2/7/2012

1

Graduate School in Electronics, Telecommunications and Automation

Global Navigation Satellite Systems (GNSS): present and future

Elena Simona Lohan, Academy Research Fellow., Dr. Techwww.cs.tut.fi/tlt/pos

Your logo

Outline

4 chapters:1. Overview of satellite navigation systems

(current and planned)2. Signals and frequencies in GNSS3. Receiver signal processing4. Dealing with wireless channels impairments

(multipaths, ionospheric delays, etc)

2

Your logo

OVERVIEW OF SATELLITE NAVIGATION SYSTEMS

3 Your logo

Chapter 1 outline

Motivation , applications and classification of GNSSs3 segment architecture of GNSSSystem-by-system overview:• GPS• Galileo• GLONASS• COMPASS

GNSS services and other satellite navigationsystems (SBAS)

4

2/7/2012

2

Your logo

GNSS – some definitions

GNSS stands for Global Navigation Satellite SystemCurrently, there are 4 global navigation systems (on sky or under standardization):• Navstar Global Positioning System, GPS: US, military control,

fully functional, 32 operational satellites• GLONASS: Russia, military control, 24 operational satellites,

fully functional since Dec 2011• Galileo: Europe, civilian control, in standardization phase; first 2

satellites belonging to the final constellation launched in Oct2011; more to come in summer 2012

• COMPASS: China, former BEIDOU system, currently 10 satellites on sky (10th one launched on 1st of Dec 2011) –currently operating on a trial basis (full coverage around China)

5 Your logo

GNSS satellites

6

Sources: http://www.navcen.uscg.gov/ftp/gps/ggeninfo/gps-iif.tif, http://ilrs.gsfc.nasa.gov/satellite_missions/list_of_satellites/g115_general.html, http://www.esa.int/http://www.blog.gpscompared.co.uk/tag/sat-nav

GPS Glonass

Galileo

Compass

Your logo

With satellite-based positioning we can acquire: • global coverage (with some limitations, for example indoors) • good accuracy• Integrity, especially if several GNSS systems are used together.

Inexpensive satellite navigation receivers:

Mini-GPS handheld receiver + location finder at about 40 EUR; a Garmin NUVI215 GPS receiver with UK and Ireland maps at about 120 EURAccording to the available maps inside the GPS receiver, price can reach few hundred EUR

Motivation for GNSS

7 Your logo

transportation (air, land, maritime), surveying and mapping, agriculture, telecommunications, natural resources exploration, commercial, etc.

The current global market of applications and services of positioning systems is estimated to about few tens billion EUR (according to the EU statistics) and it is expected to grow about 7 times over next 5 years.

Applications for satellite-based positioning

8

2/7/2012

3

Your logo

Classification of satellite navigation systems

Global coverage (GNSS):• GPS (US)• GALILEO (Europe)• GLONASS (Russia)• COMPASS (China)

Local/regional coverage (RNSS)• QZSS (Japan)• IRNSS (India)

GPS/GNSS augmentation : they offer support to GPS/GNSS; do not offer stand-alone navigationsolutions

9 Your logo

- GNSS systems are quite complex, involving many different components. The satellites are the most visible part, but they require a heavy ground infrastructure in order to deliver the right signals with the right parameters to the users.

- All GNSS systems are based on the same architecture (3-segment architecture):

- Space segment: satellites- Ground segment: monitoring, controlling and

uploading stations- User segment: user community/GNSS receivers

GNSS system architecture (I)

10

Your logo

- The number of satellites and monitor stations differ according to the GNSS system (GPS, Glonass, Galileo, ...)- The geographical placement and number of ground stations is important, especially with respect to the integrity of the positioning solution. (Integrity is the characteristic that allows a user to evaluate the confidence he can attribute to the positioning the receiver is providing)

GNSS system architecture (II)

11

Space segment

User segmentGround segment

downlinkUplink/downlink

Your logo

Tasks of space segment

The space segment is formed by the satellites, also abbreviated by SV (Satellite Value). The functions of a satellite are:

It receives and stores data from the ground control segment.It maintains a very precise time. In order to achieve such a goal, each satellite usually carries several atomic clocks of two different technologies (e.g., cesium and rubidium), depending on the generation of the satellite.It transmits data to users through the use of several frequenciesIt controls both its altitude and position.It may enable a wireless link between satellites

12

2/7/2012

4

Your logo

Where are the satellites situated?

Typically in MEO (Medium Earth Orbit) orbits

13

LEO : Low Earth OrbitMEO: Medium Earth OrbitGEO: Geo-synchronous Earth Orbit,(orbital period that synchronizes with the Earth’s speed of rotation)

IGSO = Inclined Geo-SynchronousOrbit – belongs to GEO type (the satellite is placed in orbit directly above the Earth’s equator at a precise height, so that it maintains the same position relative to the Earth’s surface)

Your logo

Where are the satellites situated?

14

Type LEO MEO GEO

Height (above Earth) Up to 1600 km From about 6000 km

to about 32000 kmAt about 35874 Km

Time in LOS 15 min 2-4 hrs 24 hrs

Cycle around Earth Up to 18 times/day Around twice/day Once/day

Advantages

Lower launch costs Very short round trip delays Small path loss

Moderate launch cost Small roundtrip delays

Covers 42.2% of the earth's surface Constant view No problems due to Doppler

Disadv. Very short life 1-3 month

Larger delays Greater path loss

Very large round trip delays Expensive rx due to weak signal

Your logo

Tasks of ground segment

15

The main functions of the ground segment are to:Monitor the satellites; activate spare satellites (if available) to maintain system availability; check the SV healthEstimate the on-board clock state and define the corresponding parameters to be broadcast (with reference to the constellation’s master time)Define the orbits of each satellite in order to predict the ephemeris data, together with the almanac;Determine the altitude and location orders to be sent to the satellites to correct their orbits.

Ephemeris = precise orbit and clock corrections for the satellites. Each satellite broadcast only its ephemeris data. In GPS, ephemeris is broadcast every 30 s.Almanac= coarse orbital parameters of the satellites (valid for up to several months)

Your logo

Tasks of user segment

The main functions of a GPS receiver are:Receive the data from the satellites belonging to one or several constellations (e.g. GPS; Galileo) on one or several frequencies. If several constellation => multi-system receivers. If several frequencies => multi-frequency receivers (dual-frequency GPS-GLONASS receivers are rather common nowadays)Acquire the signal from each satellite on sky (acquisition= identification of satellite code and coarse estimation of time delays and Doppler shifts)Track the signal received from the satellites on sky (tracking =fine estimation of time delay and Doppler shifts);Estimate the PVT solution (position, velocity time estimation).

16

2/7/2012

5

Your logo

User segment (cont.)

Block diagram of a GNSS receiver

17

ADC= Analog-to-Digital converterASIC = Application Specific Integrated Circuit (Note: alternative software GNSS receivers are also possible)PVT = Position Velocity Time computation

Your logo

GNSS frequency bands

GNSS occupy the so-called L-band (wavelengths between15 and 30 cm, frequencies between 1 and 2 GHz)

18

7.2.20

fc

Your logo

Next slides:

Examples of architectures (focus on space & ground segments) for:• GPS• GALILEO• GLONASS• COMPASS

19 Your logo

Controlled by US Department of Defense (DoD) Initially for military use only, later on/nowadays also for civilian applications

Its 3 segments are as follows:Satellite constellation: currently 32 satellites; they provide the ranging signals and navigation data messages to the user equipment. Originally, there were 24 satellites.Ground control network: 1 Master Control Station (MCS), 3 uploading stations, and 11 monitor (surveillance) stations; this segment tracks and maintains the satellite constellation by monitoring satellite health and signal integrity, and by maintaining satellite orbit configuration.User equipment: it receives signals from the satellite constellations and computes user Position, Velocity and Time (PVT).

(Navstar) GPS - Overview

20

2/7/2012

6

Your logo

Space segment: GPS constellation

Currently, there are 32 GPS satellites on the sky.(http://tycho.usno.navy.mil/gpscurr.html)

Originally -> Block II GPS satellites only (24):- In 6 nearly circular orbital planes.-Min 4 satellites/plane.- Orbital plane inclination of 55 degrees.- altitudes of 20200km above the earth.- Period: 11 h 58 min

21

Source: http://mgitecetech.files.wordpress.com/2010/12/figure-1.jpg

Period: 14 hr 22 min

Your logo

1 Master Control Station (MCS): located at Schriever Air Force Base, Colorado; here the data from the monitor stations are processed 24 h a day in real time

3 uploading stations: located at Ascension, Diego Garcia and Kwajalein; they operate in S-band (2 - 4 GHz).

11 monitor (surveillance) stations: every satellite can be seen from at least two monitor stations; the monitor stations track all satellites in their range and collect data of the satellite signals. The raw data are then sent to the mater control station where the data are processed.

GPS ground segment (I)

22

Your logo

GPS ground segment (II)

The monitor stations measure signals from the Satellite Values which are incorporated into orbital models for each satellites: precise orbital data (ephemeris) and SV clock corrections.At the Schriever Air Force base, Colorado, the back-up “master clock” of the United States Naval Observatory is located, which has a stability of less than 1 s in 20 million years.

23 Your logo

GPS Control Segment – after 2005Source:http://www.kowoma.de/en/gps/control_segment.htm, taken fromEarthmap:NASA

24

2/7/2012

7

Your logo

The Galileo Programme (so named to honor the great European scientist Galileo Galilei) is a joint initiative of the European Commission (EC) and the European Space Agency (ESA) to provide Europe with its own independent global civilian controlled satellite navigation system.It is an autonomous system, interoperable with GPS and globally available.It is based on the same technology as GPS (i.e., DS-CDMA) and provides a similar - and possibly higher - degree of precision, thanks to the structure of the constellation of satellites and the ground-based control and management systems planned.

