Research Article An Innovative Architecture of UTC GPS/INS ...
Transcript of Research Article An Innovative Architecture of UTC GPS/INS ...
![Page 1: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/1.jpg)
Research ArticleAn Innovative Architecture of UTC GPS/INS System withImproved Performance under Severe Jamming
Xueyun Wang, Jingjuan Zhang, Wei Wang, and Pengyu Gao
School of Instrument Science and Opto-Electronics Engineering, Beihang University, Beijing 100191, China
Correspondence should be addressed to Jingjuan Zhang; [email protected]
Received 17 January 2014; Accepted 12 February 2014; Published 19 March 2014
Academic Editor: Guanghui Wen
Copyright ยฉ 2014 Xueyun Wang et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Ultratightly coupled (UTC) architecture is believed to be the best architecture for Global Positioning System (GPS) and InertialNavigation System (INS) integration system due to the advanced data fusion strategy and effective mutual assistance between thesubsystems. However the performance of UTC GPS/INS system will be degraded by severe jamming interference, especially whenlow-grade inertial measurement unit (IMU) is used. To solve this problem an innovative architecture of UTC GPS/INS system isproposed. Since GPS receiverโs antijamming ability is closely related to tracking loop bandwidth, adaptive tracking loop bandwidthbased on the fuzzy logics is proposed to enhance antijamming ability for GPS receiver. The bandwidth will be adapted througha fuzzy logic controller according to the calculated carrier to noise intensity ratio (๐ถ/๐0). Moreover, fuzzy adaptive integrationKalman filter (IKF) is developed to improve estimation accuracy of IKF when measurement noises change. A simulation platformis established to evaluate the innovative architecture and results demonstrate that the proposed scheme improves navigationperformance significantly under severe jamming conditions.
1. Introduction
GPS/INS system is the best solution for many navigationapplications and has been a research hotspot since the dayit was proposed [1โ3]. By the integration level, integrationsystems can be classified into three types: loosely, tightly,and ultratightly (also called deeply). Authors differ on whatexactly constitutes each of the three couplings [4โ6]; thedefinitions used here are closest to those given by Gao in[5]. If the GPS receiver is assisted by INS, the GPS/INSsystem is defined as ultratightly coupled system, in whicheither pseudorange (PR)/pseudorange-rate (PRR) or inphase(๐ผ)/quadrature (๐) are taken as measurements by IKF. Due tothe advanced data fusion strategy and enhanced navigationperformance, UTC GPS/INS system becomes a superiorsolution to robust and precise navigation in severe jammingenvironments. Howeverwhen low-grade inertial sensors (i.e.,gyroscopes and accelerometers), like microelectromechan-ical system (MEMS) devices, are used in UTC GPS/INSsystem, the tracking loop bandwidth of GPS has to bewide enough to calibrate sensor errors effectively [7], whichweakens the jamming resistance ability at the same time.
Therefore the antijamming property is an important issue forMEMS-based UTC GPS/INS system.
Some investigations have been conducted on this subject.Gao proposed another new architecture of UTC system witha cooperated tracking loop (COOP) inside of traditionalGPS tracking loops [5]. A method that adapts bandwidth ofdelay locked loop (DLL) to carrier to noise intensity ratio(๐ถ/๐0) was proposed to handle high interference or weaksignal environment and it indicated improvement in antijammargin over an INS/GPS with fixed bandwidths [7], butcarrier tracking loop, that is, phase locked loop (PLL), is notconsidered. Besides the adaptation is realized by a quadraticpolynomial which is derived by experiments and experiences.Ohlmeyer proposed a UTC GPS/INS system featured by abank of prefilters to estimate code delay error and theDopplerfrequency error for each satellite. This design was alsoconsideredmore robust to jamming and dynamics [8]. Effectsof inertial measurement unit (IMU) on weak signal trackingwere also analyzed. Research was conducted on performanceenhancement for ultratightly coupled GPS/INS system usinga fuzzy adaptive strong tracking unscented Kalman filter,verifying the effectiveness of fuzzy logic adaptive control
Hindawi Publishing CorporationDiscrete Dynamics in Nature and SocietyVolume 2014, Article ID 185618, 8 pageshttp://dx.doi.org/10.1155/2014/185618
![Page 2: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/2.jpg)
2 Discrete Dynamics in Nature and Society
(FLAC) [9]. Similar conclusions about advantages of FLACwere also drawn in [10โ12].
