Research Article A Cross-Layer Design Combining of AMC ...downloads.hindawi.com/journals/ijdsn/2013/145254.pdf · International Journal of Distributed Sensor Networks Encoding Channel

Hindawi Publishing CorporationInternational Journal of Distributed Sensor NetworksVolume 2013 Article ID 145254 8 pageshttpdxdoiorg1011552013145254

Research ArticleA Cross-Layer Design Combining of AMC with HARQ forDSRC Systems

Gao Yuan Zhang1 Li Min Sun2 Hong Wen1 Bin Wu3 Xiping Zhu1 and Liang Zhou1

1 National Key Laboratory of Science and Technology on Communications University of Electronic Science and Technology of ChinaChengdu 610054 China

2 Institute of Software Chinese Academy of Sciences Beijing 100190 China3 School of Computer Science and Technology Tianjin University Tianjin 300072 China

Correspondence should be addressed to Hong Wen wcdma 2000hotmailcom

Received 12 April 2013 Accepted 30 September 2013

Academic Editor Liusheng Huang

Copyright copy 2013 Gao Yuan Zhang et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

A new cross-layer design combining adaptive modulation and coding (AMC) with Hybrid Automatic Repeat Qequest (HARQ)based on rate-compatible LDPC codes is proposed for IEEE Std 80211p in DSRC systems Instead of considering AMC at thephysical layer andARQat the data link layer separately we propose a cross-layer design that combines these two layers judiciously tomaximize either spectral efficiency or throughputWith ARQ correcting occasional packet errors at the data link layer the stringenterror control requirement is alleviated for the AMC at the physical layer Numerical results demonstrate that the proposed AMC-HARQ design outperforms either application of AMC only at the physical layer or incorporation of ARQ with fixed modulationand coding scheme

1 Introduction

Vehicular ad hoc networks (VANETs) are a type of mobilead hoc networks where each vehicle serves as a nodeinterconnected by wireless links One of the most importantfeatures in VANETs is that each vehicle can only move ina predictable manner but at much higher speeds comparedwith the traditional mobile ad hoc networks (MANETs) TheDedicated Short Range Communication (DSRC) standard[1] which is currently under extensive development by theIEEE 80211p [2] standardization committee defines twotypes of communication in VANETs vehicle to vehicle (V2V)and vehicle to infrastructure (V2I) An excellent overviewon DSRC technology is also given in [3] which first givesa general description of the architecture by introducing theconcepts applications and characteristics of the technologyto be used for V2V and V2I communications DSRC iscommitted to support a suite of safety applications such ascollision warning up-to-date traffic information and activenavigation and infotainment With DSRC each vehicle onthe road is broadcasting routine traffic related messages with

the information of position current time instance drivingdirection speed accelerationdeceleration and possible traf-fic conditions The frequency spectrum between 5850 and5925GHz is allocated for DSRC which will enable vehiclesto communicate with the road infrastructure and allows fora large number of ITS applications This provides an oppor-tunity for automakers government agencies and relatedcommercial entities to improve highway safety Howeverintervehicle communications must operate effectively withintransmit power limits andunder received signal strength fluc-tuations and Doppler spread DSRC for intervehicle wirelesscommunications can provide numerous safety applicationsbut these require reliable communications at a reasonablecost

The DSRC standard employs convolutional codes forForward Error Correction (FEC) The performance analysisof them has been extensively studied in the literature suchas [4ndash9] For instance a feasibility study of delay-criticalsafety applications over vehicular ad hoc networks based onthe emerging DSRC standard is conducted in [4] Underthe current static backoff schemes the infrastructure data

2 International Journal of Distributed Sensor Networks

collection mode of IEEE 80211p standard does not performwell [5] The performance of the DSRC system is evaluatedunder three different channels with convolutional codesregular LDPC codes and quasi-cyclic (QC) LDPC codes[6 7] It has shown that LDPC codes provide an attractivetradeoff between performance and complexity and should beconsidered as alternative error correction codes for DSRCsystems Similarly Amditis and Uzunoglu [8] simulated thephysical layer (PHY) of the upcoming vehicular communi-cation standard IEEE 80211p in V2V situation through twodifferent scenarios under different modulation schemes In[9] the results of evaluation of the performance of IEEE80211p PHY employing turbo coding are presented In allthese results the LDPC codes are considered the bettercandidates due to their good performance and low encodingand decoding complexity which is important to the real timecommunications in the VANETs

In wireless communication networks the demand forhigh data rates and quality of service (QoS) is growing ata rapid pace The cross-layer design approaches are likelyto provide much better results in practice The study of[10] tries to achieve links scheduling and power assignmentwhile meeting the data rate and peak power level constraintssuch that the resulting throughput is maximized In [11] across-layer design along with an optimal resource allocationframework is formulated for wireless fading networks byemploying the network coding However these cross-layerdesign methods are notoriously complicated and difficult totranslate into practice

The adaptive modulation and coding (AMC) [12] havebeen studied extensively and advocated at the physicallayer which can enhance throughput in future wireless datacommunication systemThe automatic repeat request (ARQ)protocol at the data link layer that requests retransmissionsfor those received packers with error is also very importantto throughput enhancement It has proved that joint AMC-ARQ [13] design outperforms either application of AMConly at physical layer or ARQ only with a fixed modulationand coding scheme In this paper we consider the cross-layer design combining AMC and Incremental RedundancyHARQ (IR HARQ) which is the most efficient ARQ schemeIn physical layer the AMC adjusts the data rates roughlyaccording to the Channel State Information (CSI) by choos-ing different modulation modes while the data rates areaccurately adjusted by HARQ according to the CSI withchanging the maximum retransmission number in data linklayer One of the key problems for realizing AMC IR HARQis rate-compatible (RC) codes which consist of a low-ratemother code and several higher rates achieved throughcompatible puncturingHence the decoder for the lowest ratecode is compatible with that for the higher rate codes and noadditional complexity is needed In this work we will employRC-LDPC codes as the FEC codes for DSRC systems Bytaking advantage of the cross-layer design combining AMCand IR HARQ the throughput is enhanced with low systemcomplexity