Galileo system: overview

25 Your logo

1998: EU decides to develop its own satellite navigation system ”Galileo” 2000: Definition phase of Galileo starts (it was completed in 2003).Mar 2002: statement on GPS-Galileo cooperation; establishment of Galileo Joint Undertaking (GJU) for management and concession of Galileo.Nov 2003: Joint statement between European Commission and US regarding GPS-Galileo1June 2004: EU and US signed GPS-Galileo agreement:

Common civil signal: BOC(1,1) modulation was selected as the baseline for both Galileo and GPS future OS signals

Galileo history (I)

26

Your logo

July 2004: An official European Union Regulatory Authority, the European GNSS Supervisory Authority (GSA) was established. GSA replaced GJU completely in 2006.2005: first Galileo test satellite launched on orbit (GIOVE-A)May 2006: First Galileo standardization documents (for OS signal only) were made available2007: MBOC modulation adopted for common GPS-Galileo signal for civilian use 2008: second Galileo test satellite (Giove-B) launched on orbit2011: first two IOV Galileo satellites launched (in Oct). IOV= In-Orbit Validation.

Galileo history (II)

27

Your logo

Galileo Space Segment

Galileo will be based on a constellation of 30 satellites (27 + 3 reserve) and ground stations

It uses the same principles as Navstar GPS for position location

Constellation of 30 satellites at 23222 Km altitude (17 orbits in 10 days). Orbital inclination 56 degrees. 10 satellites per plane, 3 planes.

– Only 9 satellites per plane active.

– One satellite per plane is in-orbit spare.

Period: 14 hr 22 min

2/7/2012

8

Your logo

Galileo space segment – spacecraft

• Dimensions: 2.7 x 1.2 x 1.1 m3

• Overall Spacecraft: 680 Kg / 1.6 kW class

• Launcher Options:

Ariane, Proton, Soyuz• Two on-board atomic clocks:

The Rubidium Atomic Frequency standard Clock: frequency around 6 GHz, 3.5 Kg mass, 30 W powerThe Passive Hydrogen Maser Clock: frequency around 1.4 GHz, 15 kg mass, 70 W power (the most stable clock ever to fly in space)

29

Sourse: ESA pages

Your logo

Galileo - Trivia

First 2 operational satellites launched on 21st Oct2011 from Europe's Space Port in Kourou, FrenchGuianaPress release: http://ec.europa.eu/enterprise/

policies/satnav/galileo/satellite-launches/index_en.htm

Galileo satellites will be named with namesof EU children (one child per country, winner drawn according to a drawingcompetition)

30

Your logo

2 redundant Galileo Control Centers (GCS) (Europe): control of the satellites and navigation mission management; in Oberpfaffenhofen(Germany, near Munich) and Fucino (Italy, near Rome)10 mission up-link stations (ULS) (worldwide) in C-band (5.925–6.425 GHz)29-30 Galileo Sensor/surveillance stations (GSS) (worldwide) forming the Galileo Mission control System (GMS) together with the up-link stations: they will follow the satellite signals in order to provide the definition of the data to be uploaded5 Tracking, Telemetry and Control (TT&C) stations (worldwide) in S-band (2 - 4 GHz): these TT&C stations form the Galileo Ground Control System (GCS): monitoring and control of the satellites.

The number of stations is not finalized; modifications are still possible.

Galileo Ground Segment

31

Your logo

Galileo architecture -details

32

Galileo Control Center (GCC) – 2 centers

Galileo space segment

5 TTC 10 ULS 29-30 GCS

S-band C-band L-band

2/7/2012

9

Your logo

Dual-use (civil/military) system provided by RussiaCivil signals provided to all, free of user fees; It started in 1983; now it’s updatedGLONASS modernization directive, as of January 2006:

Constellation of 18 SV by end of 2007Full operationalcapability by end of 2009-> done in Dec 2011Comparableperformance with GPS and Galileo by 2010

GLONASS system: overview

Glonass currently usesFDMA (different from allthe other CDMA-basedsatellite navigationsystems)

33

Your logo

24 active satellitesAltitude - 19100 kmInclination - 64.8 degreesPeriod - 11 h 15 min.8 satellites per plane; 3 orbital planes

GLONASS space segment

34

Your logo

1 system control center (SCC) located in Krasnoznamensk, Moscow region5 TT&C stations, all located in Rusia30-40 surveillance

stations, distributedaround the world

GLONASS ground segment

35

Source: http://www.glonass.it/images/glonass_network.jpg

Your logo

System under development by China; starting point was BEIDOU, later renamed as Compass 30 MEO satellites + 5 GEO satellites (MEO= Medium Earth Orbit; GEO= Geostationary Earth Orbit)Claimed accuracy: positioning 10 meters, velocity - 0.2 meter per second timing accuracy - 50 nanoseconds.Could institute user chargesCurrently, they have already launched 10 satellites (GEO/MEO and IGSO); more are planned to be launched before end of 2012

COMPASS system overview

36

2/7/2012

10

Your logo

COMPASS space segment

Maximum 30 MEO satellites and 5 GEO satellites; Orbit height: 21500 km (MEO) and about 36000 km (GEO)Orbit inclination: 55 degrees

37 Your logo

COMPASS ground segment

1 master control station2 upload stations30 monitor stations

First and only Interface Control Document (ICD) for Compass published on 27th of Dec 2011 (13 pagesonly); however important information still missing fromthere (source: http://www.insidegnss.com/node/2879)

38

Your logo

System comparison

39

GPS GALILEO GLONASS COMPASS

First launch 1978 2011 1982 2007

Full OperationalCapability(FOC)

1995 2017 (?) 2011 2015(?)

Number of SV 32 30 24 35 (?)

Orbital planes 6 3 3 3

Inclination(degrees)

55 56 64.8 55

Multiple access CDMA CDMA FDMA/CDMA CDMA

Orbital period 11h58m 14h05m 11h15m 12h50m

Total of121 SV possiblein the future

Your logo

Type of services offered by GNSS

Open services: free for everyone, typically no quarantee of serviceCommercial and/or authorized services: improved performance with limited accessibility

40

2/7/2012

11

Your logo

GPS services

2 types of services specified in the US Federal RadionavigationPlan (1999):

Precise Positioning Systems (PPS) andStandard Positioning Service (SPS)

PPS Predictable Accuracy is: 22 meters horizontal accuracy, 27 meters vertical accuracy, and 200 nanosecond time (Universal Time Clock/UTC) accuracy.SPS Predictable Accuracy (for civil users; no military restriction) is: 100 meters horizontal accuracy, 156 meters vertical accuracy, and 340 nanosecond time (Universal Time Clock/UTC) accuracy.

41 Your logo

Open Service (OS): free for everyone; mass market applications, simple positioningSafety of Life (SoL): for professional applications; integrity; authentication of signalCommercial Service (CS): for maritime, aviation and train applications; encrypted; high accuracy; guaranteed servicePublic Regulated Service (PRS): encrypted; government-regulated; integrity; continuous availabilitySupport to Search and Rescue service (SAR):humanitarian purpose; near real time; precise; return link feasible

Galileo services

42

Your logo

GLONASS services

Civil services: open, free for everyone, standardprecision navigation signalsMilitary/Authorized service: high precisionnavigation signalsSearch and rescue service

43 Your logo

Two kinds of global services:

• Open Service free and open to users

Positioning Accuracy: 10 m

Timing Accuracy: 20 ns

Velocity Accuracy: 0.2 m/s

• Authorized Service: ensures highly reliable use even in complex situation.

COMPASS services

44

2/7/2012

12

Your logo

About Regional Satellite Navigation Systems

QZSS: Quasi-Zenith Satellite System (Japan):Civil system for Asia Pacific regionSignals on the same frequency bands as GPS developped in collaboration with USInitial objective of 3 satellites, with possibility of more extensiveconstellation afterwardsAltitude 42164 km, 45 degree orbital inclination

IRNSS: Indian Regional Navigational Satellite System (India):

Constellation of 7 GEO satellites, due for completion in next 1-3 years; no satellite yet on skyGround segment: 1 master control sation + 4 monitor TT&C stations

45 Your logo

About GPS/GNSS augmentation

SBAS=Satellite based augmentation systems:• EGNOS (Europe)• WAAS (US)• CWAAS (Canada)• GAGAN (India)• MSAS (Japan)

Not used in a stand-alone modeOffer additional support to GPS

46

Your logo

Discussion: EGNOS (I)

Project was launched in 1998.Target accuracy is below 5 meters.It consists of 3 Geostationary satellites (Inmarsat AOR-E with PRN 120, Inmarsat IOR-W with PRN126 and Artemis with PRN 124) and a network of ground stations:

34 Ranging and Integrity Monitoring Stations (RIMS) deployed in 19 countries; 4 MCC (Mission Control Centres) deployed in Spain, Italy, UK and Germany; 6 Navigation Land Earth Stations (NLESs) deployed in Italy, France, Spain and the UK some support facilities in Spain and France.

47 Your logo

Discussion: EGNOS (II)

What are EGNOS satellites transmitting?• GPS-like signals in L1 band (carrier 1575.42 MHz)• 5 times faster data rates than GPS data rate => faster corrections than

with GPS• Forward error correction code

EGNOS is providing differential corrections (for pseudorange, orbit and clock computations) and ionospheric correctionsCannot be used alone for position computation(need GPS satellites)

48

2/7/2012

13

Your logo

SIGNALS AND FREQUENCIES IN GNSS

49 Your logo

Chapter 2 outline

Multiple access techniques in GNSSSpreading codesused in GNSSModulations used in GNSS (main emphasis on BOC family)Spectral charcateristics of modulated waveformsSignals and spectra used in the 4 GNSS systems

50

Your logo

Signals’ tasks in GNSS

Can make the separationbetween satellites (multipleaccess)Enable the receiver to compute the pseudorange. Pseudoranges are estimatesof the true range (=distancefrom satellite to receiver). Pseudoranges differ from the true range due to variousrandom errors.Also carry the informationneeded by the receiver to compute the final position

51

Rx: ’reads’ satellitesclock and location;computes Position, Velocity, Time (PVT)

Your logo

Recall: multiple access definition

Sharing (by different transmitters/ satellites) of the same transmission mediumTransmission medium = wireless channelGNSS use multiple signals:• Similar signal designs from different satellites

(e.g., C/A GPS signal sent from all GPS satellites)

• Different signals providing different services, from the same satellite (e.g., Galileo E1 and E5a signals)

52

2/7/2012

14

Your logo

Multiple access techniques in GNSS

FDMA = FrequencyDivision Multiple Access (GLONASS)

53

SV= Satellite Value

CDMA = Code Division Multiple Access (Galileo,GPS, Compass, some of the future GLONASS signals)

Your logo

How is the FDMA realized?