Aimed at improving navigation performance undersevere jamming and changing noise conditions, an innovativearchitecture of UTC GPS/INS system is proposed. Firstly theFLAC is applied to adjust PLL bandwidth according to GPSsignal noise intensity which is measured by ๐ถ/๐0. AdaptivePLL bandwidth will enhance the antijamming ability of theGPS receiver. Moreover FLAC is also used in the designof adaptive IKF in which the measurement noise model istuned online according to the innovation covariance. Thestate estimation accuracy of adaptive IKF is improved if themeasurement noises change. The proposed architecture isassessed by simulations conducted through a self-developedsimulation platform.
2. Problem Description
As PLL endures most of the noise interference and dynamicstress, DLL is generally locked in if PLL is locked in, somost attention is paid to the vulnerable PLL tracking. Forrobust and precise navigation, PLL should always be lockedin. Generally severe jamming and high dynamics are thecauses that force PLL to lose track. Narrowing the trackingloop bandwidth will improve the antijamming ability butwill lose track more easily under high dynamics due tothe large Doppler shifting caused by vehicle maneuver. InUTC GPS/INS system, as most of the Doppler shifting iscompensated by INS aiding information, PLL tracking loopsonly need to eliminate the residual Doppler shifting mainlycaused by errors of INS aiding information. The contra-dictory requirement for tracking loop bandwidth is muchrelieved. However, when low-grade MEMS IMU is used, theerrors of aiding the Doppler information are considerable.Therefore, PLL bandwidth has to be wide enough to handleaiding errors and calibrate MEMS sensors as well. To obtainboth good high-dynamic property and strong antijammingability, PLL bandwidth needs to be adapted according to thenoise intensity. Nevertheless, although adaptive bandwidth isemployed, severe jamming could still cause obvious changesto the characteristics of GPS outputs (i.e., IKF measure-ments), so fixed measurement noise model for IKF describeda priori is inappropriate. Hence an adaptive IKF in whichmeasurement noise covariance matrix can be tuned in real-time is needed.
3. Methodology
3.1. Fuzzy Logic Adaptive PLL Bandwidth. When PLL is welllocked in, the errors of PLL-derived pseudorange-rate aremainly thermal noise whose standard deviation is affected byits bandwidth and ๐ถ/๐0. The relationship can be expressedas follows [13, 14]:
std ( ฬ๐) = ๐
โ2๐๐๐ฟ๐๐ท
โ๐ต๐
๐ถ/๐0[1 +
1
2๐coh โ ๐ถ/๐0], (1)
where std( ฬ๐)is standard deviation of PRR, c is light speed,๐๐ฟis carrier frequency, ๐
๐ทis the Doppler integration time,
10 20 30 40 50 600
0.2
0.4
0.6
0.8
1
MD HGLW
C/N0 (dB-Hz)
Figure 1: ๐ถ/๐0 fuzzification.
Table 1: Fuzzy rules for adaptive PLL bandwidth.
If ๐ถ/๐0 is Then bandwidth isLow (LW) Narrow (NR)Medium (MD) Normal (NM)High (HG) Wide (WD)
๐cohis coherent integration time for ๐ผ and ๐, and ๐ต๐is PLL
bandwidth.All the parameters except for ๐ต
๐and ๐ถ/๐0 are constants
for a specific GPS receiver, so if ๐ถ/๐0 decreases, ๐ต๐has
to decrease accordingly to constrain PRR error within acertain level. If conversely ๐ถ/๐0 increases, ๐ต
๐ought to
increase to obtain better performance under high dynamicsand better calibration for inertial sensors.However, in fact thebandwidth should not be adjusted thoroughly according to(1). The first reason is that PLL are not always well locked in.Secondly, the corresponding bandwidth in severe jammingcould be too much narrow that PLL easily loses track dueto vehicle maneuvers or errors of INS aiding information.As basic adaptation principles are clear, fuzzy adaptive logiccontrol is preferred for the adjustment of PLL bandwidth.
