=Paper=
{{Paper
|id=Vol-3919/short34
|storemode=property
|title=Indoor Confined Space: GNSS Multi-Frequency Location Method based on Pursuit Principle
|pdfUrl=https://ceur-ws.org/Vol-3919/short34.pdf
|volume=Vol-3919
|authors=Zhang Heng,Wang Qing,Huang Lu,Li Yaning,Cheng Jianqiang,Liu Shiyuan
|dblpUrl=https://dblp.org/rec/conf/ipin/ZhangYHLCL24
}}
==Indoor Confined Space: GNSS Multi-Frequency Location Method based on Pursuit Principle==
https://ceur-ws.org/Vol-3919/short34.pdf
Indoor Confined Space: GNSS Multi-Frequency Location
Method based on Pursuit Principle
Zhang Heng1,2, Wang Qing1,2∗，Huang lu1,2,Li Yaning1,2,Cheng Jianqiang1,2,and Liu
Shiyuan1,2
1
The 54th Research Institute of China Electronics Technology Group Corporation
2
State Key Laboratory of Satellite Navigation System and Equipment Technology
Abstract
Aiming at the problem that it is difficult to carry out effective high-precision positioning in the complex
and limited indoor space environment, this paper proposes an indoor limited space GNSS multi-
frequency PSO (Particle Swarm Optimization) positioning method. This algorithm effectively solves the
positioning technical problems which caused by inaccurate indoor spatial signal ranging. further we
analyze the factors that affect the performance of the positioning algorithm. Compared to our team’s
existing KPI-based doppler positioning algorithm, the algorithm in this paper improves the problem of
inaccurate range measurement, which caused by frequent cycle slips of signal loss. Through multi-
frequency coordination and local optimization processing, the stability and continuity of indoor high-
precision positioning are improved. The simulation results of semi-measured data show that the accuracy
of the positioning algorithm in this paper is equivalent to that of the existing algorithm. When some
signals are temporarily lost and refilled, the performance of this algorithm is better than that of the
existing algorithm, and it has better continuity and robustness.
Keywords
indoor environment, pseudolite , dual frequency, navigation and positioning, high precision, continuity 1
1. Introduction
With the full operation of the national Beidou system and the construction of the national
comprehensive PNT (Positioning Navigation and Timing) network system, the construction of the
indoor location service network is becoming more and more urgent. Of course, the relatively
mature indoor positioning technology of the market [1-7]has two type, one is Ultra-wide band
(UWB)[8,9]. The current market share of the technology is increasing year by year, showing good
application prospects, but the technology still faces major bottlenecks in capacity and dedicated;
another technology is array Bluetooth technology[10-12], Meter-level positioning accuracy can
currently be achieved with AOA (Angle of arrival) technology, but further improvement is still
needed in the high-density layout problem.
Compared with the above technologies, pseudo-satellite technology has three advantages: (1)
flexible networking mode, fast networking in small scale and large scale space; (2) passive
positioning, no user capacity problem; (3) highly compatible with GNSS system, to meet the
seamless indoor and outdoor applications of the same terminal. Therefore, this paper still focuses
on the pseudo-satellite technology compatible with navigation satellites. At present, Waseda
University in Japan has conducted more extensive technical research on this direction[13-18],
Low-speed and high-precision positioning in the indoor environment can be realized through the
existing half-wavelength array pseudo-satellite technology, doppler and ambiguity repair
technology. Based on the technical inspiration, China electric 54 put forward aa method ,which
∗
Corresponding author.
†
These authors contributed equally.
13582161539@163.com (Z.H); 17733873190@163.com (W.Q); 18642720668@163.com (H.L) ;15631149037@163.com(L
Yaning);a243945274@163.com(C Jianqiang);rcc@bupt.cn(L Shiyuan)
© 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
does not rely on half wavelength indoor homologous array pseudo satellite technology. On this
basis, through the carrier phase application, implements the indoor carrier phase based on KPI
(Known point initialization) doppler high precision positioning algorithm, based on
electromagnetic map pseudo satellite fingerprint positioning technology, etc. It solves the problem
of high-precision positioning in the indoor non-test environment[19-23]。
Combine the existing technical achievements. Facing the problem of initial point assignment, a
combined location algorithm using multi-frequency constrained ranging and local optimization is
proposed. By integrating the relationship between probability statistics and frequency and signal
transmission characteristics, the problem of obtaining absolute ranging information in indoor
environment is solved, and the method of indoor absolute location method is given.