54

FDMA

Signal fromone satellite

f f

power power

BW1 BW2

satellite1 satellite2……

Local oscillator with variable center frequencySimple transmitter implementationSeparation between signals at the receiver is done via filtering => need of tight filtering to reduce adjacent channel interference

Your logo

How is the CDMA realized?

Implemented via direct sequence spread spectrum (DS-SS) technology = spreading the narrowband signal into a wideband signal using a special (spreading) code

Spreading factor = BW2/BW1

Additional benefit of spread spectrum: inherent robustness to narrowband interference

55

spreading

spread signal

signal

f f

power power

BW1 BW2

Your logo

Transmitter (satellite) simplified model

Mod. = modulation;MUX= multiplexing,LO =LocalOscillator,PA= Power Amplifier;*Navigationdata can beabsent (e.g, pilotchannels)

56

2/7/2012

15

Your logo

Navigation data (I)

Navigation data message contains essential information for positioning computation:• Time• Information regarding satellite positions (almanac,

ephemeris)• Clock corrections• Ionospheric model• Time offsets among different frequencies and among

different systems, and with respect to Universal Time (UTC)

57

Spreadingmod.

Channel mod.

Navigationdata

Your logo

Data message channel coding

Channel coding is used to protect (to a certain extent) the navigation data againstthe wireless channel errorsForward Error Correction (FEC) is a channelcoding method based on insertingredundancy in the transmitted data. Thisredundancy allows the receiver to correctsome of the transmission errorsError checking bits (e.g., parity bits) can alsobe used to detect errors

58

Your logo

Pilot channels

Some GNSS channels are transmitting PRN codes without any navigation data (withoutdata modulation)These are the so-called ’pilot channels’ => useful in the acquisition and trackingprocesses• Absence of data bit transitions allows for longer integration times

=> better performance in noisy environments• Code sequences are known at the receiver (except the delay),

but navigation data (when present) is not known. Without data modulation => less parameters to be estimated

59 Your logo

Spreading modulation

A pseudorandom noise (PRN) code at a rate fc (chip rate) is multiplied with a data sequence, at a much lower rate, fb (bitrate or code epoch rate)

60

X

‘User’ data, bit rate fb

chipSequence(’PRN code’)at fc

spreadspectrumsignal, at fc

Example:fc = 1.023 MHzfb = 1 kHzfc /fb = 1023 = ’bandwidthexpansion’ factor or’spreading factor’

Note: navigation data rate is not the same as the bit rate fb. The current terminology comes from the communication world. In navigationworld, fb is called the code epoch rate.

Spreadingmod.

Channel mod.

Navigationdata

2/7/2012

16

Your logo

Example of spread sequence (baseband)

61

data sequence, at fb

chip sequence, at fc

Spread spectrumsignal, at fc

Spreading factor cF

b

fS

f

0 2 4 6 8 10 12-1

0

1NAV data

0 2 4 6 8 10 12-1

0

1PRN code

0 2 4 6 8 10 12-1

0

1Spread signal

Here, SF=5 chips per code epoch and data bit has 3 code epochs duration. Time axis is in code epochs.

Your logo

Some mathematics (I)

Data sequence, bn=data bits, T= bit interval =1/fb

62

(t) Dirac pulse

t

0

( ) 1t dt

( ) , 0( ) 0, 0t tt t

( ) ( ) ( ), ( )t p t p t p t

( ) ( )nn

d t b t nT

=Convolution operator

1nb

nb

1nb 2nb

Your logo

Some mathematics (II)

Chip sequence (spreading code), as sequence of +1,-1 rectangular pulses:

63

, ,1 1

( ) ( ) ( ) ( )F F

c c

S S

n k n T c T k n ck k

c t c p t kT p t c t kT

cT = chip interval =1/fc c bF

b c

f TSf T

,k nc = chip value (+1 or -1) corresponding to the k-thchip of the n-th code epoch

code epoch n code epoch n+1

chip k

( )cTp t = rectangular pulse of support

cT

cT

t

1

0

( )cTp t

cT

cT

Your logo

Some mathematics (III)

64

,1

( ) ( ) ( )F

c

S

T n k n cn k

s t p t b c t n T k T

The spread signal, at the output of the spreading operation can be modeled as:

One important aspect when choosing a spreading code c(t) relates to its auto-correlation (correlation with itself) and cross-correlation (correlation with other signals) properties

2/7/2012

17

Your logo

How to choose the spreading codes? (I)1. Good code auto-correlation properties: the code sequence for a

certain code epoch, without the rectangular pulse shaping part , is

Its auto-correlation equals to

A ’good’ auto-correlation function is the one which ressembles to a Diracpulse, that is

65

,1

( )FS

k n ck

c t kT

, ,1 1

( )F FS S

k n p n c ck p

c c t kT pT

, ,1 1

2,

1

0,

1

F F

F

S S

k n p nk p

S

k nk

c c if k p

c

Meaning:-Correlations with delayedversions of the signal aresmall-Correlation with an alignedversion of the signal is high- Such properties help in the acquisition/tracking process

Your logo

How to choose the spreading codes? (II)2. Good code cross-correlation propertiesThe cross-correlation between 2 codes (without the pulse shaping part)

equals to

, are the chip sequences for 2 codes (1) and (2)A ’good’ cross-correlation function is the one having all-zero values (or

values much lower than the maximum of the code auto-correlation)

66

(1) (2), ,

1 1( )

F FS S

k n p n c ck p

c c t kT pT

(1),k nc (2)

,p nc

(1) (2), ,

1 10

F FS S

k n p nk p

c cMeaning:

- Correlations with the codes of other satellites are small => help in detecting correctlywhich satellites are on the sky

Your logo

Other considerations about spreading codes

Number of different codes available -> to be enough for allsatellites and all signalsLength: longer codes have typically better correlationproperties, but receiver processing is easier with shortercodes. Rate of the spreading code (chip rate): determines the accuracy with which the position determination can be made. For example, with a chip rate of 1.023 MHz, each chipcorresponds to a light travel time of about 300 m.Good balance between number of ’zeros’ and ’ones’ (or +1/-1) in a code

Trivia: The code length of 1023 chips was chosen in GPS at the recommendation of J. Spilker

67 Your logo

Spreading codes used in GNSS

1. Maximal length or m-sequences (e.g., GLONASS)

2. Gold codes (e.g., GPS current signals)3. Memory codes (e.g., Galileo Open Service)4. Weil codes (e.g., GPS future L1C civil

signal)

68

2/7/2012

18

Your logo

M-sequences

Generation based on an N-tap linear feedback shift register (accordingto on an underlying generator polynomial)

69

Delay 1 Delay 2 Delay n………….

+ +

output• N= 2n-1 length sequences• Only 2 possible values of the normalized autocorrelation function, namely 1 and -1/N => good autocorrelation properties

Your logo

Gold codes

Combination (XOR) of two m-sequences (not any m-sequenceallowed; only a sub-set of them)N= 2n-1 code length (in GPS, n even, n=10)

70

4 possible valuesof the normalizedautocorrelationfunction:

(e.g., 1, -1/1023, -65/1023, and 63/1023

2 22 21 2 1 2 11, , ,

n n

N N N

Your logo

Memory codes (’random’ codes)M-sequences and Gold codes do not have very good cross-correlation properties, especially when navigation data bittransitions are also taken into accountTherefore, an optimization-based code selection was proposedfor Galileo; optimization criteria:• Good balance of ’zeros’ and ’ones’• Small auto-correlation and cross-correlation properties with both

even and odd navigation data bit patterns (e.g., 1 followed by a 1, and 1 followed by a -1)

The complete sequences are saved in the memory (of tx & rx) and read from there=> larger memory tables than for m-sequences and Gold codes

71 Your logo

Weil codes

Based on Legendre sequences and XOR additionsCan have any prime length (not limited to 2n-1 length)Motivated by the need of finding codes of length N=10230 chips with good auto- and cross-correlation properties (e.g., for GPS L1C signal, Weil codes of prime length 10223 are used and padded with 7 chips in order to reach the 10230 chip length)Reported to have better cross-correlation properties than Gold codesNot as flexible as memory codes, e.g., not possible to drivethem to Autocorrelation Sizelobe Zero (ASZ) property (ASZ= zero auto-correlation value when the code is correlated with a delayed version of itself of 1 chip)

72

2/7/2012

19

Your logo

Channel modulation

Modulation task: some characteristics (amplitude, phase or frequency) of a carrier wave are varied in accordance with an information bearing signal

73

digitalmodulation

digitaldata analog

modulation

radiocarrier

basebandsignal

1-111-11-1-11

Channel Modulator

Spreadingmod.

Channel mod.