On the basis of the analyses above an adaptive PLLbandwidth based on fuzzy logic is designed. The fuzzy logiccontrollerโs input is ๐ถ/๐0 (in unit of dB-Hz) and output isPLL bandwidth (in unit of Hz). The membership function(MF) for fuzzification and defuzzification is shown in Figures1 and 2. Note that the bandwidth starts at 0.1 Hz, not 0. Fuzzyrules are listed in Table 1.
The designed fuzzy logic controller belongs to the type ofMamdani.Max-Min inferencemethod is applied to define theresults of fuzzy rules and centroid defuzzification is employedto extract outputs.
A vital issue arises on how to determine the input, ๐ถ/๐0.Here variance summing method (VSM) is used to calculate๐ถ/๐0 using the sampled GPS signals after downconverting[15, 16].
![Page 3: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/3.jpg)
Discrete Dynamics in Nature and Society 3
0 1 2 3 4 5 60
0.2
0.4
0.6
0.8
1
Bandwidth (Hz)
NR NM WD
Figure 2: PLL bandwidth defuzzification.
The integrate-and-dump results at sampled time i,inphase ๐ผ
๐and quadrature ๐
๐, can be divided into signal and
noise:
๐ผ๐= ๐ผ๐
๐+ ๐ผ๐
๐=๐ด๐๐ธ๐ท๐
โ2โ cos (๐ฟ๐
๐) + ๐ผ๐
๐,
๐๐= ๐๐
๐+ ๐๐
๐=๐ด๐๐ธ๐ท๐
โ2โ sin (๐ฟ๐
๐) + ๐๐
๐,
(2)
where ๐ด and ๐ท๐are signal amplitude and navigation data,
respectively, and ๐ฟ๐๐denotes PLL phase error. ๐
๐ธis the
integration number yielding ๐๐ธ= ๐Int/๐๐ where ๐Int is
coherent integration interval and ๐๐ is sample interval.
If PLL is in good lock-in condition, which means ๐ฟ๐๐is
fairly small, ๐ผ๐and ๐
๐approximate to
๐ผ๐=๐ด๐๐ธ๐ท๐
โ2+ ๐ผ๐
๐,
๐๐= ๐๐
๐.
(3)
Since binary phase shift keying (BPSK) technique is usedinGPS signalmodulation, navigation data yields๐ท
๐= ยฑ1. Set
a new variable by the following equation:
๐๐= ๐ผ2
๐+ ๐2
๐= (๐ด๐๐ธ๐ท๐
โ2)
2
+๐ด๐๐ธ๐ท๐
โ2โ ๐ผ๐
๐+ (๐ผ๐
๐)2+ (๐๐
๐)2.
(4)
Assume ๐ผ๐๐and๐๐
๐are both zero-mean noise and have the
same standard deviation ๐ฟ๐๐, so the mean and variance of ๐
๐
are as follows, respectively:
๐ง =(๐ด๐๐ธ)2
2+ 2๐ฟ2
๐๐, (5)
๐ฟ2
๐ง= 2(๐ด๐
๐ธ)2๐ฟ2
๐๐+ 4๐ฟ4
๐๐. (6)
Thus equations of the average carrier power ๐๐and average
noise power ๐๐take the following forms:
๐๐=(๐ด๐๐ธ)2
2= โ(๐ง)
2โ ๐ฟ2๐ง,
๐๐= 2๐ฟ2
๐๐= ๐ง โ โ(๐ง)
2โ ๐ฟ2๐ง.
(7)
Finally the carrier to noise density ratio C/N0 can bedetermined by the following formula:
๐ถ
๐0
= 10log10(๐๐
๐๐
โ 1
๐Int)
= 10log10(โ(๐ง)2โ ๐ฟ2๐ง๐
๐ง โ โ(๐ง)2โ ๐ฟ2๐ง
โ 1
๐Int).
