical biometric recognition system, after the biometric acquisitionmodule, some processing is applied in order to obtain the biometric template, b which is then stored in a database.

Here H is called the hamming space of length N e.g. { } 20,1 , N N

F Η = = where { }2 0,1F = . Here g′ is an invertible

function which maps R in to an n-bit error vector of weight α . However, the biometric data is never stored in thedatabase to prevent it from being stolen. Instead, after the biometric has been acquired and the biometric templatehas been generated, a cryptographic function will be applied to it [9]. The result of this operation will then be storedin the database; this will be referred to in the rest of the paper as the secure biometric template. It should be pointed

out that it is impossible to recover any biometric data from this secure template as the cryptographic function is notinvertible.




57

(a) Enrollment Phase (figure 1)

During the verification stage (figure1(b)), the probe biometric is acquired and the corresponding template, b′ , is

generated.. The problem here is that b itself is not stored in the database, but only a encrypted version of it. Torecovered the original biometric template b from the database, if the user is who he claims, or something

completely different if he is not. Therefore, the output of the feature extractor b′ needs to be encrypted. Only then,is the result compared to the encrypted that is stored in the database. If the ( )P E z′ and ( )P E z are equal, then the user

is validated to be who he claims to be. With this system, the three requirements above are verified. In particular, it is

possible to generate many different secure biometric templates from the same biometric trait; it is just a matter of using a different set of error vector( e ) . It is also easy to cancel a secure template by simply deleting thecompromised template and generating a new one by using different error vector ( e ). Finally, since the biometricdata is never stored in a database, this guarantees that this information remains private.

UserBiometric (U)

Acquisition Pre-processing FeatureExtraction

TemplateGenerationb ∈ Η

PublicKey

( P )

( )g b b P= ×

( )P

E z b P e= × ⊕ Database

ErrorVector

( )

,

e g R

where R Id r

′=

=

Biometric

Identifier ( Id )

Random No.( r )




58

(b) Verification Phase (figure 1)

4.1 AcquisitionThe acquisition module, absolutely necessary in a real biometric verification system, has not been implemented buthere Instead of implementation, it is replaced by a large database of iris images , like the one developed by theChinese Academy of Sciences’ Institute of Automation (CASIA) [20] and code from [18]. This database consists of 22051 iris images from more than 700 subjects. All iris images are 8 bit gray-level JPEG files, collected under near

infrared illumination.

4.2 Pre-processingIn this step after acquisition is to extract the iris from the input eye images. The iris area is considered as a circularcrown limited by two circles. The iris inner (pupillary) and outer (scleric) circles are detected by applying thecircular Hough transform [21], relying on edge detection information previously computed using a modified Cannyedge detection algorithm [22].The eyelids often occlude part of the iris, thus being removed using a linear Houghtransform [23] .The presence of eyelashes is identified using a simple thresholding technique.

User

Biometric(U)

Acquisition Pre-processing FeatureExtraction

TemplateGeneration

b H ′ ∈

Database

( )g b b P′ ′= × ( )P E z′ FuzzyDecision

Acce t

Public Key(P)

R e j e c t

1( ( )( )P f E z SGP −′

( )P E z

Biometric

Identifier ( Id )

Random No.( r )

Error

Vector ( )

,

e g R

where R Id r

′=

=

V

e r i f i e d




59

4.3 Feature Extraction Once the iris texture is available, features are extracted from it to generate a more compact representation, also calledthe biometric template. Reader can also read [19] in more detail, to know how Iris Recognition Works. To extractthis representation, the two-dimensional normalized iris pattern is convolved with a Log-Gabor wavelet [3]. Theresulting phase information is quantized, using two bits per pixel. The resulting iris template is composed of 9600bits, stored as a 20×480 binary matrix.

4.4 Privacy-Protection and Error-CorrectionThis is main module of this scheme, in this scheme, we are using McEliece cryptosystem, which add some randomerror at the time of encryption that makes the original template more secure than poorly chosen passwords and othercryptosystem due to it’s randomness.At the time of enrollment phase, inputs are biometric template ( b ), error vector( e ) which is chosen randomly and

public key( P ) which has generating matrix that defines an error correcting code. Here g′ is an invertible function

which maps R in to an n-bit error vector of weight α . The output of this phase is encrypted template which one isstored on system or on a data card (i.e. smart card). Now, it is not easy to gain the template from this data without theknowledge of key and error vector.At the time of verification phase, a similar procedure is used with a new acquire template b′ with same error vector,and key and error correction coding is used to correct biometric templates. In this stage, the probe template of alegitimate user is (error) corrected in order to recover the original template, obtained during enrollment; this shouldbe possible because both templates are fairly similar. However, for an illegitimate user, whose probe template isfairly different from the one originally enrolled by the legitimate user, it should not be possible to recover the

original from the probe template. Now we calculates 1( )( ) f c SGP −′ and finally get the template. Here we get

Id from the R to know the identification of the machine. Which is unique for each machine so same biometricinformation should not be able to link template corresponding to the same individual for different machine.Therefore, the selected error correcting code should be strong enough to correct templates of legitimate users, but notso strong as to also correct the templates of illegitimate users. Therefore, µ be the biometric measure of U and τ

is a recognition threshold, b′ will be accepted if ( , )b b µ τ ′ ≤ , else rejected.

5. Security AnalysisThe accuracy of any biometric system depends on the ability of that system to separate genuine users from imposters.