Navigationdata

Your logo

Binary Phase Shift Keying (BPSK)If p(t) denotes the basic pulse used in the construction of the binary data stream, then the BPSK-modulated PRN sequencecan be written as:

No pulse shaping currently used in GNSS => (rectangular pulse)Therefore, the modulated signal can be seen as the convolution between a code part c(t) (including navigationdata) and a modulation pulse

74

,1

1 for binary symbol "1"( ) ( ),

1 for binary symbol "0"

FS

n k n c nn k

s t b c p t nT kT b

( ) ( )cTp t p t

,1

( ) ( ) ( ) ( ) ( )F

c

S

T n k n c BPSKn k

s t p t b c t nT kT s t c t

( ) ( )cB P SK Ts t p t

Your logo

BPSK (II)

The expression given in the previous slide is the expression for the baseband signal (in continuous-time form), which will bethe basis of our models in what follows

The analog passband signal (at carrier frequency) becomes:

Notation: BPSK(n) means a chip rate of n*1.023 MHz

75

( ) ( ) cos(2 )c carrierm t A s t f t

Your logo

(Power) Spectral Density (PSD)

PSD is the Fourier transform of the auto-correlation function of a signalSince a BPSK-modulated signal has both discrete-time and continuous components, the PSD has two parts :

= the Fourier transform of the pulse waveform

= the PSD of the transmitted discrete-time symbol train

For infinite-length ideal codes (i.e., Dirac-shaped auto-correlation), we have = > the modulated-signal PSD is given by the square of the pulse shape frequency response:

76

2 21( ) ( ) ( )c

c

j fTs T b

c

G f P f G eT

2( ) ( )c c

j ftT TP f p t e dt

2( )cj fTbG e

2( ) 1cj fTbG e

21( ) ( )

cs Tc

G f P fT

2/7/2012

20

Your logo

Notes on terminology

Strictly speaking, we deal with finite-time signals => finiteenergy signalsThat is why, sometimes the preferred term is ’energyspectral density’ (or simply ’spectral density’) instead of PSD

77 Your logo

Equivalent baseband representation of modulated signals

78

Baseband spectrum

Spectrum at carrierfrequency (passband)

0f

0 ffcarrier-fcarrier

( )sG f

( )s carrierG f f* ( )s carrierG f f

Your logo

PSD of BPSK signals (I)

Sinc-shaped PSD, maximum energy at 0 frequency (baseband representation)

79

cTt

1

0

( )cTp t

2 21 sin( )( ) ( ) sin , sin ( )cs T c c

c

xG f P f T c fT c xT x

f

( )sG f

cT

1

cT01

cT2

cT2

cT

=

Your logo

PSD of BPSK signals (II)

80

Power within main frequency lobe (2fc bandwidth) is about 90% of the whole signal power

2/7/2012

21

Your logo

Binary Offset Carrier (BOC) modulation

81

Sine BOC with NB=2fc/fsc•Square sub-carrier modulation, where the PRN code (of chip rate fc) is multiplied by a rectangular sub-carrier of frequency fsc, which splits the spectrum of the signal. •Typical notations:

•BOC(fsc; fc) or •BOC(m; n), m = fsc/1.023 MHz, n = fc/1.023 MHz.

•BOC modulation order

•Sine-BOC time waveform given by:

2 scB

c

fNf

( ) sin BSinBOC

c

N ts t signT

Your logo

Notes on Sine BOC

From the point of view of baseband characterization, BOC-modulated signal is fully defined by two parameters, namely NB and chip frequency fc

From the point of view of passband signal, also the carrierfrequency should be specified.The sine-BOC modulated signal can be written in a similar formas BPSK-modulated signal:

Equivalent representation of sine-BOC waveform [LLR06b] as an alternating sequence of +1 and -1

82

,1

( ) ( ) ( ) ( ) ( )FS

SinBOC n k n c SinBOCn k

s t s t b c t nT kT s t c t

1

0( ) ( ) ( 1) ,

B

B

Ni c

SinBOC T B Bi B

Ts t p t t iT TN

Your logo

Cosine BOC

83

Cosine BOC with NB=2fc/fsc

( ) cos BCosBOC

c

N ts t signT

• Cosine-BOC time waveform given by:

•Cosine BOC can be modeled as a double sine BOC modulation: first stagewith modulation order NB, and secondstage with modulation order 2 [LLR06b]

11

0 02

( ) ( ) ( 1) ,2

B

B

Ni k B

CosBOC T Bk i

cB

B

Ts t p t t iT k

TTN

Your logo

Modeling BPSK, sine BOC and cosine BOC

Generic modulated-signal model

Generic modulation waveform:

Modulation factors:BPSK case

Sine BOC(m,n) case

Cosine BOC(m,n) case

84

mod , mod1

( ) ( ) ( ) ( ) ( ),mod ,FS

n k n cn k

s t s t b c t nT kT s t c t BPSK SinBOC orCosBOC

2 1

1 2 1 221 1 2

1 1

mod0 0

( ) ( ) ( 1) , ,B B

B

N Ni k c c

T B B B Bk i B B B

T Ts t p t t iT kT T TN N N

1 21, 1B BN N

1 22 , 1B B

mN Nn

1 22 , 2B B

mN Nn

2/7/2012

22

Your logo

PSD of Sine BOC-modulated signalsObtained by squared Fourier transform ofRandom part of the modulated signal ignored (e.g., assumptionthat we have ideal codes).

The above expressions are normalized to unit energy overinfinite bandwidth

85

mod ( )s t

2

sin sin1

( )cos

cc

Bs

c c

B

Tf fTN

G fT Tf f

N

For SinBOC(m,n) with even NB

2BmNn2

sin cos1

( )cos

cc

Bs

c c

B

Tf fTN

G fT Tf f

N

For SinBOC(m,n) with odd NB

Your logo

PSD of Cosine BOC-modulated signals

86

2

sin sin sin21

( )cos cos

2

c cc

B Bs

c c c

B B

T Tf f fTN N

G fT T Tf f f

N N

For CosBOC(m,n) with even NB

2BmNn

For CosBOC(m,n) with odd NB

2

sin sin cos21

( )cos cos

2

c cc

B Bs

c c c

B B

T Tf f fTN N

G fT T Tf f f

N N

Your logo

Examples of BOC spectra (I)

87 Your logo

Examples of BOC spectra (II)

88

2/7/2012

23

Your logo

Effect of non-idealities of the code

89 Your logo

Power containment

Power within a certain (double-sided) bandwidth (e.g., filled area below)

90

Your logo

Power containment

91

Examples: about 85% within 4 MHz double-sidedbandwidth for sineBOC(1,1)About 80% within 20 MHz for cosineBOC(15,2.5)

Your logo

Ideal ACF shaped by the modulation waveform

Auto-correlation function (ACF) is given by

Examples (over infinite bandwidth)• BPSK

• Sine BOC

92

mod mod( ) ( ) ( ),mod ,R t s t s t BPSK SinBOC or CosBOC

( ) ( ) ( ) ( )B B BT T TR t p t p t t

1 1

0 0( ) ( 1) ,

B B

B

N Ni k c

T B B Bk i B

TR t t iT kT TN

BT BT

( )BT t

t

0

2/7/2012

24

Your logo

Examples of ideal ACFs

93 Your logo

ACF role in receiver processing(I) Signal from the satellite is acquiredbased on the correlation betweenthe incoming signaland a reference PRN code, with varioustentative delays and tentative Dopplershifts

94

Your logo

ACF role in receiver processing (II) Correlationshape (given bythe ACF of the employedmodulation) influences alsothe trackingproperties

95 Your logo

Bandwidth impact on ACF (I)

Bandwidth-limiting filter impulse response p(t) should be taken into account:

Bandwidth limitation occurs due to the front-endfiltering stagesSampling frequency is determined by the front-end bandwidth (Nyquist condition): higherbandwidth => higher sampling rate => highercomplexity

96

( ) ( )R t p t

2/7/2012

25

Your logo

Bandwidth impact on ACF (II)

Correlationlosses’Rounding’ of peaksPossiblewidening of the correlation main lobeBW=double-sidedbandwidth

97 Your logo

Multiplexed BOC (MBOC)

98

(1,1) (6,1)10 1( ) ( ) ( )11 11MBOC SinBOC SinBOCG f G f G f

Proposed (in 2004) for Galileo E1 signals and for modernized GPS L1C signalPower spectral density of the MBOC signal has to satisfy:

Several implementations possible

Your logo

Time-Multiplexed BOC (TMBOC)TMBOC: time-multiplexed sine BOC(1,1) symbols with sine BOC(6,1) symbols; 2-level waveform. Example: 10 PRN chips; chips 2 and 6 are SinBOC(6,1)-modulated

99 Your logo

Composite BOC (CBOC)

Weighted combination of SinBOC(1,1) and SinBOC(6,1) code symbolsCBOC(+)CBOC(-)CBOC(+/-): odd chips use CBOC(+) and even chips use CBOC(-) (or viceversa)

100

( ) 1 (1,1) 2 (6,1)( ) ( ) ( )CBOC SinBOC SinBOCs t w s t w s t

( ) 1 (1,1) 2 (6,1)( ) ( ) ( )CBOC SinBOC SinBOCs t w s t w s t

2/7/2012

26

Your logo

CBOC

Example: typical values

Note that

101

1 210 1,11 11

w w

2 21 2 1w w

Your logo

Ideal ACF with MBOC

102

Your logo

PSD of MBOC

103 Your logo

Notes on MBOC PSD

Typically, the MBOC PSD is achievedwith a certain combination of data and pilot channelsFor example, in Galileo E1 Open Signal, data channel uses CBOC(+) modulationand pilot channel uses CBOC(-) modulation

104

2/7/2012

27

Your logo

Advantages of MBOC over SinBOC(1,1)

Reported bettertrackingcapability (ifsufficiently highbandwidth)Reported bettermultipathmitigation[HRV06]

105 Your logo

Complexity issues in MBOC

Higher bandwidth consumption (due to extra SinBOC(6,1) component) => highersampling rates if full processingPossibility to process MBOC signals withSinBOC(1,1) receiver => mass-marketimplementations

106

Your logo

Alternate BOC (AltBOC)