(8)
3.2. Fuzzy Logic Adaptive IKF. Details about standard algo-rithm of the Kalman filter can be found in many papers[17]. Applications based on KF not only demand correct statetransition matrix and measurement matrix but also presumecomplete and accurate a priori information about the processnoise and measurement noise. If any of the requirements arenot met, KF may propagate suboptimal estimates or evendiverge. Previously substantial work has been done to obtaincorrect state transition matrix and measurement matrix forGPS/INS system,making it not a problemat all.Moreover, thestochastic statistics of the process noise, which mainly refersto inertial sensorsโ white noise and the driving noise (alsowhite) of the Gauss-Markov process, can be determined withmethods such as the Allen variance. Besides, characteristicsof process noises do not change greatly unless faults happen.However it is different for the measurement noises. Thecovariance of measurement noise cannot be constrainedwithin a certain level under various jamming conditions,even though PLL bandwidth has been adapted. Once mea-surement noise changes, its covariance matrix needs to becorrected so that divergence caused by noise model inaccu-racy could be avoided and better state estimates could bemade. That is the exact purpose of adaptive IKF. There areseveral ways to accomplish this goal and they are classifiedinto four categories, namely, Bayesian, maximum likelihood,correlation (autocorrelation), and covariance matching [18].The fourth approach implemented by fuzzy logic control isselected for the innovative UTC GPS/INS system and thealgorithms are as follows.
Innovation, also named as residual, is critical informationfor the Kalman filter. It is defined as the discrepancy betweenactual measurements and predicted measurements:
๐๐= ๐ง๐โ ๐ป๐๐ฅโ
๐. (9)
In the equation, k is the time sequencemark and ๐ง๐repre-
sents the actualmeasurements.๐ป๐is themeasurementmatrix
and ๐ฅโ๐is the formerly estimated state vector.Theoretically in
normal conditions, the innovation is zero-mean white noise
![Page 4: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/4.jpg)
4 Discrete Dynamics in Nature and Society
Table 2: Fuzzy rules for adaptive IKF.
If covariance ratio is Then scale factor is
Low (LW) Smaller (SM)
Medium (MD) Normal (NM)
High (HG) Bigger (BG)
0.6 0.8 1 1.20
0.2
0.4
0.6
0.8
1
Covariance ratio
MD HGLW
Figure 3: Fuzzification for covariance ratio.
whose covariance is related to the covariance of process noiseand the measurement noise, yielding
๐ถ๐๐= ๐ป๐(๐น๐/๐โ1๐๐โ1๐น๐
๐/๐โ1+ ๐๐โ1)๐ป๐
๐+ ๐ ๐โ1, (10)
where ๐น๐/๐โ1
is state transition matrix from time ๐ โ 1 to k,๐๐โ1
is covariance matrix of estimation errors at time ๐ โ 1,๐๐โ1
is covariance matrix of process noise, ๐ ๐โ1
is covariancematrix of measurement noise, and ๐
๐โ1and ๐
๐โ1are set
as diagonal matrices by the assumption that all the processnoises and measurement noises are uncorrelated.
The actual innovation covariance is defined as the averag-ing covariance of the๐ newest sampled innovations and thenumber๐ is decided by the length of moving window:
๐๐๐=1
๐
๐
โ๐=๐0
๐๐โ ๐๐
๐, (11)
where ๐0= ๐ โ ๐ + 1 is the start of the moving window. N is
applied empirically to provide smoothing and the exact valueemployed in this paper is 10.
If KF proceeds normally, the theoretical and actualcovariance of innovations should be identical; otherwisethe covariance matrix of measurement noise ๐
๐โ1must
be impropriate provided that ๐ป๐, ๐น๐/๐โ1
, and ๐๐โ1
are allcorrect. So the fundamental principle of fuzzy logic adaptiveIKF is to adjust measurement noise covariance matrix aswhat it should be according to the discrepancy between thetheoretical and actual covariance of innovations.
0 0.5 1 1.5 20
0.2
0.4
0.6
0.8
1
Scale factor
SM NM BG
Figure 4: Defuzzification for scale factor.
The input for the fuzzy logic controller is the ratio ofthe actual innovation covariance and theoretical innovationcovariance:
๐ผ๐= diagonal(
๐ถ๐๐
๐๐๐
) . (12)
Only the diagonal elements are reserved to be the inputs.As the fuzzy logic controller is a single-input-single-outputsystem, each of the diagonal elements in ๐ผ
๐has to go
through the fuzzy logic controller one after another toachieve a corresponding result. Finally all the results willform the controllerโs output, which is also a diagonal matrixcontaining scale factors for the correction of measurementnoise covariance matrix. For the next integration navigation,the corrected ๐
๐is shown as follows:
๐ ๐= ๐ฝ๐โ ๐ ๐โ1, (13)
where ๐ฝ๐is the scale factor matrix.