2. Signal features
The base station signals analyzed in this paper are accomplished by homologous design. First,
the time-frequency control module tamed the local clock in real time to ensure the clock output of
1PPS and reference clock 10 MHz consistent with GNSS. The reference clock realizes the
generation of multiple frequency signals required for signal generation through the clock
frequency division module. Under the control of 1PPS, the baseband signal modulation of each
frequency is completed in the baseband signal generation module, and finally the RF signal
combined circuit output is completed by the upconversion module. Since the signals between
multiple channels and between different frequency points are controlled by the same PLL, so all
signals have the same time-frequency characteristics. At the same time, for each frequency point
signal transmitted on the same channel, because the hardware transmission path of each signal is
the same, we can think that the device delay between the frequency point signals of the same
channel is the same, so the multiple frequency point signals of each channel have a fixed phase
relationship.The pseudo-satellite base station principle is shown in Figure 1.
Figure 1: The pseudo-satellite base station principle
3. GNSS MFCR Localizatin Algorithm
3.1. Multi-frequency joint ranging
Figure 2: The principle of distance measurement
As shown in Figure 2, the multi-frequency ranging principle is similar to the pursuit principle.
Assuming that on the circular track, A athletes run 1 within unit time t and B athletes run 2
within unit time t , then the AB distance varies 1 2 within unit time. Therefore, we can
judge how far athletes A and B each run according to the cumulative change of the distance
difference. According to this principle, we assume that the wavelengths of the two frequency points
are also respectively, so we give the equation between the propagation distance and wavelength of
the two frequency points is 1 and 2
1 1 N1 1 init,1 1
(1)
2 2 N2 2 init,2 2
In the formula: 1 and 2 are the propagation distance of two frequency signals; 1 and 2 are
the wavelength; N1 and N 2 are the cycle count; 1 and 2 are the carrier phase of one
cycle; 1 and 2 are the noise error value; init ,1 and init ,2 are two frequency point signal
transmission initial phase distance value, assumed here is 0, then the formula can be simplified into:
1 1 N1 1 1
(2)
2 2 N 2 2 2
Thus, the double-frequency difference value can be obtained as follows.
21 21 2 N2 1 N1 21 (3)
The pursuit principle is also available.
21 mod((2 1 ) * N2 , 1 ) (4)
Assuming that the resulting integer value is M, then the upper formula can be changed to
21 (2 1 ) N2 1 M (5)
1
P round ( ) , P is the number of cycles to walk. N 2 MP .Then can be obtained
2 1
1
21 (round ( ))
2 1
21 2
(2 1 ) P 1 (6)
N1
1
Finally, we get the distance value. 1 and 2
3.2. Data integrity strategy
The data fusion processing strategy is shown in Figure 3.
Figure 3: Data fusion processing strategy
3.3. Ranging fusion strategy
Assuming that when there are n frequency signals, in accordance with the selection Cn2 , then
there are n 1 combination, each frequency ranging values have n(n 1) 2 times ranging values.
Therefore, there are altogether num n(n 1) 2 ranging results for the current position at the
current moment. The distance between the transmitting channel and the receiving point can be
estimated by one-dimensional normal sample expectation.
num
ps j i , j num (7)
i 1
Among them, j is the j channel, ps j is the j channel signal to reach the receiving point of the
evaluation distance value, i , j is the j channel signal to reach the i frequency point of the
distance value.
3.4. The PSO location search
Based on the estimated speed of the maximum doppler value, and set the Doppler value as f d , then
the speed is v .
fd
v c (8)
f
The search radius is defined as
2*v
r (9)
iter
Where iter is the output rate of raw observation data per second of the receiver in hz
If the position xk 1 yk 1 zk 1 at the previous moment is given, the position at the current
moment k is searched for
x ( xk 1 , xk 1 r cos( ) sin( ))
y ( yk 1 , yk 1 r cos( ) cos( )) (10)
z ( zk 1 , zk 1 r sin( ))
The particle swarm optimization algorithm initializes the constants are c1 c2 r1 r2 and w . In the
three-dimensional search space, there are a total of M particles forming a group, the i th particle
space position is xi is the potential optimal solution of the optimization problem, the speed of the
particle is recorded vi ,the contemporary best position pi ,best , the historical best positio pbest , the n-
1 generation particle, the velocity and position of the n generation equation is
vin wvin 1 c1r1 ( pi ,best xin 1 ) c2 r2 ( pbest xin 1 )
(11)
xin xin 1 vin
Among them,the w is inertial weights; c1 and c2 are self-learning factor and group learning factor,
respectively, taking positive values; r1 and r2 are two random numbers varying within [0,1].
4. Simution Validation
4.1. Simulation Setting
The simulation test is set based on the existing array pseudo-satellite platform. Here, L1C, B1I
and B3I signals are selected for simulation. The signal characteristics are shown in the Table 1.
Table 1 The Signal Characteristics
name frequency (Mhz) tape width modulation mode PRN number
(Mhz)
L1C 1575.42 2.046 B PSK 1~8
B1I 1561.098 4.092 B PSK 6~13
B 3I 1268.52 20.46 B PSK 6~13
During the indoor application of pseudo satellite signals, frequent recatch usually occurs, which
will introduce cycle jump. Based on the existing phenomenon, the simulation conditions are set as
follows:
1. The simulation time is 100s, the output frequency is 2 hz, and the signal coordinate values
are set as shown in Table 2 below.