Four signals modulated onto the two phases of orthogonal sub-carriers [ALI08]Sub-carrier waveforms are chosen so as to obtain a constant envelope at the transmitterMotivated by the need to transmit 2 complex signals (data and pilot, each with I and Q components) into 2 frequency bandsin Galileo

107 Your logo

AltBOC briefly

1. Form two Single Side-Band (SSB) carriers from SinBOC and CosBOCmodulations (via their sum and difference)

2. Modulate the 4 signals via the 2 SSB carriers

3. Use a 8-PSK mapping (look-up table) for the resulting symbols, in order to get a constant envelope signal

108

2/7/2012

28

Your logo

Example: AltBOC(15,10)

109

Main lobes separation of 30 MHz (=distance between E5a and E5b center frequencies in Galileo)

Your logo

Spectral properties (I)

Spectral Separation Coefficient (SSC) between 2 signals s1(t) and s2(t) [Bet00,Bet01]: should be as low as possible

Maximum value of the spectrum: shouldbe as small as possible => lessinterference

110

1 2( ) ( )

W

SSC s sB

G f G f df

1( )max

W

MVS sf B

m G f

Your logo

Spectral properties (II)

RMS Bandwidth of the signal: the smaller, the better

Power containment factor (discussed before): the higher, the better demodulation process

111

2 ( )W

BW sB

RMS f G f df

2

2

( )

W

W

B

sB

G f df

Your logo

SSC illustration

SSC between 2 BPSK signals (left) and betweenBPSK and sine BOC(1,1) signal, BT=8 MHz (right)

112

2/7/2012

29

Your logo

Numerical example

Source: Lohan&al[LLR05]

BT=double-sidedbandwidth BW

113 Your logo

Notes on the modulation

Modulation ’shapes’ the spectrum of the transmitted signal => defines the level of spectral interference with other GNSS signalsModulation ’shapes’ the ACF => influences the acquisition and tracking capabilities of a signal

114

Your logo

GPS, Galileo, Compass frequencies

115

Compass1561.098

RNSS=radionavigation satellite service;ARNS=aeronautical radionavigation service (radar, tactical air navigation, trafficcollision avoidance system, etc

Your logo

GLONASS frequencies

116

Source: http://www.oosa.unvienna.org/pdf/icg/2007/icg2/presentations/05.pdf

2/7/2012

30

Your logo

Frequencies and signal separation

Some overlapping between the frequency bandsof the 4 GNSS systemsSignal separation between different systemsdone via:• Different PRN codes (GLONASS also uses a

CDMA spreading code, same for all satellites)• Modulation type (spectral separation between

split-spectrum modulations and modulationswith peak energy at carrier frequency)

Inter-system interference still remaining

117 Your logo

GPS signals and modulations

Civil signals• C/A (Coarse/Acquisition) code on L1 band (since the

beginning; 1980s): BPSK(1) (it means a chip rate 1.023 MHz)

• L2C on L2 band (since 2005): BPSK(1)• L5 on L5 band (since Apr 2010): BPSK(10)• L1C on L1 band (to come): TMBOC

Restricted/military/encrypted signals:• P(Y) (Precise) code on L1 and L2 bands: BPSK(10)• M-code on L1 and L2 bands : SinBOC(10,5)

118

Your logo

Example: GPS L1 Spectra

Basebandrepresentation(offset 1575.42 MHz from 0 center frequency)Signals:

• C/A: BPSK(1)• P(Y):BPSK(10)• M: SinBOC(10,5)• L1C:

TMBOC(6,1,29/33)

119 Your logo

GPS frequency bands

120

Band Edge frequencies[MHz]

Carrier frequency[MHz]

L1 1563-1587 1575.42 = 154 x 10.23L2 1215-1237 1227.60 = 120 x 10.23L5 1164-1191.795 1176.45 = 115 x 10.23

2/7/2012

31

Your logo

GPS code lengths

1023 chips for L1 C/A10230 chip length on L1C2 codes: one with 10230 chip length and another one with 767250 chip length for L2C2 codes of 10230 chip length on L5

121 Your logo

Galileo signals

Open Service (OS) signals (E1, E5a, E5b)Safety of Life Service (SoL) signals (E1, E5b)Commercial Service (CS) signals (E1, E6)Public Regulated Service (PRS) signals (E1 E6)Search and Rescue (SAR) signals: separate frequency band

122

Your logo

Explanations on OS signals

Defined for mass-market applicationsFree of direct user chargeIntended for interoperability with GPSNo integrity information and no serviceguarantee

Note: all details subject to change. These are in accordancewith current Galileo SIS-ICD, published in 2011

123 Your logo

Explanations on SoL signals

Defined for transport applications where lifescan be endangered without the real-time noticeof system degradationHigh-integrity levelGalileo Operating Company or Concessionaire(GOC) will guarantee SoL

124

2/7/2012

32

Your logo

Explanations on CS signals

For applications that need higher performancethan OSAccess to signal limited by encryption; fee-based value-added servicesService quarantees, precise timing services

125 Your logo

Explanations on PRS signals

For government-authorized use (police, customs, coast-guard, …)Protection against spoofing and jammingAccess limited by encryption

126

Your logo

Galileo modulations

CBOC(+) modulation for E1-B (data channel)CBOC(-) modulation for E1-C (pilot channel)AltBOC(15,10) modulation for E5BPSK(5) for E6 (Commercial Service CS)CosBOC(10,5) for E6 (Public regulated Service PRS)

127 Your logo

Example: Galileo E1 spectra

128

Basebandrepresentation(offset 1575.42 MHz from 0 center frequency)Signals:• E1-C/E1-B

(OS/SoL): MBOC• PRS:

CosBOC(15,2.5)

2/7/2012

33

Your logo

Galileo frequency bands

129

Band Edge frequencies[MHz]

Carrier frequency[MHz]

E1 1587-1591 1575.42 = 154 x 10.23E6 1260-1278.75 1278.75= 125 x 10.23E5 1164-1214 1191.795 = 116.5 x

10.23E5a 1164-1191.795 1176.45 = 115 x 10.23E5b 1191.795-1214 1207.14 = 118 x 10.23

Your logo

Galileo code lengths

4092 chip length for E1 signals (Note: the spreading factor is still 1023)10230 chip length on E55115 chip length on E6

130

Your logo

GLONASS signals

Standard accuracy signals:• L1 open FDMA• L2 open FDMAHigh-accuracy signals:• L1 authorized FDMA (encrypted codes)• L2 authorized FDMAL3 signal (details not yet defined)

131 Your logo

GLONASS modulations

BPSK(0.5) on L1 and L2 (i.e., 0.511 MHz chip rate. Note: spreadingmodulation is still used in Glonass, although the multiple access schemeis mainly FDMA)Not yet public for L3 and L5 (probably, higher chipping rates)

132

2/7/2012

34

Your logo

GLONASS frequency bands

133

Band Edge frequencies [MHz] Carrier frequencies [MHz]

L1 1591-1610 1602 + 0.5625n , n=-7,-6, …6L2 1237-1260 1246+0.4375n, n=-7,-6,…,6L3 (TBD) 1191.795-1214 (same as

E5b)1208 *

L5 (TBD) 1164-1191.795 1176.45 *= 115 x 10.23

TBD =To Be Defined* On L3 and L5 we will have GLONASS CDMA signals

Your logo

Example: GLONASS spectrum, L1

134

Basebandrepresentation(offset 1602MHz from 0 center frequency)Signal:• BPSK(0.5) and

FDMA

Your logo

Compass/BEIDOU signals

B1 Open signalB1 Authorized signalB2 Open signalB2 Authorized signalB3 Authorized signal

Note: scarce public information regarding Compass. May be subjected to changes.

135 Your logo

Compass modulations

MBOC for B1 open signalSine BOC(14,2) for B1 authorized signalAltBOC(15,10) for B2 open signalCosine BOC(15,2.5) and QPSK(10) for B3 authorized signal

136

2/7/2012

35

Your logo

COMPASS frequency bands

137

Band Edge frequencies[MHz]

Carrier frequency[MHz]

B1 1559.052-1591.788 1575.42 = 154 x 10.23B2 1166.22-1217.37 1191.795 =116.5 x 10.23B3 1250.618-1286.423 1268.52 = 124 x 10.23

Your logo

Example: Compass spectra in B1

138

Basebandrepresentation(offset 1574.42MHz from 0 center frequency)Signal:• B1 open: MBOC• B1 authorized

SinBOC(14,2)

Your logo

GNSS CDMA signals together

139

GPS

Galileo

Compass

Glonass

Your logo

GNSS navigation data rates

140

Type GPS Galileo Glonass CompassBit rates [kbps] 50/200 50/125/200 50 25/50/100/

500

Bit durations [ms] 20/5 20/4/5 20 40/20/10/2

2/7/2012

36

Your logo

Pilot presence in GNSS

Type GPS Galileo Glonass CompassNo pilot C/A - all N/ATime-multiplexedpilots

L2C E1 - N/A

Quadrature-phasemultiplexed pilots

L5 E5 - N/A

Power/codemultiplexed pilots

L1C E6 - B1, B2

141 Your logo

GNSS comparison

GPS Galileo GLONASS CompassMultipleaccess

CDMA CDMA FDMA/CDMA* CDMA

Chip rates 1.023/5.115/10.23 1.023/2.5575/5.115/10.23

0.5115 1.023/2.046/2.5575/10.23

Channel coding

None/FEC FEC Parity bits N/F

Pilot channels Only in modernizedGPS

yes no yes

Modulationtypes

BPSK; TMBOC; SinBOC

BPSK; CBOC; SinBOC; CosBOC;AltBOC

BPSK BPSK; QPSK;MBOC; SinBOC; CosBOC;AltBOC

PRN codes Gold and Weil codes Memory codes M-sequences N/F

Frequencybands

L1,L2, L5 E1, E6, E5, E5a, E5b

L1, L2, L3*,L5* B1,B2,B3

142

* =planned; N/F=not found

Your logo

RECEIVER SIGNALPROCESSING

143 Your logo

Chapter 3 outline

Baseband processing:• Acquisition:

- Search Stage- Detection Stage

• Tracking:- Code tracking- Carrier Tracking

• Challenges dueto BOC modulation

Navigation processing:• Pseudoranges• Forming the location estimation• Satellite geometry

144

2/7/2012

37

Your logo

GNSS receiver processing

145

Basic operations are shown in the following block diagram

Navigation unit(computation of position, velocity and time: PVT

solution)

Sat 1Sat 2Sat 3Sat 4

Front-end and ADC

Baseband processing (acquisition,

tracking, data extraction

Coordinatesconversion ->map

position

01011...