The input and output MFs of the fuzzy logic controllerfor adaptive IKF are shown in Figures 3 and 4. Fuzzy rulesare listed in Table 2. All configurations about the fuzzy logiccontroller are identical with those of PLL bandwidth.
Since fuzzy logic adaptive controller is designed andapplied to improve the antijamming ability of IKF, themathematic model of the IKF in this innovative architectureis different from conventional IKF.The mathematic model ofnew IKF can be divided into two parts and they are detailedas follows.
![Page 5: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/5.jpg)
Discrete Dynamics in Nature and Society 5
DLL
IF data
โโ
โ
PLL tracking
Integrateand dump
DLL tracking
Ii, Qi
calculation
NCO
Phasediscriminator
๐ฟ๐
Fuzzyadaptive
bandwidth
Loopfilter
Bn
fDfL
INS
aidingINS
IMU data
๐GPS๐INS
๏ฟฝฬ๏ฟฝGPS
๏ฟฝฬ๏ฟฝINS
ฮ๐
ฮ๏ฟฝฬ๏ฟฝ
Fuzzyadaptive
IKF
C/N0
Figure 5: Scheme diagram of the innovative architecture of the UTC GPS/INS system.
Part 1.The algorisms of the Kalman filter are
๐พ๐= ๐โ
๐๐ป๐
๐(๐ป๐๐โ
๐๐ป๐
๐+ ๐ ๐) ,
๐๐= ๐ง๐โ ๐ป๐๐ฅโ
๐,
๐ฅ๐= ๐ฅโ
๐+ ๐พ๐๐๐,
๐ฅโ
๐+1= ฮฆ๐๐ฅ๐,
๐๐= (๐ผ โ ๐พ
๐๐ป๐) ๐โ
๐,
๐โ
๐+1= ๐น๐+1/๐๐๐๐น๐
๐+1/๐+ ๐๐.
(14)
The meanings of the symbols that are not explained in thispaper could be found in [17].Part 2.The adjustment of measurement covariance matrix is
๐ถ๐๐= ๐ป๐๐โ
๐+1๐ป๐
๐+ ๐ ๐,
๐๐๐=1
๐
๐
โ๐=๐0
๐๐โ ๐๐
๐,
๐ผ๐= diagonal(
๐ถ๐๐
๐๐๐
) ,
๐ฝ๐+1= FLAC (๐ผ
๐) ,
๐ ๐+1= ๐ฝ๐+1โ ๐ ๐.
(15)
After the adjustment of measurement covariance matrix,algorisms of KF will be conducted again with the newmeasurement covariance matrix. Through correcting ๐
๐the
fuzzy logic controller improves the antijamming ability of theconventional IKF so that the UTC INS/GPS system gains amuch better performance under severe jamming conditions.
The scheme diagram of the innovative architecture ofUTC GPS/INS system is shown in Figure 5.
4. Simulation and Analysis
4.1. Simulation Configurations. To verify the effectiveness ofthe proposed architecture, a simulation package is developed
GPSsatellitevelocityposition
GPSconstellation
simulator
Trajectory
simulator
IF datagenerator
IMU datagenerator
IF data IMU data
IF noise IMU noise
Vehicleattitudevelocityposition
GPSSDR IKF INS
โ
Navigationsolutions
Performanceevaluation
Figure 6: Simulation platform compositions.
and its compositions are shown in Figure 6. First of all, thetrajectory simulator (TS) gives a specific trajectory whichincludes vehicle attitude, velocity, and positions, based onwhich the IMU data generator calculates the actual accel-eration and all the angular rates for the inertial sensortriads in body frame. Secondly the intermediate frequency(IF) data generator takes the actual trajectory from TS, aswell as all kinematics parameters of GPS satellites fromthe GPS constellation simulator, to produce the IF data forGPS software defined receiver (SDR). At last IKF takes PRand PRR from GPS SDR and navigation data from INS toperform integration navigation. The innovative scheme willbe evaluated by comparing its navigation solutions with theactual trajectory. The GPS SDR is developed on the basis ofthe fundamental work previously done by Kai Borre and soforth [19] and it has been verified by processing real GPSIF data to obtain correct navigation results. The IF datagenerator and GPS constellation simulator are then certifiedby the verified GPS SDR.