Here we describe a possible attack to the scheme and identify ways of preventing it. It is possible for an attacker toimitate a signer by obtaining a copy of their biometric data. For example, see [10] for methods of duplicatingfingerprints. After obtaining a copy of the signer’s biometric data, the attacker can sign a forged message that willappear genuine on verification by the signer. To prevent this attack, genuine messages can be signed in the presenceof a trusted witness.Some issue of security in stored template consider here as,(1). Stored Template should not reveal any data and no close replica made from the stored data.(2). Multiple system using the same biometric information should not be able to link template corresponding to thesame individual.(3). If the stored data is compromised, remove that one and reissue a new one.Solutions of this issue are as,Explanation. of issue 1.Here we use goppa code in McEliece, first we encrypt a user biometric template and at the time of encryption an

error vector of fixed weight α is added. To reveal any template; attacker should now the solution of decodingproblem for unknown weight α of error vector which is very hard to solve. Coding theory based cryptosystem aresecure because decoding is hard without the knowledge of secret.Explanation. of issue 2.

If we consider error vector as ( )e g R′= here g is an invertible function which maps R into an n -bit error vector

of weightα . Where R Id r = and Id is machine identification and r is secret pseudo random vector. Since,

each system has unique Id so same biometric information should not be able to link template corresponding to thesame individual.




60

Explanation. of issue 3.It is possible to generate many different secure biometric templates from the same biometric trait; it is just a matterof using a different set of error vector( e ) . It is also easy to cancel a secure template by simply deleting thecompromised template and generating a new one by using different error vector ( e ).The randomness property of error vector is also required to prevent cross-matching of subjects across databases.In this scheme, we are using McEliece cryptosystem, which add some random error at the time of encryption thatmakes the original template more secure than poorly chosen passwords and other cryptosystem due to it’srandomness. In addition of randomness, McEliece cryptosystem is also probabilistic which give more susceptibilityof template towards brute force attacks.

It also provides non-repudiation i.e. a legitimate user may access the facilities offered by an application andthen do not claim that an intruder had circumvented the system. A bank clerk, for example, may modify the financialrecords of a customer and then can’t deny responsibility by claiming that an intruder could have possibly stolen herbiometric data. So our proposed scheme enhances the biometric security and accuracy from the previous availableliterature.

6. ConclusionUsing a public key cryptosystem to construct a commitment is a way to achieving non-repudiability andauthentication, a property which can not be offered by hash functions alone. By using McEliece in fuzzycommitment scheme error vector e used to enhance the security of the function hiding, particularly against matrixfactorization attacks. Main enhancement in this approach is randomness of the error vector, we can not obtain anyinformation about the positions in which the error occurs. Thus the information rate is increase and informationleakage rate decrease.We specifically focus on attacks designed to elicit information about the original biometric data of an individualfrom the stored template. As soon as identical templates are stored in multiple databases or datasets, it is possible toperform cross matching between them. We discuss the importance of adopting the error vector as a function of identification of machine and random number to enhance the uniqueness of biometric templates for differentmachine for same the individual. Since, each system has unique Id so same biometric information should not beable to link template corresponding to the same individual for different machine.The randomness property of error vector is also required to prevent cross-matching of subjects across databases.

References

[1] M. Blum, “Coin flipping by telephone: a protocol for solving impossible problems”, Proc. IEEE Computer Conference, pp. 133-137, 1982.

[2]. A.Juels and M.Wattenberg, “ A fuzzy commitment scheme”, In Proceedings of the 6 th

ACM Conference on

Computer and Communication Security, pp.28-36, November 1999.[3]Sunil V.K. Gaddam, Manohar Lal “Efficient Cancellable Biometric Key Generation Scheme for

Cryptography” International Journal of Network Security, Vol.11, No.2, pp.61–69, Sept. 2010.[4] A. K. Jain, S. Pankanti, S. Prabhakar, L. Hong, A. Ross, “Biometrics: A Grand Challenge”, Proc. of the

International Conference on Pattern Recognition, Vol. 2, pp. 935–942, August 2004.[5] J. Wayman, A. Jain, D. Maltoni, D. Maio, Biometric Systems:Technology, Design and Performance Evaluation,

Springer-Verlag, 2005.[6] D. Maltoni, D. Maio, A. K. Jain, S. Prabhakar, Handbook of Fingerprint Recognition, Springer, 2003.[7] Daugman,J “How Iris Recognition Works”, IEEE Transactions On Circuits andsystems for Video Technology, 2004,14 (1), pp.23-30

[8] Deo Brat Ojha, Ajay Sharma “ A fuzzy commitment scheme with McEliece’s cipher”Survey in Mathematics and

Its Application Vol.5(2010) pp73-83.[9] Ajay Sharma, Deo Brat Ojha,“Application of Coding Theory in Fuzzy Commitment Scheme”, Middle-East

Journal of Scientific Research 5 (6): 445-448, 2010.[10] T. van der Putte and J. Keuning. Biometrical Fingerprint Recognition: Don’t Get Your Fingers Burned.

Proceedings of the Fourth Working Conference on Smart Card Research and Advanced Applications, 2000.[11] Andrew Burnett, Adam Duffy, Tom Dowling “A Biometric Identity Based Signature Scheme”,

eprint.iacr.org /2004/176.pdf -[12] V.Pless, “ Introduction to theory of Error Correcting Codes”, Wiley , New York 1982.