The rest of the paper is organized as follows An overviewof DSRC system is discussed in Section 2 Section 3 describesthe proposed cross-layer design combiningmethods of AMC

Table 1 DSRC physical layer parameters

Modulation BPSK QPSK 16 QAM 64 QAMData rate 3 45 6 9 12 18 27MbpsCoding rate 12 23 34Number of subcarriers 52Subcarrier spacing 15625 KHzNumber of pilot tones 4Guard interval 16120583secOFDM symbol duration 8 120583secSignal bandwidth 10MHz

Bit stream Scrambler FEC coder Blockinterleaver

Pilotinsertion

Cyclic prefixinsertion IFFT Symbol

mapping

Figure 1 The DSRC transmitter model

and IR HARQ The simulation results are presented anddiscussed in Section 4 Section 5 concludes our work

2 Overview of DSRC System

DSRC physical layer has the same frame structure modula-tion and training sequences as specified in the IEEE 80211astandard and its signal bandwidth is 10MHz (half the IEEE80211a bandwidth) The basic DSRC parameters are shownin Table 1

The block diagrams of the DSRC transmitter and receiverare shown in Figures 1 and 2 respectively [6 7] Theinput bit stream is first scrambled and then encoded witha convolution code for which the constraint length is 7and code rate is 12 The interleaving redistributes the bitsbefore transmission which reduces the effects of burst errorscaused by the fading channel Perfect timing and frequencysynchronization is also assumed and the received signal istransformed to the frequency domain using a Fast FourierTransform (FFT) The resulting signal is then mapped to bitsand deinterleaved Finally the soft information is input tochannel decoder and the output bits descrambledTheoutputof the scrambler is encoded using code rates 119877 = 12 23or 34 depending on the desired data rate The convolutionencoder has rate 119877 = 12 and constraint length 7 (memorylength119898 = 6) as shown in Figure 3Higher rates are obtainedby puncturing the output bit stream

3 The Cross-Layer Design Combining AMCand IR_HARQ

31 AMC Method Based on LDPC Codes In this section weprovide the AMCmethod based on LDPC codes which havedemonstrated impressive error-correcting qualities underBPSK modulation and are also good codes to use in com-bination with higher-order modulations Figure 4 showsa rate-adaptive coded modulation communication system

International Journal of Distributed Sensor Networks 3

Packet detection estimation

Cyclic prefixremoval FFT

De-scrambler

FECdecoder

Block de-interleaver

Channelestimation

SymboldemappingBit sink

T and F

Figure 2 The DSRC receiver model

Input data

Output data A

Output data B

Tb Tb Tb Tb Tb Tb

Figure 3 The convolutional encoder for a DSRC system based onthe IEEE 80211p standard

The encoded bits are mapped into M-QAM constellationsaccording to Gray code mappings For each M-QAM con-stellation we construct an LDPC code such that the lengthsof all codes are integer multiple of M-QAM symbols in thecorresponding constellation

Here our objective is to maximize the data rate whilemaintaining the required performance Let 119873 denote thetotal available number of transmission modes We assumeconstant power transmission and separate the total SNRrange into 119873 + 1 with no overlapping consecutive intervalswith boundary points denoted as 120574

119894119873

119894=0 If the channel

estimated SNR is 120574 we have

120574 isin [120574119894 120574119894+1

) mode 119894 is chosen (1)

The boundary point 120574119894is determined for that the instan-

taneous block error rate (BLER) or bit error rate (BER) isguaranteed to be not greater than target BER

0or BLER

0 So

the value of 120574119894is obtained by solving the following equation

for each mode

BER (120574119894) = BER

0 (2)

or

BLER (120574119894) = BLER

0 (3)

as the functions of the estimated SNR 120574 BER(120574119894) and

BLER(120574119894) are the BER and BLER for mode 119894 respectively

32 AMC-HARQ Design In this section we develop a cross-layer design which combines AMC at physical layer withHARQ at data link layer in order to maximize systemthroughput under performance requirements

Encoding

Channelanalyzer

DemodulateDecoding

CSI

Data

Data

z

aM-QAM

Figure 4 The ACM system based on LDPC codes

Since only finite delays and buffer sizes can be affordedin practice the maximum number of ARQ retransmissionshas to be bounded We denote the maximum number ofretransmissions allowed per block by119873

max Note that a blockis dropped if it is received incorrectly after a maximumnumber of 119873

max+ 1 transmissions that is after 119873

max

retransmissions Let BLERdenote the average block error rateand BLER = 119901 The average number of transmissions perblock is computed as

119873(119901119873max

) = 1 + 119901 + 1199012+ sdot sdot sdot + 119901

119873max

(4)

When mode 119894 is used each transmitted symbol will carry119877119894= 119877119888log2119872119894information bits for mode adhering to a 119872

119894-

QAM constellation and a code with rate 119877119888 Define 119875

119903(119894) as

the probability that the mode 119894 will be chosen Therefore theaverage spectral efficiency achieved at physical layer withoutconsidering possible retransmission is

119878119890phy =

119873

sum

119894=1

119877119894119875119903 (119894) (5)

When HARQ is implemented each block and thus eachinformation bit is equivalently transmitted 119873(119901119873

max)

times Hence the overall average spectral efficiency as afunction of 119873max is given by

119878119890link =

119878119890phy

119873(119901119873max)=

1

119873 (119901119873max)

119873

sum

119894=1

119877119894119875119903 (119894) (6)

For a certain channel and modulation mode the averagespectral efficiency 119878

119890phy is a constant So from (6) we can opti-mize the overall average spectral efficiency 119878