![Page 6: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/6.jpg)
6 Discrete Dynamics in Nature and Society
118118.02
118.04118.06
40
40.05
2000
4000
6000
Longitude (deg)Latitude (deg)
Hei
ght (
m)
Figure 7: 3-dimensional position.
The 3-dimensional position of the 50-second simulatedtrajectory is shown in Figure 7 and magnitudes of velocityand acceleration are shown in Figure 8.
A four-satellite GPS constellation with a nominal GDOPof 2.52 is used by the single frequency SDR whose integrate-and-dump operation rate and PR/PRR update rate are both1000Hz. Second-order tracking loops are engaged for DLLandPLL.Thebandwidth ofDLL is 0.1Hzwhile the bandwidthof PLL will be discussed specially in the next section. Theerror characteristics of IMU are basically determined inaccordance with generic MEMS sensors and the essentialparameters are listed in Table 3. The IMU data samplinginterval is 1 millisecond while the integrated navigation isoperated every 100 milliseconds. Although Earth-Centered-Earth-Fixed (ECEF) coordinates are used to calculate PR andPRR and so forth, navigation results are finally provided ingeographic coordinates. As each satellite channel is indepen-dent and equivalent, only the loop tracking and PRR resultsof satellite number 3 (pseudorandomnumber 3) are displayedto draw conclusions in the following sections.
4.2. Evaluation of the Proposed Architecture. A wide rangeof jamming is artificially simulated from 20 s to 38 s toevaluate the performance of fuzzy adaptive PLL bandwidthand fuzzy adaptive IKF. Jamming forces ๐ถ/๐0 to decreasesharply from 45 dB-Hz to 25 dB-Hz at 20 s and keeps it at25 dB-Hz until 30 s. Then jamming gradually weakens and๐ถ/๐0 recovers to 45 dB-Hz at an increment of 2.5Hz-dBevery second. The true ๐ถ/๐0, calculated ๐ถ/๐0, and theadapted PLL bandwidth are shown in Figure 9. For standardUTC GPS/INS system the PLL bandwidth is fixed at 3Hz.However for the innovative one the bandwidth is adapteddown to about 0.8Hz according to the calculated ๐ถ/๐0whose accuracy is better than 3 dB-Hz. As a result of thenarrowed PLL bandwidth, the PRR accuracy of innovativearchitecture has improved significantly compared with thestandard one, from 3.02m/s to 0.85m/s in terms ofmaximumerror as shown in Figure 10. It is worth pointing out that ifthe PLL bandwidth of standard architecture is set at 0.8Hz,
0 10 20 30 40 500
100
200
300
Time (s)
Velo
city
(m/s
)
0 10 20 30 40 500
5
10
Time (s)
Acce
lera
tion
(g)
Figure 8: Acceleration and velocity.
Table 3: IMU error characteristics.
Error sources Error valuesGyroscopes Accelerometers
Bias 60 deg/h 10mgScale factor 200 ppm 200 ppmRandom walk 0.2 deg/โh 0.12m/s/โh
PLLwill lose track during the initial speedup because the highacceleration causes great dynamic pressure that the narrowedbandwidth is not able to handle.Therefore, the fuzzy adaptivePLL bandwidth is a more intelligent and robust solution forsevere jamming interference.
Nevertheless, in spite of being improved, the PRR noisestill increases compared with that when jamming is notinvolved, from 0.10m/s to 0.23m/s in terms of standarddeviation, so the covariance matrix of measurement noiseshould be corrected accordingly. The PRR noise standarddeviations that are used to form the measurement noisecovariance matrices of standard IKF and innovative oneare shown, respectively, in the upper plot of Figure 11. Theratios of real and theoretical innovation variances of satellitenumber 3 are shown in unit of dB in the lower plot ofFigure 11. They indicate how well the measurement noisemodel corresponds to reality. In normal situations, the ratiosshould be around 1 (0 dB). At first the PRR noise standarddeviations of both architectures are set as 0.06m/s, which is alittle bit lower than the true value 0.10m/s.Themeasurementnoise model of standard IKF is not modified. However, inthe adaptive IKF, the PRR noise standard deviation is quicklyadjusted by the IKF fuzzy logic controller to about 0.1m/suntil the jamming happens. After jamming disappears at38 s, PRR noise model is tuned back to 0.1m/s again. Forthe standard architecture, the ratio of real and theoreticalinnovation variance ranges from โ5 dB to 15 dB, but for theinnovative architecture, the ratio is kept around 0 dB all thetime, implying that the measurement noise model is correctand accurate.