61

[13].A.A.Al-saggaf,H.S.Acharya,“A Fuzzy Commitment Scheme” IEEE International Conference on Advances in

Computer Vision and Information Technology 28-30November 2007 – India.[14] F. J. MacWilliams and N. J. A. Sloane, Theory of Error-Correcting Codes. North Holland, 1991.[15] F. Hao, R. Anderson, and J. Daugman, “Combining crypto with biometrics effectively,” IEEE Transactions on

Computers, vol. 55, no. 9, pp. 1081–1088, 2006.[16] A. Cavoukian and A. Stoianov, “Biometric encryption: A positive-sum technology that achieves strong

authentication, security and privacy,” Information and privacy commissioner of Ontario, White Paper, March2007.

[17] E. Krichen, B. Dorizzi, Z. Sun, S. Garcia-Salicetti, and T. Tan, Guide to Biometric Reference Systems and

Performance Evaluation. Springer-Verlag, 2008, ch. Iris Recognition, pp. 25–50.[18] L. Masek, P. Kovesi, MATLAB Source Code for a Biometric Identification System Based on Iris Patterns,

School of Computer Science and Software Engineering, University of Western Australia, Australia, 2003.[19] J. G. Daugman, “How Iris Recognition Works”, IEEE Transactions on Circuits and Systems for Video

Technology, Vol. 14, No. 1, pp. 21–30, January 2004.[20] CASIA website, http://www.cbsr.ia.ac.cn/IrisDatabase.htm[21] T. Kawaguchi, D. Hidaka, M. Rizon, “Detection of eyes from human faces by Hough transform and separability

filter”, Proc.of the IEEE International Conference on Image Processing, Vol. 1, pp. 49-52, Vancouver, Canada,2000.

[22] J. Canny, “A Computational Approach to Edge Detection”, IEEE Transactions on Pattern Analysis and

Machine Intelligence, Vol. 8, pp. 679-714, 1986.

[23] R. Duda, P. Hart, “Use of Hough Transformation to Detect Lines and Curves in Pictures: Graphics and ImageProcessing”, Communications of the ACM , Vol. 15, pp. 11-15, 1972.

[24] J. Bringer and H. Chabanne,“An Authentication protocol with encrypted biometric data”, proc. Int. con

cryptology. Africacrypt .pp-109-124, 2008.




62

Statcom In Eight Bus System For Power Quality Enhancement

1. G.Sundar 2. S.RamaReddyResearch Scholar, Bharath University, ChennaiProfessor, Jerusalam College of Engg., Chennai

[email protected] , [email protected]

AbstractThis paper presents the detailed analysis of a STATCOM in eight bus system for harmonic content achievement

in order to enhance the voltage regulation by connecting a heavy load. For fast response, the STATCOM in intendedto replace the widely used static Var compensator (SVC). STATCOM display a low harmonic rate and injectsreactive power in the load. The results of simulation are presented.

Key words: STATCOM, voltage regulation, reactive power compensation




63

1. Introduction

The rapid growth in electrical energy use, combined with demand for low cost energy, has gradually led tothe development of generation sites remotely located from the load center. The generation of bulk power atremote locations necessitates the use of transmission line to connect generation sites to load centers. With longdistance ac power transmission and load growth, active control of reactive power is indispensable to stabilize thepower system and to maintain the supply voltage. The static synchronous compensator (STATCOM) using

voltage source inverters has been accepted as a competitive alternative to the conventional Static Varcompensator (SVC) using thyristor-controlled reactors STATCOM functions as a synchronous voltage source. Itcan provide reactive power compensation without the dependence on the ac system voltage. By controlling thereactive power, a STATCOM can stabilize the power system, increase the maximum active power flow andregulate the line voltages. Faster response makes STATCOM suitable for continuous power flow control andpower system stability improvement. The interaction between the AC system voltage and the inverter-composedvoltage provides the control of the STATCOM output. When these two voltages are synchronized and have thesame amplitude, the active and reactive power outputs are zero.

As with all static FACTS devices the STATCOM has the potential to be exceptionally reliable but with theadded capability to: sustain reactive current at low voltage (constant current not constant impedance), reduceland use and increase re-floatability (footprint 40% of SVC) and, be developed as a voltage and frequencysupport (by replacing capacitors with batteries as energy storage). Although currently being applied to regulatetransmission voltage to allow greater power flow in a voltage limited transmission network in the same manner

as a static var compensator (SVC), the STATCOM has further potential. By giving an inherently faster responseand greater output to a system with a depressed voltage, the STATCOM offers improved quality of supply. Themajor applications are: voltage stability enhancement, damping torsional oscillations, power system voltagecontrol, and power system stability improvement. These applications can be implemented with a suitable control(voltage magnitude and phase angle control).

This paper is aimed at the development of multilevel STATCOM for power quality enhancement. Theresults reveal that the three-level STATCOM offers higher efficiency and reduced voltage and current harmoniclevels. The concept of the proposed multilevel STATCOM is supported by MATLAB/SIMULINK results.

2. Basic Principle of STATCOM

The Static Synchronous Compensator (STATCOM) is shunt connected reactive compensation equipment,which is capable of generating and/or absorbing reactive power whose output can be varied so as to maintaincontrol of specific parameters of the electric power system. A single line diagram of STATCOM is shown in

Fig.1.The STATCOM basically consists of a step-down transformer with a leakage-reactance, a three-phaseGTO/IGBT voltage source inverter (VSI), and a DC capacitor. The AC voltage difference across the leakagereactance produces reactive power exchange between the STATCOM and the power system, such that the ACvoltage at the bus bar can be regulated to improve the voltage profile of the power system, which is the primaryduty of the STATCOM.