119890link at data linklayer by changing the maximum number of retransmissions119873

max The system model and layer structure of our cross-layer design are shown in Figures 5 and 6 respectively Let119868 denote the total number of AMC available modes For eachmode we define themaximumnumber of retransmissions forIR HARQ as 119873

max If 119873max is 119869119894under 119894th AMC mode the

total available number of transmissionmodes of this system is1198731015840= 1198691+1198692+sdot sdot sdot+119869

119894+sdot sdot sdot+119869

119868We assume that the instantaneous

BLER is guaranteed to be not greater than target BLER0at


Encoding

Channelanalyser

DemodulateDecoding

CSI

Data

AchieveYesNo

ARQ

Data

M-QAM

z

a

BLERlink

Figure 5 Cross-layer combining AMC with HARQ transmissionsystem

Data link layer

Physical layer

Higher layers

HARQ

AMC

Higher layers

Data link layer

Physical layer

Figure 6 Cross-layer structure combining AMC with HARQ

physical layer The BLER is not greater than BLER119873max+1

0at

data link layer So we obtain

BLER119873max+1

0le BLERlink (7)

From (7) we have

BLER0le BLER(1(119873

max+1))

link (8)

The whole SNR range is divided into 119873 + 1 intervalswith thresholds 120574

119894 119894 = 0 1 119873The instantaneous channel

signal to noise ratio SNR is 120574 When 120574 isin [120574119894 120574119894+1

) mode 119894 ischosenThe threshold 120574

119894is determined according to (3) In (3)

the target BLER0at physical layer is determined according to

(8)By always selecting the acceptable code producing the

highest spectral efficiency for a given channel estimatedSNR value we will maximize the spectral efficiency of theoverall system while the BER or BLER requirement is stillmaintained

4 Performances

41 Extraction of the LLR of QAM Modulation In LDPCcoded QAM system the calculation of the Log-LikelihoodRatio (LLR) on the coded bits is an important function ofpractical wireless receivers In this section the exact bit LLRfor the M-QAM signal is presented [14]

As an example let us consider a 16-QAM transmittedsymbol sequence 119904

119896= 119904119868119896

+ 119895119904119876119896 with 119904

119868119896 119904119876119896

isin plusmn1 plusmn3

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

Figure 7 16-QAM bit to symbol mapping

According to the mapping rule 119904119896

= 119872(1199061198961 1199061198962 1199061198963 1199061198964)

each modulation symbol is obtained from the informationbits 119906

119896119894 119894 = 1 4 Without loss of generality we assume

119904119868119896

= 119872119868(1199061198961 1199061198963) and 119904

119876119896= 119872119876(1199061198962 1199061198964) An example of

Gray code mapping is given in Figure 7Let 119883

119896denote the baseband received signal sample

corresponding to symbol 119904119896

119883119896= 119909119868119896

+ 119895119909119876119896

= 119886radic119864119904119904119896+ 119899119896 (9)

In (9) 119886 isin 119877+ 119864119904represents the received signal symbol

energy and 119899119896= 119899119868119896

+ 119895119899119876119896

is additive white complex noisewith 119899

119868119896and 119899119876119896

being independent Gaussian processes withzero mean and variance 120590

2= 11987302

Let Λ(119906119896119894) indicate the LLR of bit 119906

119896119894given119883

119896 and in the

case of equiprobable symbols

Λ (1199061198961) = log

119875 119906119896119894

= 1 | 119883119896

119875 119906119896119894

= 0 | 119883119896

= log119875 119883119896| 119906119896119894

= 1

119875 119883119896| 119906119896119894

= 0 (10)

Define by 119878(119906119896119894

= 1) = 119902 119904(119902)

= 119872( 119906119896119894

= 1 ) and119878(119906119896119894

= 0) = 119902 119904(119902)

= 119872( 119906119896119894

= 0 ) the subsetsof symbol indexes corresponding to 119906

119896119894= 1 and 119906

119896119894= 0

respectively then

119875 119883119896| 119906119896119894

= 1 = sum

119902isin119878(119906119896119894=1)

119901 (119883119896| 119904(119902)

)

119875 119883119896| 119906119896119894

= 0 = sum

119902isin119878(119906119896119894=0)

119901 (119883119896| 119904(119902)

)

(11)


1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(b)

Figure 8 Symbol set partitioning for bit 1199061198961

and 1199061198962

1010

1

3111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

1

3

minus3

minus1

(b)


and 1199061198964

Figures 8 and 9 show the symbol set partitioning for the bits119906119896119894 119894 = 1 4 respectively From the assumption on the

noise 119899119896 we have

119901 (119883119896| 119904(119902)

) =1

1205871198730

exp(minus

10038161003816100381610038161003816119883119896minus 119886radic119864

119904119904(119902)10038161003816100381610038161003816

2

1198730

) (12)

Therefore straightforward calculation gives

Λ (1199061198961) = log 119890

minus(119883119896minus119886119896 sdot1sdotradic119864119904)221205902

+ 119890minus(119883119896minus119886119896 sdot3sdotradic119864119904)

221205902

119890minus(119883119896+119886119896 sdot1sdotradic119864119904)221205902

+ 119890minus(119883119896+119886119896 sdot3sdotradic119864119904)221205902

Λ (1199061198962) = log 119890


+ 119890minus(119909119896+119886119896 sdot1sdotradic119864119904)

221205902

119890minus(119909119896minus119886119896 sdot3sdotradic119864119904)221205902



Table 2 Transmission modes in the AMC system under AWGN channel

Model MCS1 MCS2 MCS3 MCS4 MCS5 MCS6Modulation 4QAM 4QAM 16QAM 16QAM 64QAM 64QAMCoding rate 12 23 23 34 34 56Spectral efficiency 100 133 267 300 450 500SNR thresholds 18 34 94 106 156 167

Table 3 Transmission modes cross-layer combining AMC with HARQ (the maximum number of retransmissions119873max is fixed)

Model MCS1 MCS2 MCS3Modulation 4QAM 16QAM 64QAMCoding rate 56sim12 56sim12 56sim12Spectral efficiency 16667sim100 3333sim2 5sim3Thresholds SNR under AWGN channel 48 109 16