![Page 7: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/7.jpg)
Discrete Dynamics in Nature and Society 7
20
25
30
35
40
45
50
10 20 30 40 500.5
1
1.5
2
2.5
3
3.5
Time (s)
Band
wid
th (H
z)
BandwidthTrue C/N0
C/N
0 (d
B-H
z)
Computed C/N0
Figure 9: ๐ถ/๐0 and adapted bandwidth.
10 20 30 40 50Time (s)
PRR
erro
r (m
/s)
StandardInnovative
โ1
โ0.5
0
0.5
1
1.5
Figure 10: PRR error comparison in jamming.
The east, north, and up velocity errors of both architec-tures are given in Figure 12. Clearly the innovative architec-ture has a much improved navigation performance undersevere jamming conditions than the standard one. In termsof maximum error, the velocity accuracy of the innovativearchitecture during 20 sโ38 s is 0.73m/s while the standardone only gets to 2.05m/s. In terms of standard deviation,the performances of innovative and standard schemes are0.10m/s and 0.31m/s, respectively, during the same period oftime.
5. Conclusion
An innovative architecture of UTC GPS/INS integratedsystem is proposed to reinforce the antijamming ability.The architectural features include adaptive PLL bandwidth
0 10 20 30 40 500.05
0.1
0.15
0.2
PRR
noise
Std
(m/s
)
10 20 30 40 50Time (s)
Time (s)
Inno
vatio
n va
r rat
io (d
B)
StandardInnovative
โ5
0
5
10
15
Figure 11: Performance of adaptive and standard IKFs.
0 5 10 15 20 25 30 35 40 45 50
0 5 10 15 20 25 30 35 40 45 50
0 5 10 15 20 25 30 35 40 45 50Time (s)
Time (s)
Time (s)
StandardInnovative
โ2
โ1
0
Vu
erro
r (m
/s)
Vn
erro
r (m
/s)
Ve
erro
r (m
/s)
โ1
โ0.5
0
0.5
โ0.50
0.51
1.5
Figure 12: Velocity error comparisons.
and adaptive IKF, both of which are implemented throughfuzzy logics. Firstly, PLL bandwidth is adjusted according tocalculated ๐ถ/๐0, so the antijamming ability of GPS receiveris reinforced. Secondly the measurement noise covariancematrix of IKF is tuned online according to innovationcovariance, so the estimation accuracy is improved when themeasurement noises change. The algorithms are detailed anda simulation platform is developed to verify the effectivenessof the proposed architecture. Results presented in this paperdemonstrate that the innovative architecture improves thenavigation performance significantly under severe jamming
![Page 8: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/8.jpg)
8 Discrete Dynamics in Nature and Society
conditions. Further work essentially lies in the algorithmverification with real data and the implementation of thesystem by hardware. By then the practical usage of fuzzylogic and additional computation burden should be dealt withdedicatedly.
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper.
Acknowledgment
This research was sponsored by the National Natural ScienceFoundation of China (Grant no. 61079017). The authors aregrateful for the support it provided.
References
[1] L. Semeniuk and A. Noureldin, โBridging GPS outages usingneural network estimates of INS position and velocity errors,โMeasurement Science and Technology, vol. 17, no. 10, pp. 2783โ2798, 2006.
[2] Q. Zhang, X. Niu, Q. Chen, H. Zhang, and C. Shi, โUsingAllan variance to evaluate the relative accuracy on differenttime scales of GNSS/INS systems,โ Measurement Science andTechnology, vol. 24, Article ID 085006, 2013.
[3] L. Zhao and Q. Wang, โDesign of an attitude and headingreference system based on distributed filtering for small UAV,โMathematical Problems in Engineering, vol. 2013, Article ID498739, 8 pages, 2013.