(x,y,z; t)

fD, ...

Sat =SatelliteRuka skiSlope 10

Your logo

Code shift and Doppler frequency estimation are needed for reliable performance of any CDMA systemThe code synchronization task is typically split into:• coarse synchronization (or acquisition stage) and• fine synchronization (or code tracking stage).

Acquisition is used to get a rough timing estimate, say within +/- 1 chipTracking means finding and maintaining fine synchronizationCode tracking is much easier given the initial acquisitionCode acquisition, however, is usually considered as one of the most challenging tasks in a DS-SS systems

Acquisition and tracking problems

Your logo

Timing uncertainty: determined by the transmission time ofthe transmitter and the propagation delay (and clock uncertainties)• usually much longer than a chip duration.• due to a search through all possible delays, a larger timing

uncertainty means a larger search areaFrequency uncertainty: determined by the Doppler shift and mismatch between the transmitter and receiver oscillators.Angular uncertainty: determined by the direction of arrival of the signal. Separation in angular domain applies only for antenna array receivers (not applicable for current GNSS).

Uncertainty regions

Your logo

One correlation output forms a time-frequency bin. Several bins form a time-frequency window. The whole search space = time-frequency unertainty window.

Time-frequency uncertainty

2/7/2012

38

Your logo

Typically performed in 2 stages:• Search stage: build the time-frequency window with various

delay and frequency candidates (via correlations).• Detection stage - form a decision variable and compare it with a

threshold in order to detect whether the signal is present or absent. Threshold choice is an important step in designing a good acquisition stage. Either constant or variable (e.g., according to SNR) thresholds can be used, as it will be discussed later on.

The larger the time or frequency uncertainty is the greater time will take to achieve acquisition, or the greater receiver complexity is required for a given acquisition time requirement.

Code acquisition

Your logo

The search strategy can be:• serial: there is only one bin per window low complexity (only

one complex correlator needed), high acquisition time (many bins to be searched; acquisition time is proportional with the code epoch length, i.e. 1023 chips for standard GPS, up to 10230 for Galileo and modernized GPS).

• hybrid: there are several bins per window and there are several windows in the whole search space tradeoff between complexity and acquisition time; general case.

• parallel: there are more than one bin per window and there is only one window in the whole search space high complexity (need many correlators), low acquisition time.

Acquisition: Search stage

Your logo

Basic principle

The basic principle of code acquisition is: try and match each possible phase of the reference spreading sequence to the received data=> correlator

151

BPF= Band-pass filterED= Envelope DetectorSS= Spread Spectrum

Your logo

Examples

Left: signal present; right: signal absent (note: x-yaxes are different in these examples)

152

05

1015

20

0

100

200

3000

0.2

0.4

0.6

0.8

1

x 10-3

Time indexFrequency index

2/7/2012

39

Your logo

Correlation is done between the received signal and a reference (local) code at the receiver (e.g., C/A code)In time domain:

Time vs Frequency correlation

Discrete (I&D= Integrate and Dump)

Analog

In frequency domain:

FFT X

FFT

IFFT*

Rx signal

Ref. signal

Rx signal Rx signal

Ref. signal

153 Your logo

Example of FFT vs time correlation

154

Your logo

Acquisition: detection stage

155

Acquisition problem is in fact a detection problem: detect signal in noise, or, equivalently, separate between hypothesis H1 (signal plus noise arepresent) and hypothesis H0 (noise only is present).

Your logo

Acquisition: threshold choice

Threshold can be chosen as a tradeoff between a good detection probabilityand a sufficiently low false alarm (detection probability increases when false alarm probability increases). Below: PDF= Probability Density Function

The Probability Distribution Functions (PDFs) under each hypothesis arederived according to the decision statistic (see previous slide), which can be, for example, the maximum in a certain correlation window

2/7/2012

40

Your logo

Detection probability and false alarm probability: ’borrowed’ from classical detection problem of signal in noise

Acquisition: performance measures (I)

• A false alarm occurs when the decision statistic exceeds the threshold for an incorrect hypothesized phase•A miss occurs when the decision statistic falls below the threshold for a correct hypothesized phase.

157 Your logo

Mean Acquisition Time (MAT)Example:

Assume that the time uncertainty corresponds to N pseudorandom chips (or NTc = T seconds, where Tc = chip duration. Assume that there is no carrier frequency uncertainty (e.g., assisted acquisition)Assume also that detection probability at the correct hypotheses is Pd= 1 and the false alarm probability at incorrect hypotheses is Pfa = 0 (ideal case).What is the MAT time for a dwell time D if the timing search update is in half-chip increments (0.5Tc)?Answer: there are Q = 2N timing positions (hypotheses, bins) to be searched and the time to search one time bin is D Tacq = 2N D

If all hypotheses are equally probable the mean acquisition time can be approximated by half of Tacq => MAT N D

Acquisition: performance measures (II)

158

Your logo

Time to First Fix (TTFF):• The time from receiver turn on until the first navigation solution is

obtained.• The complete navigation solution (min 4 satellites in view) is needed to

compute TTFF• Typical values of TTFF: 30-40s without assistance; 1s with network

assistance (e.g., A-GPS)The TTFF is commonly broken down into three more specific scenarios:

• Cold (or factory): The receiver is missing, or has inaccurate estimates of, its position, velocity, the time, or the visibility of any of the GNSS satellites.

• Warm or normal: The receiver has estimates of the current time within 20 seconds, the current position within 100 kilometers, and its velocity within 25 m/s, and it has valid almanac data (i.e., coarse orbital information of satellites).

• Hot or standby: The receiver has valid time, position, almanac, and ephemerisdata (i.e., accurate orbital data of satellites), enabling a rapid acquisition of satellite signals.

Acquisition: performance measures (III)

159

Your logo

Once the coarse estimates of code delay and Doppler frequency are available, we move to the tracking stage, in charge with fine estimation of code delays and maintaining the lock:• Code tracking: fine delay estimation• Carrier tracking (needed for the carrier replica at the receiver)

Tracking

160

2/7/2012

41

Your logo

The purpose of a tracking loop is to reduce the estimation error of the delay between the received signal and the replica code at the receiver.A coarse delay estimate is obtained from the acquisition stage; the tracking loop is further refining/improving this estimate and trying to keep track of delay changes (e.g., due to the relative receiver-transmitter movement).In practice, we need both carrier tracking (e.g., achieved with a PLL) and code phase tracking; here we focus only on the code tracking part. A short discussion on carrier tracking can be found towards the end of this lecture.

Code tracking

161

Your logo

The basic code tracking loop is the Delay Locked Loop or Early-Minus Late (EML) correlator with 1 chip early-late spacing (also called wideband early-minus-late correlator). Both coherent and non-coherent (see below) DLLs exist.

Code tracking: DLL

162

Your logo

Looking for maximum correlation is equivalent with searching for the zero-crossing of the derivative of the correlation functionThe derivative can be approximated via early-minus-late correlation difference

Principle of DLL

E<L => delay the signal E=L => signal ’in tune’; stay there!

163

Your logo

The S-curve or discriminator curve translates the correlation values into a delay error index. For example:

S-Curve of Discriminator curve

The zero-crossings (from above) of the S-curve signal the presence of a channel path they signal the delay error. Note: if late-minus-early (instead of early-minus-late) curve need to consider the zero crossings from below

164

2/7/2012

42

Your logo

S-curve example

165

Your logo

According to the early-late correlator spacing, the early-minus-late correlator can be divided into:• Wide correlator (classical): =1 chip• Narrow correlator: Introduced by Dierendonck and Fenton in

1992, for GPS in order to reduce the code tracking error in the presence of noise and multipaths: < 1 chip

Narrow versus wide correlator

166

Your logo

DLL basic discriminator types

167 Your logo

Tracking noise variance: closed-form formulas are typically available for single-path static channelsMultipath Error Envelopes (MEEs): performance in multipathsMean Time to Lose Lock (MTLL, or MTTLL)

Tracking: performance measures

168

2/7/2012

43

Your logo

Problem: in single-path situation and in the presence of background noise (modelled as white Gaussian), which is the variance of the code delay tracking error?

Answer:• It depends on the discriminator type.

• Case studies:- Non-coherent early minus-latecorrelator

- Dot-product discriminator

Tracking: Noise variance

Early-late correlator spacing Code loop bandwidth

Signal-To-Noise or Carrier-To-Noise ratio Coherent integration interval

169

Your logo

When early-late spacing decreases tracking loop error variance decreases => benefit of narrow correlator.

Example of tracking noise variances

Early-minus-late power Dot product

Your logo

’Lock’ condition: delay error smaller than a certain threshold (in absolute value); e.g., 0.5 chips for GPS C/A code and classical early-minus-late correlatorDue to noise, delay error might become too high and we go out the convergence region of the S-curve=> ’loss of lock’MTLL is a statistical measure showing the ability of a discriminator (or code tracker) to maintain the lock. The higher, the better.

Mean Time To Lose Lock (MTLL)

171

Your logo

Needed to obtain carrier information (i.e., to build an exact carrier replica at the receiver)Phase Locked Loops (PLL) are used to obtain carrier phase informationFrequency Locked Loops (FLL) are used to obtain carrier frequency information

Frequency and phase tracking (FLL/PLL)

172

2/7/2012

44

Your logo

Similar principle with DLL. Different discriminator equations (e.g., atan(Q/I), ...)