Table 4 Transmission modes cross-layer combining AMC with HARQ under AWGN channel (the maximum number of retransmissions119873

max is changeable)

Model MCS1 MCS2 MCS3Modulation 4QAM 16QAM 64QAMCoding rate 56sim12 56sim12 56sim12Spectral efficiency 16667sim100 3333sim2 5sim3Maximum number of retransmissions 4 3 2 4 3 2 4 3 2SNR thresholds 36 49 595 10 11 125 151 161 174

Λ (1199061198963) = log 119890



221205902



Λ (1199061198964) = log 119890



221205902



(13)The similar conclusion can be obtained for any otherMQAMschemes in DSRC standard

42 Simulation Results To evaluate the performance of ouralgorithm we have performed the simulations under AWGNchannel The transmission modes of AMC system are listedin Table 2 in ascending order of rate The (2304 1920) LDPCcode with rate 56 is employed as a mother code fromwhich (2304 1728) code with 34 rate (2304 1536) codewith 23 rate and (2304 1152) code with 12 rate [15] canbe obtained by puncturing Decoded with Belief Propagation(BP) algorithm [16] is set to be the maximum numberof iterations equal to 50 BER and BLER performances ofmodulation modes in Table 2 under AWGN channel andRayleigh channel are shown in Figure 10 Let the constrainedperformance in physical layer be BER

0= 10minus3 from which

we can determine the thresholds 120574119894with (2)

The transmission modes of AMC-HARQ system withfixed maximum number of retransmissions based on RC-LDPC codes are listed in Table 3We still use (2304 1920) RC-LDPC codes with 56 rate as FEC from which LDPC codeswith 12 23 34 and 45 rate can be obtained by puncturingTherefore the maximum number of retransmissions of ARQis 4The BLER for 56 rate LDPC code as well as its punctured

0 2 4 6 8 10 12 14 16 18SNR (dB)

BER

minus210minus6

10minus5

10minus4

10minus3

10minus2

10minus1

100

MCS1MCS2MCS3

MCS4MCS5MCS6

Figure 10 BER performance of modulationmodes in Table 2 underAWGN channel

codes under AWGN channel is shown in Figure 11 Let theconstrained performance in data link layer be BLER

0= 10minus2

From (8) the performance target BLER in physical layer isBLER

0le 0001

15= 02512 Let 45 rate punctured code as

criterion thresholds 120574119894can be determined according to (3)

which is shown in Table 3


2 4 6 8 10 12 14 16 18SNR (dB)

MCS1 34MCS1 45MCS1 56MCS2 34MCS2 45

MCS2 56MCS3 34MCS3 45MCS3 56

BLER

10minus5

10minus4

10minus3

10minus2

10minus1

100BLER = 02512

1205741 34 = 23dB

Figure 11 BLER for 23 rate LDCP code and its punctured codesunder AWGN channel

0 2 4 6 8 10 12 14 16 18

AMC

SNR (dB)minus2

0

05

1

15

2

25

3

35

4

45

5

Thro

ughp

ut

AMC + HARQ with changeable Nmax

AMC + HARQ with fixed Nmax

Figure 12 Average spectral efficiency versus average SNR for threedifferent transmission systems under AWGN channel

The transmission modes of AMC-HARQ system withchangeable maximum number of retransmissions based onRC-LDPC codes are listed in Table 4 RC-LDPC codes (23041920) [15] with 56 rate also are employed as FEC The totalnumber of modulation modes is MCS1 MCS2 and MCS3We consider three values for the maximum numbers ofretransmissions 119873

max= 2 3 and 4 for each modulation

mode Given the performance constraint at the data link layeris BLER

0= 10minus2 we can get the performance target BLER

in physical layer for each mode and the thresholds 120574119894can

be obtained according to (3) which are shown in Table 4The simulations are performed for comparing the throughput

0 2 6 8 10 12 14 16 180

05

1

15

2

25

3

35

4

45

5

SNR (dB)

Thro

ughp

ut

4minus2

AMC + HARQ with CCAMC + HARQ with TCAMC + HARQ with LDPC

Figure 13 Average spectral efficiency versus average SNR underthree channel coding methods

of three systems under AWGN channel We assume perfectchannel estimation so that the receiver can obtain the exactchannel quality SNR and the feedback channel is error-free The stopping-waiting IR HARQ protocols are adoptedThe maximum iterative decoding numbers of LDPC codesare 50 The results are shown in Figure 12 from whichwe can know that AMC-HARQ system with changeablemaximum number of retransmissions outperforms the othertwo systems We replace the LDPC codes by rate compatibleconvolutional codes (CC) and turbo codes (TC) in [2] andthe simulation results are shown in Figure 13 The proposedAMC-HARQ system based on LDPC codes illustrates thebest performance

5 Conclusions

In this paper we developed a cross-layer design which com-bines adaptive modulation and coding at the physical layerwith IR HARQ at the data link layer in order to maximizesystem spectral efficiency under prescribed performanceconstraints In physical layer the AMC adjusts the data ratesroughly according to the CSI by choosing different modula-tion modes In data link layer HARQ adjusts the data rateaccurately according to the CSI by changing the maximumretransmission number Numerical results demonstrated therate improvement of our cross-layer design over AMC aloneas well as AMC-HARQ with fixed maximum number ofretransmissions In our proposed scheme we employed RC-LDPC as channel codes due to their decoder for the lowestcode rate being compatible with the ones for higher coderates Our new method can improve the performance andno additional complexity is needed All of our simulationsare conducted under AWGN channel and in the future


we intend to perform further experiments under a morerealistic channel model for the DSRC system

Acknowledgments

The work is supported by the NSFC (Grant nos 6103200361271172 and 61071100) RFDP (Grant nos 20120185110025and 20120185110030) NCET (Grant no NCET-09-0266) andSRF for ROCS SEM