The principle of STATCOM operation is as follows. The VSI generates a controllable AC voltage sourcebehind the leakage reactance. This voltage is compared with the AC bus voltage system; when the AC busvoltage magnitude is above that of the VSI voltage magnitude, the AC system sees the STATCOM as aninductance connected to its terminals. Otherwise, if the VSI voltage magnitude is above that of the AC busvoltage magnitude, the AC system sees the STATCOM as a capacitance connected to its terminals. If thevoltage magnitudes are equal, the reactive power exchange is zero. If the STATCOM has a DC source or energystorage device on its DC side, it can supply real power to the power system. This can be achieved adjusting thephase angle of the STATCOM terminals and the phase angle of the AC power system. When the phase angle of

the AC power system leads the VSI phase angle, the STATCOM absorbs real power from the AC system; if thephase angle of the AC power system lags the VSI phase angle, the STATCOM supplies real power to ACsystem.




64

Fig.1. Single line diagram

3. Multilevel STATCOM

The Multilevel STATCOM configuration consists of a voltage source inverter, dc side capacitors (C ) withvoltage Vdc on it, and a coupling reactor or a transformer. The ac voltage difference across the coupling reactorproduces reactive power exchange between the STATCOM and the power systems at the point of commoncoupling (PCC). If the output voltage of the STATCOM (Vc) is more than the system voltage (Vl), in that case,reactive power is supplied to the power system, while reactive power goes to STATCOM if vc is less than thatof Vl. To take effect of this bidirectional flow of reactive power, the STATCOM output voltage should be variedaccording to requirement of reactive power compensation, and this can be accomplished in two ways: i) bychanging the switching angles while maintaining the dc capacitor voltage at a constant level (inverter type Icontrol) or ii) keeping switching angles fixed and varying the dc capacitors voltages (inverter type II control).The variation of dc capacitors voltages is simply achieved by varying the active power transfer betweenSTATCOM and power system by adjusting phase angle between Vc and Vl Each of these control schemes hastheir own merits and demerits. In general, inverter type II control is preferred where very fast voltage control isnot required such as in power system applications because THD injected can be minimized in this case. In thiswork inverter type II control scheme has been applied.

4. Harmonic in voltage source inverter (VSI)

The DC capacitor and series inductor can cause resonance at low order harmonics that can be present in thesystem contingency conditions or due to harmonic source in the vicinity (nonlinear loads). An unbalance due tofaults results in negative sequence (fundamental frequency) voltage which in turn results in second harmoniccomponents on the DC side. A second harmonic voltage on the DC capacitor results in both negative sequence(fundamental) component and a positive sequence third harmonic component. Depending on the value of theDC capacitor, it is possible that the negative sequence (fundamental) current and positive sequence thirdharmonic current in the AC are magnified. The use of multilevel inverter eliminates the need for harmonicfilters in the case of a STATCOM. The voltage and power ratings are expected to increase. The use of STATCOM in distribution systems has become attractive, not only for voltage regulation, but also foreliminating harmonics and improving power quality.

5. Simulation Results

A complete model of eight bus system using the simulation of STATCOM is presented in this paper. Theeight bus system model without STATCOM is shown in Fig.5a. Each line is represented by series impedancemodel. An additional load is added in parallel with load 1 by closing the breaker. At 0.2sec, additional load isconnected. The P & Q across the load 1 is as shown in Fig.5b.




65

Fig. 5a Model of 8-bus system without STATCOM

Fig. 5a Model of 8-bus system without STATCOM

Fig.5b Voltage, Real &Reactive Power across load-1




66

Fig.5c Model of 8-bus system with STATCOM

Fig.5d Model of STATCOM




67

Fig.5e Voltage, Real &Reactive Power across load-1

At t=0.2sec, the STATCOM is switched and connected to eight bus system between buses 3& 6 byswitching on the circuit breaker. The 8 bus system with STATCOM is shown in Fig.5c. The STATCOM modelis shown in Fig.5d. At this time, say t=0.2sec, the load is added to the buses, therefore, more reactive powercompensation is still required. The voltage phase displacement of STATCOM increases and therefore, the DCcapacitor voltage increases. The STATCOM injects Q in to the load 1. The regulated voltage, P & Q across theload 1 is shown in Fig.5e.

6. Conclusion

A STATCOM is simulated in eight bus system with MATLAB SIMULINK. The simulation results of eightbus system with & without STATCOM presented. This work has proposed to harmonic content achievement forvoltage regulation in eight bus system by connecting additional load. The simulation is based on the assumptionof load. The simulation results are presented and they are in line with the predictions.

7. References

[1] Ashwin and Thyagarajan (2006) “Modeling and simulation of VSC based STATCOM” , iicpe06 , pp303-307..

[2] Jianye Cuen, Shan Song, Zanji wang, (2006) “Analysis and implement of Thyrister based STATCOM”,

International conference on Power System technology.[3] J.Kumar, B.Das and P.Agarwal, (2008) “Selective harmonic elimination technique for a Multilivel

inverter”, Fifteenth national power system conference (NPSC),IIT Bombay, pp.608-13.[4] J.Kumar, B.Das and P.Agarwal, (2010) “Optimized Switching scheme of a Cascade Multilevel Inverter”,

Electric Power Components and systems, Vol.38, No.4, pp.445-64.[5] KEPRI Electric Power System Technology Group (2003) “Development of FACTS Operation Technology

(phase II: Pilot plant Development and construction), KEPRI final report. [6] MATLAB Version 7.3 [7] N.G. Hingorani and L. Gyugyi, (2000) Understanding FACTS, concepts and technology of Flexible AC

Transmission systems, piscatway, NJ:IEEE press. [8] R. Mienski, R. Pawelek and I. Wasiak, (2004) “Shunt Compensation for Power Quality improvement

using a STATCOM controller: Modeling and simulation”, IEEE Proce,Vol.51, No. 2.




