Benefits of combining systems The Receiver s Perspective Dr Philip G Mattos October 2011
Contents Who we are What s missing in GPS alone Other constellations available Improving GPS only receivers Add other constellations What vulnerabilities remain Viability of the all-constellation receiver Technical Economic Prospects for a future-proof receiver Ready for as yet unspecified signals 2
Who Philip G Mattos Cambridge/British Telecom Research in the 70 s Chief engineer of ST s GPS and Navigation team Working on GPS since 1987 Consultant on Galileo since ~1997 Designed receivers for Loran, GPS, Galileo, Glonass and Compass STMicroelectronics 10 billion dollar semiconductor company Major supplier of chips for satnavs, in-car navigation etc since 1991 Dedicated GPS/Navigation processors since 1995 3
What s missing in GPS alone Different for different markets Government, safety of life -integrity, independence The argument for Galileo Everyone, infrastructure : single point of failure, no backup Very weak signal Easily jammed, accidentally or maliciously Consumer : limited accuracy/availability indoors/urban People gather in towns, and indoors GPS was designed 30+ years ago We should be able to do better now! 4
What have we learnt? Sensitivity GPS was designed to be used with signals from 50dB Cno down to 35dB Cno Consumer receivers now position at below 15dB Cno, and track down to 10dB Cno 35-15 = 20dB, ie 1% of the signal Multipath Reflections Distorts main signal and hence accuracy Or allows tracking in urban canyons! New signals claim to improve accuracy 5
Other constellations available Glonass Russian system of the 80s/early 90 s Declined with Russian economic problems Now almost up to full strength again Galileo European system with first real satellites this month (GIOVE test satellites for several years) Compass (Regional, B2) East Asia Coverage 2012 (B2 Signal) World FOC 2020 (B3 signal) 6
Glonass 24 satellites in 3 planes Frequency 1602MHz centre FDMA each satellite on a different frequency Spread spectrum used only for timing/nav, not for CDMA Historically difficult due to FDMA/expensive receiver As silicon capability improved, satellites died! 2011 STA8088 Teseo-II first consumer GPS/ Glonass/ Galileo Receiver Future version will be CDMA on GPS/Galileo frequency no spec yet 7
Galileo European equivalent to GPS 30 satellites planned, but very delayed 2 test satellites flying, but only test signals First 2 real satellites launch October 20 th Same centre frequency as GPS Minimal extra cost Use as one giant constellation of 60 satellites Availability In-town accuracy 8
Compass/ Beidou 3 Generations BD1 is active loop delay, Chinese govt use only BD2 is regional GEO/IGSO system for East Asia 1561 MHz signal, spec not published Signal Available now, Completion 2012 BD3 Worldwide system 1575 MHz BOC(1,1) compatible with GPS/Galileo Also no specification Completion 2020 9
Improving GPS only receivers Add other constellations More satellites More coverage in town Better geometry 100% availability Gives reliability inter-system RAIM Independence keeps governments/infrastructure happy What vulnerabilities remain All GNSS are weak, microwave signals Tunnels and buildings Jammable (but it is harder to build a multi-band jammer) 10
RF architecture Integrated LNA Common GPS/Glonass RF input from antenna RF amp LNA Mixer Amp HPF IF filter PLL AGC ADC GPS 3bits + clk TCXO/ XTAL Glonass is around 30MHz after mixer filter /n Clockgen CPU clks Fed to second mixer to convert to 8MHz Separate IF and ADC LDO Control SPI GAL-GPS Mixer key IF filter AGC GLONASS ADC Glonass 3bits + clk 11
Baseband GGG 4fo fully parallel correlators for 4 satellites IF0 4fo FIR Down sample Notch Filter Mixer Acquisition engine Acquisition memory 32fo->16fo 8fo IF1 8fo FIR Down sample Notch Filter Mixer 64fo->32fo 32 Tracking channels GAL-GPS GLONASS Processor interface ARM 9 CPU 12
Sats tracked All sats tracked, even in canyons (by reflection) One tunnel reduces sats to zero 13
Canary Wharf Blue : GNSS (patch) Green : GPS only(linear) 14
Canary Wharf Blue : GNSS (patch) Green : GPS only(linear) 15
GNSS = more sats Blue = GNSS GPS+GLONASS Plenty of sats Red = GPS only Too few sats example Green = GPS only Linear antenna Accepts reflections On this occasion, beneficial 16
Tokyo GNSS 17
Multiconstellation benefit Road Test/1 Tokyo Dallas Competitor GPS TII GPS+GLO TII GPS TII GPS+GLO
Multiconstellation benefit Road Test/2 TII GPS TII GPS+QZSS
Viability of the all-constellation receiver Technical Not a problem Easiest on GPS/1575.42 but remains jammable Already done on GPS/Galileo/Glonass, nearby frequencies All that is needed is the will, see economic below (and the spec!) Economic Extra Costs per unit are small The cost is in development of both silicon and software Cannot be done without guaranteed volume sales Tens of Millions of units, or hundreds of millions in the phone market Risk of designing for the future with no spec! Cannot design new chips fast enough Must make them generic to reduce the number of designs. 20
Prospects for a future-proof receiver Ready for as yet unspecified signals Any frequency (in the set of allocated bands)(1600, 1200 MHz) Any code length, structure, chipping rate Hardware(SR) generated codes LFSR,Gold,Weill. Truncated,Appended Random(memory) codes Any modulation scheme (BPSK, QPSK, BOC ) Any symbol rate, datarate, FEC Flexible hardware Configured by software with fine feature granularity All but fastest DSP in software Software designed with multi-constellation in mind ANY is a big word despite only having 3 letters 21
Dual/Multi frequency GPS/Galileo etc use 2 bands, 1.2 and 1.6 GHz Some cost involved to support dual band Filters, architectural redesign hardware and software Actual silicon cost is low. Benefits are small in consumer May penetrate buildings/trees a little better Marketing tick box Anti Jamming redundancy Multipath/fading redundancy Indian IRNSS proposes 2.4GHz Interference issues Stands alone...no critical mass of constellations 22
Conclusions 1 In the short term, all receivers will receive all satellites transmitting on L1 (1575MHz) Minimum GPS Galileo Many will also handle either GPS-Glonass GPS-Compass Regional In the long term all receivers will receive everything 2-3 years for the chips Many will also support Dual frequency?3-5? years for the chips?5-7? Years in the field 23
Conclusions 2 Superb accuracy and availability in town But still not indoors from GNSS! Patriotic sensitivities recognised Same chips, same mobile, can be used everywhere I can retire. Thankyou 24