References

[1] C Cseh ldquoArchitecture of the dedicated short-range commu-nications (DSRC) protocolrdquo in Proceedings of the 48th IEEEVehicular Technology Conference (VTC rsquo98) pp 2095ndash2099Ottawa Canada May 1998

[2] Task Group p ldquoIEEE P80211p Draft Standard for InformationTechnology Telecommunications and information exchangebetween systems Local andmetropolitan area networks Specificrequirements Part 11 Wireless LAN Medium Access Control(MAC) and Physical Layer (PHY) specificationsrdquo IEEE Com-puter Society 2009

[3] R A Uzcategui and G Acosta-Marum ldquoWave a tutorialrdquo IEEECommunications Magazine vol 47 no 5 pp 126ndash133 2009

[4] J Yin T Elbatt G Yeung et al ldquoPerformance evaluation ofsafety applications over DSRC vehicular ad hoc networksrdquo inProceedings of the 1st ACM International Workshop on VehicularAdHoc Networks pp 1ndash9 Philadelphia Pa USA October 2004

[5] Y Wang A Ahmed B Krishnamachari and K Psounis ldquoIEEE80211p performance evaluation and protocol enhancementrdquo inProceedings of the IEEE International Conference on VehicularElectronics and Safety (ICVES rsquo08) pp 22ndash24 Columbus OhioUSA September 2008

[6] N Khosroshahi and T A Gulliver ldquoLow density parity checkcodes for dedicated short range communication (DSRC) sys-temsrdquo in Proceedings of the IEEE Pacific Rim Conference onCommunications Computers and Signal Processing (PACRIMrsquo09) pp 802ndash807 August 2009

[7] N Khosroshahi and T A Gulliver ldquoQuasi-cyclic low densityparity check (LDPC) codes for dedicated short range commu-nication (DSRC) systemsrdquo in Proceedings of the 23rd CanadianConference on Electrical and Computer Engineering (CCECErsquo10) pp 1ndash5 May 2010

[8] A Amditis and N K Uzunoglu ldquoSimulation-based perfor-mance analysis and improvement of orthogonal frequencydivision multiplexingmdash80211p system for vehicular communi-cationsrdquo IET Intelligent Transport Systems vol 3 no 4 pp 429ndash436 2009

[9] G C Kiokes G Economakos A Amditis and N KUzunoglu ldquoRecursive systematic convolutional code simulationfor Ofdmmdash80211p system and FPGA implementation usingan ESL methodologyrdquo in Proceedings of the 12th EuromicroConference onDigital SystemDesign ArchitecturesMethods andTools (DSD rsquo09) pp 791ndash798 August 2009

[10] G Kulkarni V Raghunathan and M Srivastava ldquoJoint end-to-end scheduling power control and rate control in multi-hop wireless networksrdquo in Proceedings of the IEEE GlobalTelecommunications Conference (GLOBECOM rsquo04) vol 5 pp3357ndash3362 December 2004

[11] K Rajawat N Gatsis and G B Giannakis ldquoCross-layer designsin coded wireless fading networks with multicastrdquo IEEEACMTransactions on Networking vol 19 no 5 pp 1276ndash1289 2011

[12] M-S Alouini and A J Goldsmith ldquoAdaptive modulation overNakagami fading channelsrdquoWireless Personal Communicationsvol 13 no 1 pp 119ndash143 2000

[13] Y Z Bang F Pin and C Z Gang ldquoCross layer design forservice diferentiation in mobile ad hoc networksrdquo in ProceedingofIEEEInternational Symposium on Personal Indoor andMobileRadio Communication vol 1 pp 778ndash782 2003

[14] S Allpress C Luschi and S Felix ldquoExact and approximatedexpressions of the log-likelihood ratio for 16-QAM signalsrdquoin Proceedings of the Conference Record of the 38th AsilomarConference on Signals Systems and Computers pp 794ndash798Pacific Grove Calif USA November 2004

[15] Draft IEEE Standard for Local andmetropolitan area networks-Part 16 ldquoAir Interface for Fixed andMobile BroadbandWirelessAccess Systemsrdquo IEEE P80216eD12 2005

[16] D J C MacKay ldquoGood codes based on very sparse matricesrdquoIEEE Transactions on Information Theory vol 45 no 2 pp1645ndash1646 1999

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics


Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks



collection mode of IEEE 80211p standard does not performwell [5] The performance of the DSRC system is evaluatedunder three different channels with convolutional codesregular LDPC codes and quasi-cyclic (QC) LDPC codes[6 7] It has shown that LDPC codes provide an attractivetradeoff between performance and complexity and should beconsidered as alternative error correction codes for DSRCsystems Similarly Amditis and Uzunoglu [8] simulated thephysical layer (PHY) of the upcoming vehicular communi-cation standard IEEE 80211p in V2V situation through twodifferent scenarios under different modulation schemes In[9] the results of evaluation of the performance of IEEE80211p PHY employing turbo coding are presented In allthese results the LDPC codes are considered the bettercandidates due to their good performance and low encodingand decoding complexity which is important to the real timecommunications in the VANETs

In wireless communication networks the demand forhigh data rates and quality of service (QoS) is growing ata rapid pace The cross-layer design approaches are likelyto provide much better results in practice The study of[10] tries to achieve links scheduling and power assignmentwhile meeting the data rate and peak power level constraintssuch that the resulting throughput is maximized In [11] across-layer design along with an optimal resource allocationframework is formulated for wireless fading networks byemploying the network coding However these cross-layerdesign methods are notoriously complicated and difficult totranslate into practice

