GNSS Navigation message analysis and perspectives

-1-GNSS Navigation message analysisand perspectivesInternational Technical Symposium on Navigationand TimingDr Axel GARCIA-PENA - ENACNovember 2014-2-Introduction (1/2)• The GNSS navigation message is a key element of a GNSS system since it carriesessential information to calculate the user PVT (Position Velocity Time) Essential information: Satellites Ephemeris and Satellite Clock Error corrections(CED), message time stamp• Traditionally, less crucial information such as the satellites almanacs, the ionosphereand the troposphere correction, etc, was also carried by the message but the totalamount of data remained small and easily manageable.• Therefore, the implementation of the communication part of the GPS L1 C/A signalwas subordinated to the implementation of the synchronization part-3-Introduction (2/2)• The communication part of a signal can be divided into two main functions:1) Transmission of the information: How the information is transmitted and whichC/N0 is required to access this information Fields: Digital Modulation, Channel Coding2) Information Integrity: How the transmitted information integrity is guaranteed Fields: Channel Coding, Authentication (not presented)• Nowadays, the continuous growth of GNSS applications results in the considerableincrease of the information to be broadcast such as Precise Positioning, Safety of Life(SoL)• Moreover, due to the expansion of the GNSS mass market applications in suburban,urban and indoor environments, the GNSS signal has to be adapted to suburban,urban and indoor environments• Therefore, the GNSS navigation signal and message must continue to evolve in orderto meet the ever evolving user needs.-4-Outline1.

Navigation Message Evolution2. New demodulation performance methodology in mobile channels3. Code Shift Keying (CSK) modulation4. GNSS Signal with variable data rate-5-Outline1. Navigation Message Evolutiona) Navigation message GPS L1 C/Ab) Navigation message GPS L2C, L5 and SBAS L1c) Navigation message GALILEO E1 OSd) Navigation message GPS L1Ce) GNSS signal comparison in an AWGN channel2. New demodulation performance methodology in mobile channels3. Code Shift Keying (CSK) modulation4.

GNSS Signal with variable data rate-6-1. Navigation Message Evolution• A good perspective of the navigation message evolution is necessary to design a newsignal/message. The following characteristics of the message are discussed:1) Message structure2) Forward Error Correction:• Channel code is implemented to allow the message correction at reception.• Demodulation performance depends on the modulation and FEC channel code3) Message content Integrity:•How the data integrity is checked and its limitations4) Message content:• Whether the message is repeated from one frame/subframe to next one Allow the coherent message accumulation  C/N0 increase• Only meaningful evolutions will be commented for each message-7-1a.

Navigation message GPS L1 C/A (1/2)Message Structure (sent to the channel):• Continuous stream of frames: a frame of 1500 bits is constituted of 5 subframes• Each subframe of 300 bits is constituted of 10 codewords• Each codewords is constituted of 30 bits: 24 information bits and 6 parity bits• Rigid structure: Contents of each subframe follow a cyclic scheduleForward Error Correction (FEC):• No channel code is implemented to apply the FEC• The GPS L1 C/A signal demodulation performance depends only on the demodulationperformance of the BPSK modulation-8-1. Navigation message GPS L1 C/A (2/2)Integrity:•GPS L1 C/A message integrity is controlled by an Extended Hamming code (32,26)•Theoretically, this code can detect up to 3 errors with a 100% probability in acodeword.•Theoretically, the probability of not being able to detect a pattern of errors in acodeword is: 226== 32 = 2−6 = 0.0156 2-9-1b.

Navigation message GPS L2C, L5 and SBAS L1 (1/2)Information Message Structure: GPS L2C and GPS L5• Continuous stream of information messages.• Each message is constituted of 300 bits: 276 information bits and 24 CRC bits• First information bits are: 8b preamble, 6b SV Id, 6b Message Id• Flexible structure: Message contents (or Id) are determined by time outsInformation Message prior to BPSK modulator:• The continuous stream of information messages are fed to an encoder block• The output of the encoder block is sent (after BPSK modulation) to the channel- 10 -1b. Navigation message GPS L2C, L5 and SBAS L1 (2/2)Forward Error Correction (FEC):• Referred as Inner channel code, it is aconvolutional channel code, CC(171,133)with = 1/2, = 7• Therefore, each information word isconverted into 600 coded bits.~5dB• The introduction of a CC makes thenavigation message more robust to the noise: For a BER=10-5, about 5dB less will be requiredIntegrity:• GPS L2C message integrity is controlled by a Cyclic Redundancy Code, CRC-24Q• CRC-24Q performance:– It detects all single bit error, double bit error, any odd number of errors and anyburst error with length b ≤ 24 bits per code word (300 information bits)– The fraction of error bursts of length b > 24 that are undetected is: = 2−23 = 1.19 × 10-7, if b = 25 bits = 2−24 = 5.96 × 10-8, if b > 25 bits- 11 -1c.