68

G.Sundar has obtained his B.E degree from Madras University, Chennai in the year 2001. Heobtained his ME degree from SCSVMV University, Kanchipuram in the year 2005.He ispresently a research scholar at Bharath University, Chennai. He is working in the area of Power quality improvement using STATCOM.

S. Ramareddy is Professor of Electrical Department, Jerusalem College of engineering,Chennai. He obtained his D.E.E from S.M.V.M Polytechnic, Tanuku, A.P. A.M.I.E inElectrical Engg from institution of Engineers (India), M.E in Power System AnnaUniversity. He received Ph.D degree in the area of Resonant Converters from College of Engineering, Anna University, Chennai. He has published over 20 Technical papers inNational and International Conference proceeding/Journals. He has secured A.M.I.EInstitution Gold medal for obtaining higher marks. He has secured AIMO best project

award. He has worked in Tata Consulting Engineers, Bangalore and Anna University, Chennai. His researchinterest is in the area of resonant converter and Solid State drives. He is a life member of Institution of Engineers (India), Indian Society for India and Society of Power Engineers. He is a fellow of Institution of Electronics and telecommunication Engineers (India). He has published books on Power Electronics and SolidState circuits.




69

Improved Modification of single stage Ac-Ac converter for Induction

Heating Application

S.Arumugam1, S.Ramareddy

2

Research Scholar,Bharath University, Chennai, India

. [email protected] Professor

Jerusalem College of Engg Chennai, India. [email protected]

AbstractThis paper presents simulation of single stage Induction heating system with series Load Resonant.

Low frequency AC is converted in to High Frequency Ac using newly developed ZVS-PWM high frequencyinverter. This High Frequency is used for Induction Heating .single stage Ac-Ac converter system are modeledand they are simulated using matlab simulink. The simulation results of ZVS-PWM high frequency system arepresented. The effectiveness of this UFAC-to-HFAC direct power frequency converter using IGBTs forconsumer high-frequency IH appliances is evaluated and proved on the basis of simulation results.

Keywords: High frequency series load resonant inverter, Loss less capacitive snubbers, AsymmetricalPWM,ZVS,UFAC-HFAC direct Inverter, Consumer IH cooking appliances.



International Journal of Computational Intelligence and Information Security, Vol. 1 No. 6, August 2010

70

I. Introduction

In recent years, new application fields of high-frequency induction heating (IH) power technology inconsumer and industry have developed more and more in all electricity power utilization systems as energysaving. For example, these IH appliances are IH cooking heater, IH rice cooker, IH hot water producer, IHsteamer, and IH super heated steamer for cleaning, disinfecting, drying and cooking. These new IH applicationsin addition to microwave oven for food processor have been expanding dramatically with tremendous

development of the core technology of the state-of-the art high-frequency power electronics in IH technology.However, application-specific high frequency resonant inverters used for these appliances cause stage is requiredactually. The power losses of this power stage are to be significant as well as size and cost consideration. Inorder to solve these practical problems, the technological developments of new high frequency resonant invertercircuit and system topologies are necessary to use high frequency soft switching commutation as ZVS, ZCS andZVZCS. In these two power stage series load resonant high frequency inverter with a new passive PFC rectifier,switching losses and conduction losses of the power semiconductor devices used in the circuit can be reducedand achieve high efficiency for the high frequency switching operation. These can have the advantages of highperformance and the miniaturization, and use enough limit of rated characteristics of power semiconductordevices by reducing switching surges switching losses and conduction losses of power devices, getting largercooling devices and heat release systems, decreased rated ability of power devices by switching surges.Furthermore, increased EMI/RFI noise levels due to high frequency leakage current in high frequency switchingof conventional high frequency inverters operated under hard switching PWM. Therefore, the PFC rectifier.Generally, the high frequency IH power appliances based on the high frequency power electronics have PFC

rectifier stage, diode bridge rectifier stage as passive PFC rectifier with DCM of the inductor current and highfrequency resonant PWM inverter stage for supplying HFAC power to various HF-IH load structures. These twopower stage IH products using series load resonant HF inverter have high power factor and low utility ACcurrent harmonics characteristics in UFAC side. However, the IH direct inverter products actually need a lot of power semiconductor switching devices, passive resonant circuit components and bulky aluminum electrolyticsmoothing Capacitor stack From these present backgrounds, this paper deals with the one stage soft switchingPWM high frequency series load resonant direct inverter for IH applications. This high frequency resonantinverter topology has unique points as only one diode conducting mode passive PFC bridge rectifier operating atZCS.

Fig1.Block Diagram

In this paper, the characteristics of the high frequency direct inverter on the basis of computer simulation resultsare evaluated. In addition, its direct power regulation characteristics and UFAC side power quality

characteristics in periodic steady-state are illustrated including high frequency AC power regulation under theconditions of soft switching and hard switching.