The adaptive modulation and coding (AMC) [12] havebeen studied extensively and advocated at the physicallayer which can enhance throughput in future wireless datacommunication systemThe automatic repeat request (ARQ)protocol at the data link layer that requests retransmissionsfor those received packers with error is also very importantto throughput enhancement It has proved that joint AMC-ARQ [13] design outperforms either application of AMConly at physical layer or ARQ only with a fixed modulationand coding scheme In this paper we consider the cross-layer design combining AMC and Incremental RedundancyHARQ (IR HARQ) which is the most efficient ARQ schemeIn physical layer the AMC adjusts the data rates roughlyaccording to the Channel State Information (CSI) by choos-ing different modulation modes while the data rates areaccurately adjusted by HARQ according to the CSI withchanging the maximum retransmission number in data linklayer One of the key problems for realizing AMC IR HARQis rate-compatible (RC) codes which consist of a low-ratemother code and several higher rates achieved throughcompatible puncturingHence the decoder for the lowest ratecode is compatible with that for the higher rate codes and noadditional complexity is needed In this work we will employRC-LDPC codes as the FEC codes for DSRC systems Bytaking advantage of the cross-layer design combining AMCand IR HARQ the throughput is enhanced with low systemcomplexity

The rest of the paper is organized as follows An overviewof DSRC system is discussed in Section 2 Section 3 describesthe proposed cross-layer design combiningmethods of AMC

Table 1 DSRC physical layer parameters

Modulation BPSK QPSK 16 QAM 64 QAMData rate 3 45 6 9 12 18 27MbpsCoding rate 12 23 34Number of subcarriers 52Subcarrier spacing 15625 KHzNumber of pilot tones 4Guard interval 16120583secOFDM symbol duration 8 120583secSignal bandwidth 10MHz

Bit stream Scrambler FEC coder Blockinterleaver

Pilotinsertion

Cyclic prefixinsertion IFFT Symbol

mapping

Figure 1 The DSRC transmitter model

and IR HARQ The simulation results are presented anddiscussed in Section 4 Section 5 concludes our work

2 Overview of DSRC System

DSRC physical layer has the same frame structure modula-tion and training sequences as specified in the IEEE 80211astandard and its signal bandwidth is 10MHz (half the IEEE80211a bandwidth) The basic DSRC parameters are shownin Table 1

The block diagrams of the DSRC transmitter and receiverare shown in Figures 1 and 2 respectively [6 7] Theinput bit stream is first scrambled and then encoded witha convolution code for which the constraint length is 7and code rate is 12 The interleaving redistributes the bitsbefore transmission which reduces the effects of burst errorscaused by the fading channel Perfect timing and frequencysynchronization is also assumed and the received signal istransformed to the frequency domain using a Fast FourierTransform (FFT) The resulting signal is then mapped to bitsand deinterleaved Finally the soft information is input tochannel decoder and the output bits descrambledTheoutputof the scrambler is encoded using code rates 119877 = 12 23or 34 depending on the desired data rate The convolutionencoder has rate 119877 = 12 and constraint length 7 (memorylength119898 = 6) as shown in Figure 3Higher rates are obtainedby puncturing the output bit stream

3 The Cross-Layer Design Combining AMCand IR_HARQ

31 AMC Method Based on LDPC Codes In this section weprovide the AMCmethod based on LDPC codes which havedemonstrated impressive error-correcting qualities underBPSK modulation and are also good codes to use in com-bination with higher-order modulations Figure 4 showsa rate-adaptive coded modulation communication system




De-scrambler

FECdecoder


Channelestimation


T and F


Input data

Output data A

Output data B

Tb Tb Tb Tb Tb Tb




119894119873



120574 isin [120574119894 120574119894+1




0or BLER

0 So


for each mode

BER (120574119894) = BER

0 (2)

or

BLER (120574119894) = BLER

0 (3)




Encoding

Channelanalyzer

DemodulateDecoding

CSI

Data

Data

z

aM-QAM





max


119873(119901119873max

) = 1 + 119901 + 1199012+ sdot sdot sdot + 119901

119873max

(4)


119894-


119903(119894) as


119878119890phy =

119873

sum

119894=1

119877119894119875119903 (119894) (5)


max)


119878119890link =

119878119890phy

119873(119901119873max)=

1

119873 (119901119873max)

119873

sum

119894=1

119877119894119875119903 (119894) (6)











Encoding

Channelanalyser

DemodulateDecoding

CSI

Data

AchieveYesNo

ARQ

Data

M-QAM

z

a

BLERlink


Data link layer

Physical layer

Higher layers

HARQ

AMC

Higher layers

Data link layer

Physical layer



0at


BLER119873max+1

0le BLERlink (7)

From (7) we have


max+1))

link (8)









4 Performances



119896= 119904119868119896

+ 119895119904119876119896 with 119904

119868119896 119904119876119896


1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1



= 119872(1199061198961 1199061198962 1199061198963 1199061198964)



119904119868119896

= 119872119868(1199061198961 1199061198963) and 119904

119876119896= 119872119876(1199061198962 1199061198964) An example of




119883119896= 119909119868119896

+ 119895119909119876119896

= 119886radic119864119904119904119896+ 119899119896 (9)


energy and 119899119896= 119899119868119896

+ 119895119899119876119896


119868119896and 119899119876119896


2= 11987302


119896119894given119883

119896 and in the


Λ (1199061198961) = log

119875 119906119896119894

= 1 | 119883119896

119875 119906119896119894

= 0 | 119883119896

= log119875 119883119896| 119906119896119894

= 1

119875 119883119896| 119906119896119894

= 0 (10)

Define by 119878(119906119896119894

= 1) = 119902 119904(119902)

= 119872( 119906119896119894

= 1 ) and119878(119906119896119894

= 0) = 119902 119904(119902)

= 119872( 119906119896119894


119896119894= 1 and 119906

119896119894= 0

respectively then

119875 119883119896| 119906119896119894

= 1 = sum

119902isin119878(119906119896119894=1)

119901 (119883119896| 119904(119902)

)

119875 119883119896| 119906119896119894

= 0 = sum

119902isin119878(119906119896119894=0)

119901 (119883119896| 119904(119902)

)

(11)


1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(b)


and 1199061198962

1010

1

3111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

1

3

minus3

minus1

(b)


and 1199061198964



119901 (119883119896| 119904(119902)

