Advanced global navigation satellite system receiver design (797918), страница 31
Текст из файла (страница 31)
The remaining 44.4& percent is split equallybetween the B and C components. Code combinations correlating the addition orsubtraction of the B and C components can be used to recover all of the 44.4& percentof transmitter power given to these components. This is of particular importance forhigh-sensitivity receivers or those operating in weak signal environments. Additionor subtraction of codes can be performed prior to the accumulation in the correlator inorder as in [Mattos 2005]. Equally the codes may be individually correlated and213Single chip GPS and Giove-A receiversummed after the accumulation process.
Both techniques require knowledge of thepolarity of the secondary code of the C components in order to recover all oftransmitted power and decode the navigational data bit. In this research we haveavoided using code combinations because applications for patent protection have beenfiled for these techniques. Avoiding code combinations also enabled comparisonbetween the single-chip receiver and the Septentrio Galileo experimental test receiver(GETR) [De Wilde et al 2004].The single chip receiver was configured to receive the Galileo E1 B data modulatedsignal component. The B component has a chipping rate of 1.023 MHz and a codelength of 4092 chips, which results in a 4ms repeat period or code epoch.
The codeepoch is also equal to the symbol period of the navigational data, which is thereforemodulated at a rate of 250 sps. The current Galileo E1 specification defines codes,which must be stored in receiver memory. These memory codes designed to providebalanced codes with low cross-correlation values.
However, the codes transmitted byGiove-A are based on Gold codes, which are generated by shift register polynomials(see Figure 3-4). These codes were designed to be representative of Galileo signalsbut were not intended for public and have not been officially released. However,researchers and engineers first in the USA at Stanford University and in [Psiaki et al2006] have managed to decode the transmitted PRN code sequences generallythrough the use of high-gain antennas.. Since then, researchers in France have alsomanaged to decode the Giove-A PRN code. Table 10-1 shows the publicly knowngenerator polynomial for the Galileo E1 signal transmitted by Giove-A.Table 10-1, Galileo E1 PRN codes on Giove-ASignalCode lengthGeneratorStart valueSecondarySecondary codems (chips)polynomial(hexadecimal)code lengthbit pattern(hexadecimal)12(hexadecimal)E1 B4 (4092)26B131E11B831-E1 C8 (8184)201B26B11983257015B2214Single chip GPS and Giove-A receiverFor simplicity we maintain the simple serial search.
When integrating across a 1mscode period (GPS C/A) the approximate search time is 20.46sec (±5 kHz search)assuming half-chip steps in delay and 500 Hz frequency bins. Integrating across a4 ms code period (Galileo E1 B) results in an equivalent search time of 1309 secondsor 21.8 minutes. Clearly this is an unacceptable length of time to search for thesignal.
A solution is to short-cycle the code and correlate a 1 ms section of the codefor acquisition. When sub-dividing into four 1ms chunks, four code shifts aresearched every code epoch (4 ms). Therefore, the search time is only increased by thenumber of chips to be searched, a factor of 4 compared to 1ms integration. Shortcycling the code results in an approximate search time of 81.84 seconds or 1.364minutes, which is acceptable for most applications.
If further reductions in searchtime are required the receiver must devote additional correlation channels to thesearch or implement an FFT search algorithm.Table 10-2, Search timesSignalCodeIntegrationNumber ofNumber ofSearch timeperiodperiodhalf-chip codefrequency bins(s)(s)(s)shifts(±5kHz(A)requiredsearch)(B)(A××B××C)(C)GPS C/A0.0010.00120461020.46Galileo E1 B0.0040.0048184401309Galileo E1 B0.0040.00181841081.84Estimation of the quality of the Giove-A signal has been carried out under ESAcontract using the Septentrio GETR receiver. This receiver estimates the separatecarrier to noise densities for the Galileo E1 B and C components.