This high-frequency inverter is composed of a passive PFC converter operating at one diode conductingbridge circuit and asymmetrical ZVS-PWM high frequency resonant inverter without the bulky electrolyticcapacitor stage for boosted DC voltage smoothing. In addition, this proposed high-frequency resonant inverterhas only one diode conducting mode in the diode bridge rectifier with boost inductor. The operating principle of the proposed series load resonant high frequency inverter is described by using the switching mode equivalentcircuits in addition to the simulated operating voltage and current waveforms. The circuit parameters of proposedhigh frequency inverter design to achieve one diode conduction at the same time at diode bridge par t by

simulation analysis in this paper

InputAC Source Rectifier DC Source

ZVS-PWMInverter

ControlCircuit

IH




71

2. Soft Switching PWM High Frequency Direct Converter

2.1. Circuit description

Figure 1 shows the General Block diagram of single stage AC-Ac Converter for IH Application.Figure2 shows single stage ZVS-PWM high frequency power converter with passive PFC function. Theproposed single stage converter has two circuit parts; passive PFC converter part (see Fig.2 (b)) and highfrequency inverter part (see Fig.2(c)). The unique feature of the proposed high frequency direct inverter in Fig.2(a) includes the direct power frequency conversion processing from UFAC to HFAC, conducting only onediode of diagonal diodes alternatively during the one switching period with passive PFC function and voltageboost function.

The high frequency direct inverter without symmetric bidirectional switches consists of low passfilter with the mid point of C a1, C a2, Lb and La, diode bridge rectifier, boost capacitor C b, boost inductor betweenneutral point of C a1 and C a2 and midpoint of switching bridge leg, power semiconductor switches Q 1, Q2, seriesresonant tuned capacitor C r and IH load ( Ro, Lo).

2.2. Principle of Operation

Figure 3 shows asymmetrical PWM processing signal trains in the case of positive half wave of UFACside voltage source v AC . The operating voltage and current waveforms in steady state are illustrated in Fig.4around a peak value of v AC (t )>0. The definition of the duty factor D is expressed by the following equation.

Ton2 +Td D=--------------- ----- (1)

Twhere, T on2 is gate duration time of the switch S2 of Q2, T d is a dead time between

switches; Q1, Q2, and T is one cycle of HF inverter switching period.

The circuit operating modes and switching mode equivalent circuits of the high frequency series resonantdirect inverter with a new passive PFC rectifier in the case of positive half wave of UFAC voltage vAC . Thecircuit operation in a periodic steady-state is described as follows

Fig 2. a. Single stage high frequency inverter

Fig 2. b. PFC converter part




72

Fig 2. c High frequency inverter part

Fig 3.Asymmetrical PWM gate pulses

[Mode 1; t 0<t _t 1] This circuit operating mode is started by turning-on the switch S1 of Q1. In this mode, diode Ds1

only in the bridge rectifier as a passive PFC also conducts. This mode ends when the switch S 1 is turned off withZVS due to the gate pulse release.[Mode 2; t 1<t _t 2] When the switch S1 is turned off with ZVS, the circuit operating mode is shifted to Mode 2.The switch S1 can achieve ZVS turn off commutation, and at this point, the diode D s1 only conductscontinuously. In this circuit operating mode, the lossless snubbing capacitor C s1 in high side bridge arm ischarged and C s2 in low side bridge arm is discharged simultaneously from a certain boosted voltage vCb. Thisoperating mode ends when the lossless snubbing capacitors complete charging and discharging each other.[Mode 3; t 2<t _t 3] This circuit operating mode starts after charging and discharging of the lossless snubbing

capacitors; C s1 and C s2. The only one diode Ds1 conducts and D2 of Q2 is turned on. This circuit operating modeends when the switch S2 is turned on.[Mode 4; t 3<t _t 4] When the gate driving pulse is delivered to the switch S2 of Q2 during the operating period inMode 3,




73

Fig 4.Operating waveforms

The circuit operation in Mode 4 will start from this point. In this mode, the only one bridge diode D s1 conducts

and switch S2 is turned on with ZVZCS as the complete soft commutation. This circuit operating mode endswhen the only one bridge diode Ds1 is turned off with ZCS for one diode conducting and boost operating modepassive PFC rectifier.[Mode 5; t 4<t _t 5] In this circuit operating mode, the switch S2 conducts and the bridge diode Ds4 in diagonalbridge arm is naturally commutated with ZCS from Ds1. This operating mode ends when the switch S2 is turnedoff with ZVS.[Mode 6; t 5<t _t 6] This circuit operating mode starts when the switch S2 is turned off with ZVS. The only onediode Ds4 conducts continuously. In this circuit operating mode, the lossless snubbing capacitor C s1 in a high sidearm is discharged with the aid of series resonant IH load and, on the other hand, C s2 in the low side arm ischarged at the same time during a dead time interval. This circuit operating mode ends when charging anddischarging of the lossless snubbing capacitors C s1, C s2 are completely performed.[Mode 7; t 6<t _t 7] When the charging and discharging behaviors of two lossless snubbing capacitors arecompleted, the diode D1 of Q1 commutates from C s1 naturally. The low side bridge diode D s4 as a PFC rectifierconducts continuously. This operating mode ends when the only one conducting diode Ds4 is turned off with ZCS

and a high side bridge diode Ds1 is turned on with ZCS.[Mode 8; t 7<t _t 8] In this circuit operating mode, the bridge diode Ds1 is turned on with ZCS and diode D1

conducts. This circuit operating mode ends when the switch S1 of Q1 is turned on with ZVZCS.

3. Simulated Performance Evaluations

ZVS – PWM Inverter system is simulated using simulink and their results are given here Fig 5a. showsthe proposed simulation model. Driving pulses are shown in Fig 5b. Input current and voltage waveform are asshown in Fig 5c. Ouput AC voltage waveforms are shown in Fig 5d. and it enlarged output waveform is shownin Fig 5e. It can be seen that output voltage is almost sine wave. The low frequency AC input voltage is




74

converted to Dc using uncontrolled Rectifier and its output of Rectifier is converter in to high frequency Acusing ZVS_PWM Inverter.