) =1

1205871198730

exp(minus

10038161003816100381610038161003816119883119896minus 119886radic119864

119904119904(119902)10038161003816100381610038161003816

2

1198730

) (12)


Λ (1199061198961) = log 119890



221205902



Λ (1199061198962) = log 119890



221205902









max is changeable)


Λ (1199061198963) = log 119890



221205902



Λ (1199061198964) = log 119890



221205902








0 2 4 6 8 10 12 14 16 18SNR (dB)

BER

minus210minus6

10minus5

10minus4

10minus3

10minus2

10minus1

100

MCS1MCS2MCS3

MCS4MCS5MCS6



0= 10minus2


0le 0001





2 4 6 8 10 12 14 16 18SNR (dB)



BLER

10minus5

10minus4

10minus3

10minus2

10minus1

100BLER = 02512

1205741 34 = 23dB


0 2 4 6 8 10 12 14 16 18

AMC

SNR (dB)minus2

0

05

1

15

2

25

3

35

4

45

5

Thro

ughp

ut










0 2 6 8 10 12 14 16 180

05

1

15

2

25

3

35

4

45

5

SNR (dB)

Thro

ughp

ut

4minus2




5 Conclusions




Acknowledgments


References



















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation












De-scrambler

FECdecoder


Channelestimation


T and F


Input data

Output data A

Output data B

Tb Tb Tb Tb Tb Tb




119894119873



120574 isin [120574119894 120574119894+1




0or BLER

0 So


for each mode

BER (120574119894) = BER

0 (2)

or

BLER (120574119894) = BLER

0 (3)




Encoding

Channelanalyzer

DemodulateDecoding

CSI

Data

Data

z

aM-QAM





max


119873(119901119873max

) = 1 + 119901 + 1199012+ sdot sdot sdot + 119901

119873max

(4)


119894-


119903(119894) as


119878119890phy =

119873

sum

119894=1

119877119894119875119903 (119894) (5)


max)


119878119890link =

119878119890phy

119873(119901119873max)=

1

119873 (119901119873max)

119873

sum

119894=1

119877119894119875119903 (119894) (6)











Encoding

Channelanalyser

DemodulateDecoding

CSI

Data

AchieveYesNo

ARQ

Data

M-QAM

z

a

BLERlink


Data link layer

Physical layer

Higher layers

HARQ

AMC

Higher layers

Data link layer

Physical layer



0at


BLER119873max+1

0le BLERlink (7)

From (7) we have


max+1))

link (8)









4 Performances



119896= 119904119868119896

+ 119895119904119876119896 with 119904

119868119896 119904119876119896


1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1



= 119872(1199061198961 1199061198962 1199061198963 1199061198964)



119904119868119896

= 119872119868(1199061198961 1199061198963) and 119904

119876119896= 119872119876(1199061198962 1199061198964) An example of




119883119896= 119909119868119896

+ 119895119909119876119896

= 119886radic119864119904119904119896+ 119899119896 (9)


energy and 119899119896= 119899119868119896

+ 119895119899119876119896


119868119896and 119899119876119896


2= 11987302


119896119894given119883

119896 and in the


Λ (1199061198961) = log

119875 119906119896119894

= 1 | 119883119896

119875 119906119896119894

= 0 | 119883119896

= log119875 119883119896| 119906119896119894

= 1

119875 119883119896| 119906119896119894

= 0 (10)

Define by 119878(119906119896119894

= 1) = 119902 119904(119902)

= 119872( 119906119896119894

= 1 ) and119878(119906119896119894

= 0) = 119902 119904(119902)

= 119872( 119906119896119894


119896119894= 1 and 119906

119896119894= 0

respectively then

119875 119883119896| 119906119896119894

= 1 = sum

119902isin119878(119906119896119894=1)

119901 (119883119896| 119904(119902)

)

119875 119883119896| 119906119896119894

= 0 = sum

119902isin119878(119906119896119894=0)

119901 (119883119896| 119904(119902)

)

(11)


1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(b)


and 1199061198962

1010

1

3111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

1

3

minus3

minus1

(b)


and 1199061198964



119901 (119883119896| 119904(119902)

) =1

1205871198730

exp(minus

10038161003816100381610038161003816119883119896minus 119886radic119864

119904119904(119902)10038161003816100381610038161003816

2

1198730

) (12)


Λ (1199061198961) = log 119890



221205902



Λ (1199061198962) = log 119890



221205902









max is changeable)


Λ (1199061198963) = log 119890



221205902



Λ (1199061198964) = log 119890



221205902








0 2 4 6 8 10 12 14 16 18SNR (dB)

BER

minus210minus6

10minus5

10minus4

10minus3

10minus2

10minus1

100

MCS1MCS2MCS3

MCS4MCS5MCS6



0= 10minus2


0le 0001





2 4 6 8 10 12 14 16 18SNR (dB)



BLER

10minus5

10minus4

10minus3

10minus2

10minus1

100BLER = 02512

1205741 34 = 23dB


0 2 4 6 8 10 12 14 16 18

AMC

SNR (dB)minus2

0

05

1

15

2

25

3

35

4

45

5

Thro

ughp

ut










0 2 6 8 10 12 14 16 180

05

1

15

2

25

3

35

4

45

5

SNR (dB)

Thro

ughp

ut

4minus2




5 Conclusions




Acknowledgments


References



















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










Encoding

Channelanalyser

DemodulateDecoding

CSI

Data

AchieveYesNo

ARQ

Data

M-QAM

z

a

BLERlink


Data link layer

Physical layer

Higher layers

HARQ

AMC

Higher layers

Data link layer

Physical layer



0at


BLER119873max+1

0le BLERlink (7)

From (7) we have


max+1))

link (8)









4 Performances



119896= 119904119868119896

+ 119895119904119876119896 with 119904

119868119896 119904119876119896


1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1



= 119872(1199061198961 1199061198962 1199061198963 1199061198964)