Carrier tracking loops

173

Your logo

Source: Kaplan’s book.

Block-diagram of carrier-code tracking

174

Your logo

Challenges due to BOC modulation

175

Most of the Galileo signals are split-spectrum signals, that is with power spectral density content away from the middle of the band.

Your logo

Galileo – specific processing: ambiguities

- Due to the split-spectrum modulations (BOC, MBOC) => ambiguities (notches) in the envelope of the correlation function and additional side lobes

2/7/2012

45

Your logo

Ambiguities effects

- Acquisition stage:- Time-bin step in the searching process needs to be smaller (in

chips) than in BPSK modulation => longer time to spend in acquisition stage or need to remove the ambiguities

- Tracking stage:- False lock peaks- More difficult to cope with multipaths (how to make the

distinction between a ’side peak’ and a multipath)

Your logo

- Left: small bin step. Right: large bin step . - The number of states in the acquisition process is inversely

proportional to the time-bin step => a smaller step meanshigher time spent in the acquisition (or higher complexity)

Time-bin step in acquisition

Your logo

Unambiguous acquisition (I)

- Unambiguous aquisition methods: try to recover a ’BPSK-like’ correlation shape, such that a higher time-bin step can be used in the acquisition process

- As a rule of thumb, a time-bin step of ¼ of the width of the main lobe in the ACF envelope is used (e.g., 0.5 chips in GPS BPSK-modulated signals, about 0.35 chips in Galileo MBOC/SinBOC(1,1)-modulated signals)

Your logo

Example of unambiguous acquisition

Betz&Fishman unambiguous acquisition:

2/7/2012

46

Your logo

Example of performance: unambiguous acq.

Mean Acquisition Time MAT

181 Your logo

In the tracking stage, the main problem related to Galileo signals is how to deal with the additional side lobes (lock to a false peak), while preserving the narrow width of the main correlation lobeDelay accuracy is directly proportional to the width of the main correlation lobe (its envelope)The unambiguous acquisition methods are not suitable in the tracking stage, because the width of the main correlation peak is increased in unambiguous acquisition => poorer tracking performance

Unambiguous tracking (I)

182

Your logo

Unambiguous tracking (II)

Sidelobe cancellation/mitigation approaches (SCM)

183 Your logo

Recall: GNSS receiver processing

184

Navigation unit(computation of position, velocity and time: PVT

solution)

Sat 1Sat 2Sat 3Sat 4

Front-end and ADC

Baseband processing (acquisition,

tracking, data extraction

Coordinatesconversion ->map

position

01011...

(x,y,z; t)

fD, ...

Sat =SatelliteRuka skiSlope 10

2/7/2012

47

Your logo

Underlying principle: trilateration

1 Satellite 2 Satellites => intersection is the red circle

3 Satellites=> 2 points of intersection; one point is outside Earth

185 Your logo

Pseudoranges

186

7.2.20

Velocity x Time = Distance

Radio waves travel at the speed of light, roughly299 792 458 m/s (i.e., around 3*108 m/s).E.g., the time needed by a radio signal sent from a GPSsatellite to reach the Earth surface is about20000km/(3*108) m/s = 66.7 msIn the above equation, time = tR (Receiver time) – tT(Transmit time).

Pseudorange= estimate of the true sv-tx range (distance)

Your logo

Pseudorange concept

187

Pseudorange = true range + errors3,2.1, iiiii d

Errors:- Clock errors (always present)- ionospheric and troposheric errors (satellite transmitters

only)- Multipath errors, etcErrors can be common to all transmitters (e.g., clock errors with stable transmitter clocks, such as in satellite case) or transmitter-dependent (e.g., multipaths)

Your logo

Navigation equations based on pseudoranges

188

Receiver clock may have an (unknown) bias compared with the transmitterclock

2/7/2012

48

Your logo

Satellite positions: ephemeris

Navigation data carries information about almanac(coarse orbit information) and the ephemeris (pl. ephemerides) or the accurate information about the satellites orbits and movements

189

Examples of ephemeris data:a: Semi-major axis, e : Eccentricity (deviation) of orbit,i: Inclination angle,l : Right ascension of ascending node, …

Your logo

Navigation data decoding

190

The output from the tracking loop is the value of the in-phase arm of the tracking block truncated to the values 1 and 1. Theoretically we could obtain a bit value every ms. However, we deal with noisy and weak signals =>a mean value for is computed and truncated to 1 or 1.

In GPS, one navigation bit has a duration of 20 ms => mean over 20 ms (see an example in the figure)

Your logo

Satellite geometry

191

When the satellites are all in the same part of the sky, measurements will be less accurate.

Your logo

Satellite geometry: GDOP

A measure of the ’goodness’ of satellite geometry is the Geometrical Dilution of Precision (GDOP)GDOP has several components (e.g., Vertical DOP, Horizontal DOP)An intuitive physical measure of DOP is the inverse of the volume of the tetrahedron that is formed by the satellites and the receiver: 1/V

Lower DOP => better geometry

192

2/7/2012

49

Your logo

Carrier-phase enhancements (I)

With code phase measurements, at the speed of light a microsecond (i.e., about 1 chip in GPS) is almost 300 meters of error. Carrier frequency has a cycle rate of over a GHz (which is 1000 times faster!)E.g., the wavelength of L1/E1 signals is

If we can measure the carrier cycles between the transmitter and the receiver (instead of the code delays) => centimeter level accuracy

193

cmm 05.1910*575.1

10*39

8

Your logo

Carrier-phase enhancements (II)

Sine carrier:

We can express the phase cycles as:

194

))(sin()2sin()( 0 tAftAts

2*)( 0tft cycle

Your logo

Carrier-phase enhancements (III)

195

. ..

Finding N= ambiguity resolution problem -> N unknownbecause the receiver merely begins counting carrier cycles from the time a satellite is placed in activetracking mode

Source: Kaplan book

Your logo

Solving carrier-phase ambiguity

Reception of signal of two frequencies or via 2 antennas=> same N factor on both => the error can be corrected

Integer ambiguity remains the same, as long as nocycle slips occur

196

2,1,, ifN tropoionoiclockbiasii

Phase observation

2/7/2012

50

Your logo

High precision GNSS positioning

Combinations of carrier-phase and code-phasemeasurements are possibleIn addition to carrier-phase and code-phasemeasurements, we can also have the Precise Pointpositioning (PPP):- Uses a single receiver + reads correction data from some

databases in order to model and cancel errors: Satellite orbiterrors, Satellite and/or receiver clock errors, Fractional cyclebiases, Ionosphere errors, Troposphere errors, Antenna-relatederrors, Relativistic effects

197 Your logo

WIRELESS CHANNELS: MULTIPATHS AND OTHER CHANNEL IMPAIRMENTS

198

Your logo

Chapter 4 outline

Wireless channel impairments:• Multipaths• Ionospheric and troposheric delays• Other interference sources

Mitigation approaches:• Advanced dekay tracking structures

199 Your logo

What is multipath?

200

-splitting of signal into 2 or more components due to reflections, scattering, refractions, dispersion, etc.

- replicas of the same tx signalarrive at the rxwith different attenuations (amplitudes), phases and delays

20/08/2009

2/7/2012

51

Your logo

Multipath terminology

LOS = Line Of Sight (directgeometrical non-obstructed pathbetween transmitter and receiver)NLOS = Non Line Of Sight (anymultipath component which is notLOS)TOA = Time Of Arrival

201 Your logo 202

GNSS Signal Transmission: Line Of Sight

Line of sight is the ability to draw a straight line between two objects without any other objects getting in the way.

GNSS receivers assume Line Of Sight (LOS) transmissions.

Obstructions such as trees, buildings, or natural formations may prevent clear line of sight.

Your logo

Multipath effect on position

203

Triangulation principle for PVT computation is based on LOS TOAIf incorrect LOS estimate or NLOS case only => link-level errors will affect the final PVTExact amount depends on how many links are affected and what is the final algorithm for PVT computation

A rule of thumb at link-level:

e rr delay error => e r rc distance error.

c=3*108 m/s (speed of light)

Your logo

Example 1: multipath effect

2 channelpaths, shown in red dotsWithoutmultipathmitigation=> trackingerrors

204

2/7/2012

52

Your logo

Example 2: multipath effect

The paths canadd togetherconstructively ordestructively(according to their phases)The multipatherrors are notnecessarilyincreasing withthe number of paths

205 Your logo

Challenges in GNSS

Falsepeaks dueboth to multipathsand BOC modulation(whenpresent)

206

Your logo

Criteria to evaluate multipath effects

207

Link-level criteria:• Most used criterion is the Multipath Error Envelope (MEE),

see next slide• Multipath delay error mean/variance/root mean square

error (RMSE)• Probability Distribution Functions (PDF) of the delay

errors• Carrier to Noise Ratios (CNR) needed to achieve a

certain performance levelSystem-level criterion• Ultimately, the error on the estimated PVT (mean,

variance, RMSE,…) would be the most meaningful (also the hardest to evaluate during the algorithm design, since the whole chain including the navigation algorithms should be simulated)

Your logo

With multipath => zero-crossing point in S-curve is displaced, causing errors in the estimation. The horizontal error shown below is called MEE and it gives a measure of the discriminator ability to deal with multipaths

Multipath Error Envelopes (MEEs) (I)

208

2/7/2012

53

Your logo

Multipath Error Envelope (MEE) (II)

2 static pathseither in-phase (0 degrees phaseshift) or out-of-phase (180 degrees phaseshift)Area in betweenthe curves definesthe error region

209

Example

Your logo

How to cope with multipaths?

Stand-alone GNSS:• At baseband link-level: try to reduce/remove

the multipath effect for each satellite.• At navigation layer: system-level (multi-

satellite)• Combined link-level/system-level solutionsWith some external aid: • sensors, WLAN, inertial navigation systems

(INS), etc.