Fig 5. a Simulation of proposed inverter

Fig 5. b. Driving pulses

Fig 5. c. Input current & voltage waveform




75

Fig 5. d. Output voltage waveform

5. e. Enlarged waveform

Fig 5. f. output & input voltage waveform

4. Main Features

The advantageous features are summarized as follows, compared with two stage power frequencyconversion processing scheme. (a) The diodes of the diode bridge circuit are turned on and off with ZCS. The diode recovery currents and theirpower losses could be minimized.(b) Only one diode switch can conduct in bridge rectifier stage for passive PFC converter equipped with utilityAC side grid. The diode conduction losses in the utility side diode bridge rectifier can be reduced in principle.Ds1, Ds4 for v AC >0 and Ds2, Ds3 for v AC <0 can achieve ZCS.(c) The capacitance in DC boost link can be reduced. The film capacitor could be used in replace of the DCelectrolytic capacitor. The ESR of the boosted DC capacitor can be lowered. Its power loss and temperature rise




76

might be minimized. The total power factor in UFAC side becomes unity and line harmonic current componentsin the UFAC side can be reduced without complex specific control procedure with sensor less scheme(d) The high frequency AC power for IH load can be regulated by the simple asymmetrical PWM under theconditions of the soft commutation and constant frequency.(e) Total system efficiency could be higher. The high power density might be achieved under simple coolingscheme and energy saving.(f) The DC component of working coil in IH load can be zero due to the series resonant load with tuned seriescapacitor.(g) The envelope of output high frequency current has the same sine wave as that in UFAC side.

5. Conclusions

This paper has presented analysis, Modeling and Simulation of single stage Ac-Ac converter for a varietyof consumer induction heating (IH) appliances such as IH cooker, IH hot water producer, IH steamer and IHsuper heated steamer. This high frequency series load resonant tuned direct inverter using lossless snubbingcapacitors has a high efficiency passive PFC rectifier operating only one diode conduction and boosting mode inthe diode full bridge rectifier. Furthermore, the proposed direct high frequency inverter circuit topology haslowered line current harmonic contents and THD characteristics in UFAC side, and utility power factorcharacteristics in UFAC side. Also this system has advantages like low switching loss and reduced stress Themodels are developed and they are successfully used for simulation studies. The simulation results are in linewith predictions.

In the future, it should do the comparative studies between the conventional type two-stage highfrequency resonant inverter and proposed one stage high frequency resonant direct inverter on the basis of experimental results.

References

[1] R.Ordonez, H.Calleja, “Induction heating inverter with power factor correction”, VI IEEE InternationalPower Electronics Congress (CIEP)pp.90-95, 1998.

[2] H.Calleja, R.Ordonez, “Improved induction-heating inverter with powerfactor correction”, Proceedings of 30th Annual IEEE Power ElectronicsSpecialists Conference (PESC) Vol.2, pp.1132-1137, 1999.

[3] H.Tanimatsu, T.Ahmed, I.Hirota, K.Yasui, T.Iwai, H.Omori, N.A.Ahmed,H.W.Lee, M.Nakaoka, “Two-switch boost-half bridge and boost active clamped ZVS-PWM AC-AC converters for consumer highfrequency induction heater”, Proceedings of Twentieth Annual IEEE Applied Power Electronics Conferenceand Exposition (APEC), Vol.2, pp.1124-11302005

[4] N.A.Ahmed, Y.Miura, T. Ahmed, E.Hiraki, A.Eid, H.W.Lee, andM.Nakaoka, “Quasi-Resonant Dual ModeSoft Switching PWM and PDM High-Frequency Inverter with IH Load Resonant Tank” Proceedings of IEEE Power Electronics Specialists Conference (PESC), pp.2830-2853, Brazil, June, 2005

[5] H.Sugimura, A.M.Eid, S.K.Kwon, H.W.Lee, E.Hiraki, M, Nakaoka, “High frequency cyclo-converter usingone-chip reverse blocking IGBT based bidirectional power switches”, Proceedings of the EighthInternational Conference on Electrical Machines and Systems (ICEMS), Vol.2 pp.1095-1100, 2005.

[6] B.Saha, H.W.Lee, M.Nakaoka, “Utility Frequency AC to High Frequency AC Power Converter with Boost-Half Bridge Single Stage Circuit Topology”, Proceedings of IEEE International Conference on IndustrialTechnology (ICIT), pp.1430-1435, 2006.

[7] Y.Kawaguchi, E.Hiraki, T.Tanaka, M.Nakaoka, “Full bridge phase-shifted soft switching high-frequencyinverter with boost PFC function for induction heating system” Proceedings of European Conference onPower Electronics and Applications (EPE), pp.1-8, Sept., 2007

[8] C.S.Seo, J.W.Park, K.Y.Sim, H.J.Kim, J.S.Won, D.H.Kim, “A study on Single-Stage High-Power-FactorElectronic Ballast for Discharge Lamps Operating in Critical Conduction Mode,” Transactions of theKorean Institute of Electrical Engineers (KIEE), Society of Electrical Machinery and Energy ConversionSystems (EMECS), Vol.54B-12, pp.601-608, Dec. 2005.

[9] S.H.Ha, C.I.Kim, S.W.K, J.R.Nam, S.P.Mun, “The Improvement Effect of Input Current Waveform of TwoNew Main Switching Boost Rectifiers,” Journal of the Korean Institute of Illuminating and ElectricalInstallation Engineers, Vol.22, No.3, pp.15-26, March 2008.