Navigation message GALILEO E1 OS (1/2)Information Message Structure:• Continuous stream of frames, where each frame isconstituted of 24 subframes• Each subframe is constituted of 15 pages (or words)• Each page (or word) is constituted of two page parts• Each page part is constituted of 120 bitsInformation Message prior to CBOC modulator:• GPS L2C/L5 scheme is applicable to GALILEO E1 OS, but with some modifications.• Before, sending the page parts to the channel (after CBOC modulation), the pageparts are encoded, interleaved and a preamble of 10 bits is inserted at the beginning.- 12 -1c. Navigation message GALILEO E1 OS (2/2)Forward Error Correction (FEC):• The same channel code as GPS L2C is used to implement the FEC: CC(171,133)• Difference: 6 tail bits are inserted at the end of each page part– Advantage: Allows the implementation of an interleaver– Drawback: Preamble bits are no longer protected by the FEC (w.r.t.

GPS L2C)Interleaver:• Rectangular interleaver with dimensions, (30 columns x 8 rows), applied to 1 page• Function: To make the navigation message more robust to its transmission through amobile channel by breaking the burst of errors introduced by the channel- 13 -1d. Navigation message GPS L1C (1/2)Information Message Structure:• Continuous stream of frames, where each frame is constituted of 3 subframes• Subframe 1 contains the Time of Interval (TOI) and is constituted of 9 information bits• Subframe 2 contains the CED and isconstituted of 600 bits (576 info bitsand 24 CRC bits)• Subframe 3 contains variable dataand is constituted of 274 bits (250info bits and 24 CRC bits)Information Message prior toTMBOC modulator:• For the first time, subframes havedifferent sizes and are encoded withdifferent channel codes• Subframes 2 and 3 are interleavedtogether- 14 -1d.

Navigation message GPS L1C (2/2)Forward Error Correction (FEC):• Subframe 1 is encoded with an extendedBCH(52,9) with minimum distance, dmin=20• Subframe 2 is encoded with a LDPC(1200,600)and subframe 3 with a LDPC(548,274)• LDPC are very powerful codes which obtain nearshanon limit performance~2,1dB• Both LDPC codes and BCH(52,9) obtain betterdemodulation performance than the CC(171,133)Message Content Repetition• Subframe 2 replicas can be coherently accumulated for 2h and thus the received C/N0is increased with each coherent accumulation: /0 = 10 log10 - 15 -1e. GNSS signal comparison in an AWGN channel• The demodulation performance of signal in an AWGN as a function of the receivedC/N0 depends on three main factors:1) Signal modulation  BPSK for all GNSS/SBAS signals (from demodulationperformance point of view)2) Implemented channel code to apply the FEC3) Symbol rate• The Clock Error Corrections and Ephemeris Error Rate (CEDER) as a function of thereceived C/N0 in an AWGN is presented for the main GNSS signals10CED Error Rate101010100-1-2-3AWGN-GPS L1C/AAWGN-GPS L2CAWGN-GPS L1CAWGN-Gal E1OS-41618202224C/N0 [dBHz]262830- 16 -Outline1.

Navigation Message Evolution2. New demodulation performance methodology in mobile channels3. Code Shift Keying (CSK) modulation4. GNSS Signal with variable data rate- 17 -6. New demodulation Performance methodology (1/4) Developped in M. ROUDIER Ph.D. thesis at ENAC.Classical Methodology: Data set Error Rate as a function of the received C/N0•This methodology presents 2 main limitations for urban environments Limitation n°1: Received C/N0 is not constant for signals received in urbanenvironments• Objective:10CED Error Rate10100 Find a C/N0 which is constant a long timefor any urban user-1 Find a C/N0 which is representative froman operational point of view.•-2Proposition:1010-3AWGN-GPS L1C/AAWGN-GPS L2CAWGN-GPS L1CAWGN-Gal E1OS-41618202224C/N0 [dBHz]262830GNSS signals demodulation performance in the AWGNchannel with ideal code delay and carrier phase trackingCpre-urban/N0:- 18 -6.

New demodulation Performance methodology (2/4)• Operational point of view: a user moving inside an urban environment observes aconstant Cpre-urban/N0 from a satellite with a fixed elevation• Future step: Knowing the receiver architecture and interference environment, thedemodulation performance can be given as a function of the satellite elevation Limitation n°2: For a GNSS signal, only punctual instead of continuous messagedemodulations are required• Objective: Combine the next characteristics:1010GPS L1C Prieto 40°-elevation0 Urban environments have dynamic signalreception conditionsUnsatisfactorydemodulationperformance-1AWGNPrieto-all states-idealPrieto-all-states-PLLPrieto-GOOD states-idealPrieto-GOOD states-PLL-222 GNSS requires punctual demodulation242628303234Cpre-urban/N0 [dB-Hz]Cpre-urban[dBHz]363840GPS L1C CED demodulation performance with thePrieto model0(dB)Channel AmplitudeChannel module [dB]CED Error Rate10-10-20-30-40-5051015202530Time in s35404550Channel amplitude with the Prieto channel model- 19 -6.

New demodulation Performance methodology (3/4)•Proposition:1) Separating the reception message conditions into favorable receptionconditions and unfavorable reception conditions2) Providing demodulation performance for favorable reception conditions3) Providing the occurrence rate: Statistical information about the occurrenceof the message favorable reception conditions.• Occurrence rate:12 Key point: To define a criterion which providesthe ‘favorable state messages’ having the bestpossible demodulation performance (work to do).Normalized Count [%]108642X= 0Y = 0.59900510152025Number of favorable states messages in 1 hour (=200 messages)Favorable state occurrence rate, = , for GPS L1C frame using thePrieto channel model with 40° of elevation considering the Prieto channelGOOD states as favorable reception conditions- 20 -6.