210

Your logo

High Resolution Correlator or Double-Delta correlators (also called Pulse Aperture Correlator, Strobe Correlator or Very Early-Very Late Correlator)Multiple Gate Delay structures (MGD): generalization of the multi-correlator structuresMultipath Estimator Delay Locked Loop (MEDLL)Deconvolution-based mitigationNon-linear operator based (e.g., Teager-Kaiser) Peak tracking (PT)

Advanced code tracking structures

211

Your logo

It adds 2 extra (complex) correlators: very early (VE) and very late (VL) (or, equivalently, 4 extra real correlators)It forms the discriminator output via:

The factor a above is typically set to 0.5It offers more resolution in multipaths

High Resolution Correlator (HRC)

212

2/7/2012

54

Your logo

0.05 chips early-late spacing. BW=front-end bandwidth

MEE performance of narrow correlator vs HRC

213

Your logo

Variable number of correlators and variable spacings; possible to optimize those according to channelGeneral case; it covers the early-minus-late (3 correlators) and HRC (5 correlators) structures

Multiple Gate Delay structures (MGD)

214

Your logo

Maximum Likelihood (ML)-based approachReceived signal:After correlation with the reference code

Multipath Estimator Delay Locked Loop (MEDLL)

Amplitude of the l-th path of the channel Phase of the l-th path of the channel

Delay of the l-th path of the channel

215

Your logo

Number of paths is fixed to a certain value beforehand (e.g., max 2 paths) or the algorithm is repeated until a certain threshold is metAmplitudes, phases and delays are estimated for all paths, via maximum likelihood principle:

MEDLL

216

2/7/2012

55

Your logo

Example: MEDLL estimation in good CNR

217 Your logo

The received signal is the result of the convolution between the (satellite) transmitted signal s(t) and the channel impulse response h(t):

Deconvolution methods

Deconvolution attempts to estimate the channel response h(t), invert it and then convolve it with the incoming signal, to estimate the transmitted signal. Usually, solved iteratively.Also constrained deconvolution methods possible, e.g., POCS (Projection Onto Convex Sets)

218

Your logo

z(n)= correlation functionThe following non-linear transform is applied on z(n); see below an example of the correlation after the non-linear transform in 1-path and 4-path cases

Teager Kaiser estimator

219 Your logo

Taken from [BLR08]. Some of the algorithms in the plot have not been metioned here. Among those mentioned in the course, nEML=narrow early-minus-late correlator; HRC= High Resolution Correlator; MEDLL= Multipath Estimator Delay Locked Loop and TK= Teager Kaiser

Performance Example

2/7/2012

56

Your logo

Examples of advanced code tracking algorithms used by some GPS manufacturers

Company Multipath mitigation algorithm*

Ashtech Strobe CorrelatorCedar Rapids HRCMagellan Strobe CorrelatorNovatel MEDLL ,Pulse Aperture

Correlator, ...Sokkia Pulse Aperture Correlator

Your logo

Other channel impairments in GNSS

FadingAtmospheric delays:• Ionospheric delays• Tropospheric delays

Doppler shifts/spreadsNarrowband and wideband interferences

222

Your logo

Fading

Due to mobility, the channel

characteristics change over time• signal paths change• different delay variations of different

signal parts• different phases of signal partsquick changes in the instantaneous received power (short term fading)slow changes in the average receivedpower (long term fading)

t

223dB= Fading random component

s2

s1

s1+s2

s2

s1

s1+s2

Your logo 22

Atmospheric layers

Ionosphere: 50-1000 km above the Earth – introduces large, randomly fluctuating delays, frequency dependent

Troposphere: 0-50 km above the Earth – introduces small, randomlyfluctuating delay, frequencyindependent

2/7/2012

57

Your logo

Random, frequency dependent delayIt depends on latitude of the receiver, season, time of day, solar

activity, etc.

Ionospheric delays

22

2

3.40fTEC

iono

TEC= Total electron content, randomly fluctuatingf = carrier frequency

If a transmitter transmit on 2 or more frequencies, the TEC term is the same on all frequencies => we can use differential combining to reduce/remove the ionospheric delay

Your logo

Example: Ionospheric delays in Galileo

226

Your logo

Tropospheric delays

Caused by the signal refraction in the electrically non-ionized atmospheric layerTropospheric delay is a function of the satellite elevation angle and the altitude of the receiverDepends on atmospheric pressure, temperature, and water vapor pressureUnlike ionospheric effects, tropospheric ones do not depend on the signal’s frequency

227 Your logo

Chanel impairements: Doppler shifts/Doppler spreads

If the receiver or transmitter are moving, => Doppler shifts/spreads

If frequency f is transmitted, the received frequency = f +/- fD+ if mobile is moving towards the transmitter- if mobile is moving away fromthe transmitter

The Doppler shift equals

Also moving objects in the environment may introduce fast fading effects, even if the receiver and transmitter are stationary.

cosvfD

angle between the direction of motion and the direction of the wave wavelength

speed of the receiver

228

2/7/2012

58

Your logo

Narrowband and wideband interferences

Wideband interference:• Inter-system (e.g., GPS-Galileo), sharing

same frequency bans• Intra-system (e.g., OS and PRS signals in

E1 band in Galileo)Narrowband interference:• Unintentional (LightSquared)• Intentional (jamming, spoofing)

229 Your logo

Summary and conclusions (I)

More than 110 navigation satellites forecast for the next5+ years4 main GNSS systems3 frequency bands per GNSS systemCDMA technique common for all (in GLONASS, onlyused for spreading; in GPS/Galileo/Compass used bothfor spreading and multiple access)Several signals transmitted in each band (more to come in the future)Several modulation types, with BOC family as the main candidate (generic models available)

230

Your logo

Summary and conclusions (II)

Main baseband tasks:• Acquisition• Tracking

GNSS affected by several sources of errors, also specific to other wireless systems (e.g., multipath, fading, Doppler shifts, interferences)Multipath and ionospheric delays are two of the most difficult phenomenon to cope with in GNSS receivers.

231 Your logo

Trivia: Galileo open-source software

Simulink Galileo E1 and E5a baseband transmitter-receiver chains, built within the GRAMMAR (Galileo Ready Advanced Mass Market Receiver, funded from the European Community's Seventh Framework Programme FP7/2007-2013 under grant agreement n°227890.) project- available as open-source (see http://www.cs.tut.fi/tlt/pos/Software.htm)

232

2/7/2012

59

Your logo

Some references (I)

[LLR06a] E. S. Lohan, A. Lakhzouri, and M. Renfors, "Feedforward delay estimators in adverse multipath propagation for Galileo and modernized GPS signals", EURASIP Journal of Applied Signal Processing, vol 2006, Article ID 50971, 19 pages[LLR06b] E. S. Lohan, A. Lakhzouri, M. Renfors, "Binary-offset-carrier modulation techniques with applications in satellite navigation systems", Wireless Communications and Mobile Computing, Volume 7, Issue 6 (p 767-779), 2006.[LLR06c] E. S. Lohan, A. Lakhzouri, M. Renfors, ``Complex Double-Binary-Offset-Carrier modulation for a unitary characterization of Galileo and GPS signals'', IEE Proceedings on Radar, Sonar, and Navigation, vol. 153(5), pp. 403-408, Oct 2006.[Win06] J.O. Winkel, ”SPREADING CODES FOR A SATELLITE NAVIGATION SYSTEM”, ESA Patent application WO/2006/063613, http://www.wipo.int/pctdb/en/wo.jsp?WO=2006063613&IA=EP2004014488&DISPLAY=STATUS

233 Your logo

Some references (II)

[WRH07] Wallner, S., Avila-Rodriguez, J.A., Hein, G.W.: Galileo E1 OS and GPS L1C Pseudo Random Noise Codes - Requirements, Generation, Optimization and Comparison -, Proceedings of ION GNSS 2007, Fort Worth, Texas, USA, 25-28 September 2007[PP2010] B.W. Parkinson and S.T. Powers, ”Fighting to survive: fivechallenges, one key technology, the political battlefield – and GPS mafia”, GPS World, Jun 2010, pp.8-18. [Bet00] J. W. Betz, “Effect of Narrow Interference on GPS Code TrackingAccurancy,” in Proc. ION 2000, Institute of Navigation, Anaheim, CA, January26-28. 2000, pp. 16–27.[Bet01] J. W. Betz, “Effect of Partial-Band Interference on Receiver Estimationof C/N0: Theory ,” in Proc. ION 2001,Institute of Navigation, Long Beach, CA, January 22-24. 2001, pp. 16–27.[LLR05] E. S. Lohan, A. Lakhzouri, and M. Renfors ,”Benefits of using lower chip rates for Galileo OS and PRS signals”, ENC-GNSS 2005, Munich.

234

Your logo

Some references (III)

[HRV06] M. Fantino, P. Mulassano, F. Dovis, and L. L. Presti, “Performance of the Proposed Galileo CBOC Modulation in Heavy MultipathEnvironment,” Wireless Personal Communications, vol. 44, pp. 323– 339, Feb 2008. [ALI08] G. Artaud, L. Lestarquit, and J.-L. Issler, ”AltBOC for dummies oreverything you always wanted to know about AltBOC”, in Proceedings of the ION GNSS 2008, Savannah, GA, Sep 2008.[Cho09] C. Chong, Status of COMPASS/BeiDou Development , Standfordpresentation,http://scpnt.stanford.edu/pnt/PNT09/presentation_slides/3_Cao_Beidou_Status.pdf[BLR08] M.Z.H. Bhuiyan, E.S. Lohan, and M. Renfors, ``Code Tracking Algorithms for Mitigating Multipath Effects in Fading Channels for Satellite-based Positioning'', EURASIP Journal on Advances in Signal Processing , vol. 2008, article ID 863269, published on-line

235