[4] M. Lashley, D. M. Bevly, and J. Y. Hung, โAnalysis of deeplyintegrated and tightly coupled architectures,โ in Proceedings ofthe IEEE/ION Position, Location and Navigation Symposium(PLANS โ10), pp. 382โ396, May 2010.
[5] G. Gao and G. Lachapelle, โA novel architecture for Ultra-TightHSGPS-INS integration,โ Journal of Global Positioning System,vol. 7, pp. 46โ61, 2008.
[6] D. G. Egziabher, โGNSS Solutions: weighting GNSS observa-tions and variations ofGNSS/INS integration,โ InsideGNSS, vol.2, no. 1, pp. 28โ33, 2007.
[7] P. D. Groves and D. C. Long, โCombating GNSS interferencewith advanced inertial integration,โ Journal of Navigation, vol.58, no. 3, pp. 419โ432, 2005.
[8] E. J. Ohlmeyer, โAnalysis of an ultra-tightly coupled GPS/INSsystem in jamming,โ in Proceedings of the IEEE/ION Position,Location, and Navigation Symposium, pp. 44โ53, April 2006.
[9] D. Jwo and C. Yang, โPerformance enhancement for ultra-tightGPS/INS integration using a fuzzy adaptive strong trackingunscented Kalman filter,โ Nonlinear Dynamics, vol. 73, pp. 377โ395, 2013.
[10] C. Hide, T. Moore, and M. Smith, โAdaptive Kalman filteringfor low-cost INS/GPS,โ Journal of Navigation, vol. 56, no. 1, pp.143โ152, 2003.
[11] D. Loebis, R. Sutton, J. Chudley, and W. Naeem, โAdaptivetuning of a Kalman filter via fuzzy logic for an intelligent AUVnavigation system,โ Control Engineering Practice, vol. 12, no. 12,pp. 1531โ1539, 2004.
[12] H. Bian, Z. Jin, and W. Tian, โStudy on GPS attitude deter-mination system aided INS using adaptive Kalman filter,โ
Measurement Science and Technology, vol. 16, no. 10, pp. 2072โ2079, 2005.
[13] X.-L. Wang and Y.-F. Li, โAn innovative scheme for SINS/GPSultra-tight integration system with low-grade IMU,โ AerospaceScience and Technology, vol. 23, pp. 452โ460, 2012.
[14] R. J. Landry, โNew technique to improve GPS receiver per-formances by acquisition and tracking threshold reduction,โin Proceedings of the RTO SCi International Conference onIntegrated Navigation Systems, pp. 24โ26, 1999.
[15] M. L. Psiaki, O. M. Akos, and J. Thor, โA comparison of directRF sampling and down-convert & sampling GNSS receiverarchitectures,โ in Proceedings of the ION GPS Conference, pp.1941โ1952, 2003.
[16] M. L. Psiaki, โBlock acquisition of weak GPS signals in a soft-ware receiver,โ in Proceedings of the 14th International TechnicalMeeting of the Satellite Division of the Institute of Navigation(ION GPS โ01), pp. 2838โ2850, 2001.
[17] M. Liu and F. Xiong, โA fuzzy adaptive GPS/INS integratednavigation algorithm,โ Procedia Engineering, vol. 15, pp. 660โ664, 2011.
[18] J. Ali, โStrapdown inertial navigation system/astronavigationsystem data synthesis using innovation-based fuzzy adaptiveKalman filtering,โ IET Science, Measurement and Technology,vol. 4, no. 5, pp. 246โ255, 2010.
[19] K. Borre, D.M.Akos,N. Bertelsen, P. Rinder, and S.H. Jensen,ASoftware-Defined GPS and Galileo Receiver: A Single-FrequencyApproach, Springer, Washington, DC, USA, 2006.
![Page 9: Research Article An Innovative Architecture of UTC GPS/INS ...](https://reader030.fdocuments.us/reader030/viewer/2022012418/61734b2840b16b07425fdab5/html5/thumbnails/9.jpg)
Submit your manuscripts athttp://www.hindawi.com
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttp://www.hindawi.com
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com
Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Stochastic AnalysisInternational Journal of