The results ofmonitoring the receiver-estimated carrier to noise against the elevation of the Giove-Asatellite were presented in [Falcone et al 2006].215Single chip GPS and Giove-A receiverOn the 26th of June 2006, the SSTL rooftop antenna was used to receive the Giove AE1 B signal with the single-chip receiver. Figure 10-8 shows the elevation andazimuth of the satellite pass on June 26th derived from STK simulation.18080160ElevationAzimuth140Elevation (degrees)60120501004080306020Azimuth (degrees)704023:0022:4522:3022:1522:0021:4521:3021:1521:0020:4520:3020:1520:0019:4519:3019:1519:0018:4518:3018:1518:0017:4517:3017:1517:0016:45016:30016:152016:0010Time (hours:mins)Figure 10-8, Elevation and azimuth of the Giove-A pass on the 26th of June 2006The signal was successfully acquired at 19:57, the receiver estimated carrier to noisedensity as 45 dB-Hz. This figure agrees well with the results from the SeptentrioGETR receiver.
The receiver was configured to track the phase of the incomingcarrier, the IQ plot taken from Giove-A tracking is shown in Figure 10-9.Figure 10-9, I/Q plot taken from Giove-A while in PLL mode216Single chip GPS and Giove-A receiverThe receiver successfully acquired and tracked the Giove-A BOC(1,1) signal withboth DE BOC receiver technique and the BJ algorithm.
Both tracking schemes wereable to maintain lock on the central peak of the BOC(1,1) correlation.The single chip receiver provides SSTL with a compact flexible correlator andprocessor architecture for future missions. The receiver has been demonstrated withof both GPS PSK signals and the Giove-A BOC(1,1) signals. As part of their ongoinginvestment in this research SSTL are currently adapting the single chip receiverdesign to track the Giove-A BOC(15,2.5) signal. This demonstration is intended toprove the advantages of the DE BOC receiver for signal with significant distortion(asymmetry).21711 Discussion, conclusions and future workThis chapter provides a synopsis of the contributions made by this research andsuggestions on areas of future work in this area. Firstly the contributions aredescribed, which due to the strong practical emphasis of this research containindustrial contributions as well as academic.
Following this the suggestions for futurework are given, which are also contrived for both industry and academia.11.1 Academic contributionsThe core academic contributions of this research are based around the reception ofBOC modulated signals.In Chapter 4 the advantages and disadvantages of BOC modulated signals are detailedand analyzed.
A simply derived yet accurate model for the theoretical timing jitter ofBOC systems is given. This allowed derivations for the cosine BOC timing jitter,which have not yet been produced in the literature. A fair comparative analysis of thetiming jitter and multipath error performance between BOC and PSK modulations isgiven. This analysis differs significantly from those given in the literature comparingeach signal from the perspective of occupied bandwidth of the signal, rather than apredetermined signal specification. Although there are many political pressures andinfluences involved in GNSS signal definition, it is important to realize thatconsidering the occupied bandwidth of the signal BOC modulation does not providesignificant performance benefits compared with PSK.
BOC modulation provides onlya small benefit in timing jitter (3 dB maximum), potentially worse multipathperformance and can introduce acquisition and tracking problems for the receiver.The fundamental advantage of BOC modulation is to enable spectral separationbetween signals within an allocated spectral bandwidth. BOC modulation thereforeallows efficient use of spectral bandwidth while minimizing interference.Chapter 5 details the theory of search, acquisition and tracking of PSK and BOCmodulated signals.
Although the theory for PSK systems is well established, a fullcomparative analysis of BOC search and tracking techniques has not yet been218Discussion, conclusions and future workproduced. This chapter draws on the techniques proposed in the literature for BOCsearch and tracking and provides a comparative analysis considering bothperformance measures and the receiver hardware impact. The sub-carriercancellation technique [Heiries et al 2004] is found to be the leading BOC searchtechnique, considering search performance, hardware impact and compatibility withBOC tracking techniques. The bump-jumping (BJ) algorithm [Fine and Wilson 1999]is found to deliver the maximum timing precision achievable from the BOCmodulation and is the most efficient in terms of hardware resources.
The single sideband tracking technique [Martin et al 2003] and smooth multiple gate discriminator[Bello and Fante 2005] provide the most robust acquisition and tracking performancebut are less efficient with hardware resource and have severely degraded timingprecision.The issue of integrity of the BJ algorithm in presence of signal distortion andmultipath inference has been analyzed. The results show that both distortion andmultipath inference can compromise the performance and reliability of receivers usingthe BJ algorithm. This approach to solving the BOC ambiguity problem is thereforenot suitable for high-integrity applications.From Chapter 5 we identify two distinct types of BOC tracking scheme. The firsttype preserves the full timing precision but has false-lock conditions resulting inpotentially degraded acquisition performance and unreliable operation.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