About AuthorsS.Arumugam has obtained his B.E degree from Bangalore University, Bangalore in the year

1999. He obtained his M.E degree from Sathyabama University; Chennai in the year 2005.Heis presently a research scholar at Bharath University, Chennai. He is working in the area of Resonant inverter fed Induction.




77

S.Ramareddy is Professor of Electrical Department, Jerusalem Engineering College,Chennai. He obtained his D.E.E from S.M.V.M Polytechnic, Tanuku, A.P. A.M.I.E inElectrical Engg from institution of Engineers (India), M.E in Power System from AnnaUniversity. He received Ph.D degree in the area of Resonant Converters from College of Engineering, Anna University, Chennai. He has published over 20 Technical papers inNational and International Conference proceeding/Journals. He has secured A.M.I.E InstitutionGold medal for obtaining higher marks. He has secured AIMO best project award. He has

worked in Tata Consulting Engineers, Bangalore and Anna University, Chennai. His research interest is in thearea of resonant converter, VLSI and Solid State drives. He is a life member of Institution of Engineers (India),Indian Society for India and Society of Power Engineers. He is a fellow of Institution of Electronics andtelecommunication Engineers




78

IJCIIS Reviewers

A. Govardhan, Jawaharlal Nehru Technological University, IndiaAjay Goel, Haryana Institute of Engineering and Technology, IndiaAjay Sharma, Raj Kumar Goel Institute of Technology, IndiaAkshi Kumar, Delhi Technological University, IndiaAlok Singh Chauhan, Ewing Christian Institute of Management and Technology, IndiaAmandeep Dhir, Helsinki University of Technology Finland, Denmark Technical University, DenmarkAmol Potgantwar, Sandip Institute of Technology and Research Centre, IndiaAnand Sharma, MITS, IndiaAos Alaa Zaidan Ansaef, Multimedia University, MalaysiaArul Lawrence Selvakumar, Kuppam Engineering College, IndiaAyyappan Kalyanasundaram, Rajiv Gandhi College of Engineering and Technology, IndiaAzadeh Zamanifar, Iran University of Science and Technology University and Niroo Research Institute, IranBilal Bahaa Zaidan, University of Malaya, MalaysiaB. L. Malleswari, GNITS, IndiaB. Nagraj, Tamilnadu News Prints and Papers, IndiaC. Suresh Gnana Dhas, Vel Tech Multitech Dr.Rengarajan Dr.Sagunthla Engg. College, IndiaC. Sureshkumar, J. K. K. M. College of Technology, IndiaDeepankar Sharma, D. J. College of Engineering and Technology, IndiaDurgesh Kumar Mishra, Acropolis Institute of Technology and Research, IndiaD. S. R. Murthy, SreeNidhi Institute of Science and Technology, India

Hafeez Ullah Amin, KUST Kohat, NWFP, PakistanHanumanthappa Jayappa, University of Mysore, IndiaHimanshu Aggarwal, Punjabi University, IndiaJagdish Lal Raheja, Central Electronics Engineering Research Institute, IndiaJatinder Singh, UIET Lalru, IndiaIman Grida Ben Yahia, Telecom SudParis, FranceLeszek Sliwko, CITCO Fund Services, IrelandM. Azath, Anna University, IndiaMd. Mobarak Hossain, Asian University of Bangladesh, BangladeshMohammed Salem Binwahlan, Hadhramout University of Science and Technology, YemenMohamed Elshaikh, Universiti Malaysia Perlis, MalaysiaM. Surendra Prasad Babu, Andhra University, IndiaM. Thiyagarajan, Sastra University, IndiaManjaiah D. H., Mangalore University, India

Nahib Zaki Rashed, Menoufia Univesity, EgyptNagaraju Aitha, Vaagdevi College of Engineering, IndiaNatarajan Meghanathan, Jackson State University, USAN. Jaisankar, VIT University, IndiaOjesanmi Olusegun Ayodeji, Ajayi Crowther University, NigeriaOluwaseyitanfunmi Osunade, University of Ibadan, NigeriaPerumal Dananjayan, Pondicherry Engineering College, IndiaPiyush Kumar Shukla, University Institute of Technology, Bhopal, IndiaPoonam Garg, Institute of Management Technology, IndiaPraveen Ranjan Srivastava, BITS, IndiaRajesh Kumar, National University of Singapore, SingaporeRajeshwari Hegde, BMS College of Engineering, IndiaRakesh Chandra Gangwar, Beant College of Engineering and Technology, IndiaRaman Kumar, D A V Institute of Engineering and Technology, India

Raman Maini, University College of Engineering, Punjabi University, IndiaRamveer Singh, Raj Kumar Goel Institute of Technology, IndiaSateesh Kumar Peddoju, Vaagdevi College of Engineering, IndiaShahram Jamali, University of Mohaghegh Ardabili, IranSriman Narayana Iyengar, IndiaSuhas Manangi, Microsoft, IndiaSujisunadaram Sundaram, Anna University, IndiaSukumar Senthilkumar, National Institute of Technology, IndiaS. S. Mehta, J. N. V. University, IndiaS. Smys, Karunya University, IndiaS. V. Rajashekararadhya, Adichunchanagiri Institute of Technology, India




Thipendra P Singh, Sharda University, IndiaT. Ramanujam, Krishna Engineering College, Ghaziabad, IndiaT. Venkat Narayana Rao, Hyderabad Institute of Technology and Management, IndiaVasavi Bande, Hyderabad Institute of Technology and Management, IndiaVishal Bharti, Dronacharya College of Engineering, IndiaV. Umakanta Sastry, Sreenidhi Institute of Science and Technology, India