119904119868119896

= 119872119868(1199061198961 1199061198963) and 119904

119876119896= 119872119876(1199061198962 1199061198964) An example of




119883119896= 119909119868119896

+ 119895119909119876119896

= 119886radic119864119904119904119896+ 119899119896 (9)


energy and 119899119896= 119899119868119896

+ 119895119899119876119896


119868119896and 119899119876119896


2= 11987302


119896119894given119883

119896 and in the


Λ (1199061198961) = log

119875 119906119896119894

= 1 | 119883119896

119875 119906119896119894

= 0 | 119883119896

= log119875 119883119896| 119906119896119894

= 1

119875 119883119896| 119906119896119894

= 0 (10)

Define by 119878(119906119896119894

= 1) = 119902 119904(119902)

= 119872( 119906119896119894

= 1 ) and119878(119906119896119894

= 0) = 119902 119904(119902)

= 119872( 119906119896119894


119896119894= 1 and 119906

119896119894= 0

respectively then

119875 119883119896| 119906119896119894

= 1 = sum

119902isin119878(119906119896119894=1)

119901 (119883119896| 119904(119902)

)

119875 119883119896| 119906119896119894

= 0 = sum

119902isin119878(119906119896119894=0)

119901 (119883119896| 119904(119902)

)

(11)


1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(b)


and 1199061198962

1010

1

3111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

1

3

minus3

minus1

(b)


and 1199061198964



119901 (119883119896| 119904(119902)

) =1

1205871198730

exp(minus

10038161003816100381610038161003816119883119896minus 119886radic119864

119904119904(119902)10038161003816100381610038161003816

2

1198730

) (12)


Λ (1199061198961) = log 119890



221205902



Λ (1199061198962) = log 119890



221205902









max is changeable)


Λ (1199061198963) = log 119890



221205902



Λ (1199061198964) = log 119890



221205902








0 2 4 6 8 10 12 14 16 18SNR (dB)

BER

minus210minus6

10minus5

10minus4

10minus3

10minus2

10minus1

100

MCS1MCS2MCS3

MCS4MCS5MCS6



0= 10minus2


0le 0001





2 4 6 8 10 12 14 16 18SNR (dB)



BLER

10minus5

10minus4

10minus3

10minus2

10minus1

100BLER = 02512

1205741 34 = 23dB


0 2 4 6 8 10 12 14 16 18

AMC

SNR (dB)minus2

0

05

1

15

2

25

3

35

4

45

5

Thro

ughp

ut










0 2 6 8 10 12 14 16 180

05

1

15

2

25

3

35

4

45

5

SNR (dB)

Thro

ughp

ut

4minus2




5 Conclusions




Acknowledgments


References



















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1

3

1 3minus3 minus1

minus3

minus1

(b)


and 1199061198962

1010

1

3111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

minus3

minus1

(a)

1010111001100010

0011 0111 1111 1011

1001

10001100

11010101

01000000

0001

1 3minus3 minus1

1

3

minus3

minus1

(b)


and 1199061198964



119901 (119883119896| 119904(119902)

) =1

1205871198730

exp(minus

10038161003816100381610038161003816119883119896minus 119886radic119864

119904119904(119902)10038161003816100381610038161003816

2

1198730

) (12)


Λ (1199061198961) = log 119890



221205902



Λ (1199061198962) = log 119890



221205902









max is changeable)


Λ (1199061198963) = log 119890



221205902



Λ (1199061198964) = log 119890



221205902








0 2 4 6 8 10 12 14 16 18SNR (dB)

BER

minus210minus6

10minus5

10minus4

10minus3

10minus2

10minus1

100

MCS1MCS2MCS3

MCS4MCS5MCS6



0= 10minus2


0le 0001





2 4 6 8 10 12 14 16 18SNR (dB)



BLER

10minus5

10minus4

10minus3

10minus2

10minus1

100BLER = 02512

1205741 34 = 23dB


0 2 4 6 8 10 12 14 16 18

AMC

SNR (dB)minus2

0

05

1

15

2

25

3

35

4

45

5

Thro

ughp

ut










0 2 6 8 10 12 14 16 180

05

1

15

2

25

3

35

4

45

5

SNR (dB)

Thro

ughp

ut

4minus2




5 Conclusions




Acknowledgments


References



















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation















max is changeable)


Λ (1199061198963) = log 119890



221205902



Λ (1199061198964) = log 119890



221205902








0 2 4 6 8 10 12 14 16 18SNR (dB)

BER

minus210minus6

10minus5

10minus4

10minus3

10minus2

10minus1

100

MCS1MCS2MCS3

MCS4MCS5MCS6



0= 10minus2


0le 0001





2 4 6 8 10 12 14 16 18SNR (dB)



BLER

10minus5

10minus4

10minus3

10minus2

10minus1

100BLER = 02512

1205741 34 = 23dB


0 2 4 6 8 10 12 14 16 18

AMC

SNR (dB)minus2

0

05

1

15

2

25

3

35

4

45

5

Thro

ughp

ut










0 2 6 8 10 12 14 16 180

05

1

15

2

25

3

35

4

45

5

SNR (dB)

Thro

ughp

ut

4minus2




5 Conclusions




Acknowledgments


References



















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










2 4 6 8 10 12 14 16 18SNR (dB)



BLER

10minus5

10minus4

10minus3

10minus2

10minus1

100BLER = 02512

1205741 34 = 23dB


0 2 4 6 8 10 12 14 16 18

AMC

SNR (dB)minus2

0

05

1

15

2

25

3

35

4

45

5

Thro

ughp

ut










0 2 6 8 10 12 14 16 180

05

1

15

2

25

3

35

4

45

5

SNR (dB)

Thro

ughp

ut

4minus2




5 Conclusions




Acknowledgments


References



















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











Acknowledgments


References



















RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









Documents

Research Article A Cross-Layer Design Combining of AMC ...downloads.hindawi.com/journals/ijdsn/2013/145254.pdf · International Journal of Distributed Sensor Networks Encoding Channel