Table 2 pseudo-satellite coordinates
name X(m) Y(m) Z(m)
P L01 -0.17 14.67 3.44
P L02 1.07 14.31 6.47
P L03 1.51 13.32 3.28
P L04 1.19 12.33 6.16
P L05 0.33 11.71 3.36
P L06 -0.99 11.84 6.26
P L07 -1.64 12.99 3.44
P L08 -1.38 14 6.33
2. simulation data setting: the data of ranging precision analysis is simulated by using single-
channel three-frequency data, and the data loss and cycle-slip anomaly are simulated by
setting L1 frequency point 1/3/5/7/8 signal loss at 10 seconds. At 13 s, the signal
reacquisition is effective and the observation data are output. At 40s, the signal of B1
frequency point 7/8/10/11/13 loses lock, at 40.5 s, the signal of B1 frequency point 7/8 star
reacquisition, and at 63 s, the signal of B1 frequency point 6/9/13 loses lock, 7/8/10/11/13
signals were lost at B1I and B3i frequency points at 70s, and reacquisition was effective at
73s. The rest of the time the data are normal.
3. In the random error simulation the phase error of the receiver output is generally 0.02 cycle.
Here, the empirical value as the reference to complete the error simulation test within the
range of 0.02~0.08.
4. The particle swarm optimization algorithm is initialized. We randomly generate 100 points
within the range described above, set the particle velocity
to vlimit = [-0.5, 0.5;-0.5, 0.5;-0.5, 0.5;] , set the inertia weight to w = [0.5; 0.5; 0.5], the self-
learning factor c1 = [0.5; 0.5; 0.5], the population learning factor c2 = [0.5; 0.5; 0.5], set the
iterations to 100, set the particle fitness initialized to 0, the particle best fitness initialized
to 0, and the particle best position initialized to the position minimum point.
4.2. Analysis of phase error and ranging accuracy
It is found from the simulation data that when the carrier phase error is set in ± 0.01 range, the
estimation of the whole cycle of ranging is almost all in 57, and there is a ± 1 week ranging error at
some points When the carrier phase error is set in the ± 0.02 range, the whole-cycle estimation is
similar to that when the phase is set in ± 0.01, only the partial data is more in the 56-week
estimation; when the carrier phase error is in the ± 0.04 range, there are frequent fluctuations and
multi-values in the whole-cycle estimation, and the maximum range estimation can be up to about
1 m, the data can guarantee more than 90% of the correct range estimation.This is shown in Figure
4.
Figure 4: Estimation of ranging accuracy
4.3. Analysis of frequency points and ranging accuracy
Of the 1000 sets of simulation data analyzed in this section, 45% introduced a ± 0.04-week
phase error and 55% introduced a ± 0.02-week phase error. The results show that: (1) the longer
the wavelength of the two frequency signals, the more accurate the range estimation is; (2)
increasing the frequency can effectively increase the accuracy of range estimation.This is shown in
Figure 5.
Figure 5:With any combination of frequency points, the whole-cycle estimation satisfies ±1
cycle probability
According to the simulation results, the original KPI-based Dopler localization algorithm can
achieve the reusing of the signal data and participate in the localization for the instantaneous
signal loss and recapture, but the data is continuously unavailable, compared with this algorithm,
the proposed algorithm ensures the stability of the signal in continuity and location accuracy by
multi-frequency constraint and abnormal signal reacquisition and repair.
Figure 6: Location error analysis
5. CONCLUSION
In order to solve the problem of cycle slips in pseudolite indoor positioning, a new method based
on multi-frequency constraint is Particle swarm optimization. This method avoids the ranging error
caused by frequent cycle slips in indoor environment and time-based ranging problems such as Toa
and TDOA. The signal transmission distance can be calculated by using the inter-cycle phase
relation of dual-frequency signals, and then use the Particle swarm optimization for high-precision
positioning. The algorithm can effectively improve the continuity of pseudolite signal in indoor
location and phase recovery of signal recapture. The simulation results show that the multi-
frequency constraint can not only improve the application of KPI algorithm for single signal lock-
down and recapture, but also improve the ranging accuracy, the location continuity of the signal is
further improved. However, the algorithm can not locate the signal if the dual frequency of the
tracking signal is not satisfied, so the KPI algorithm has more advantages, on the premise of
ensuring accuracy, the continuous availability of positioning is further improved.
Acknowledgements
This work was supported in part by the National Key Research and Development Plan of China
and the major scientific and technological achievements transformation project of Hebei Provincial
Department of Science and Technology (project: High precision positioning, navigation and control
technology for large underground space(No. 2021YFB3900800); Study on application of high-speed
and high-precision positioning in urban highway tunnel(No.23565901D).
References
[1] Pei Ling, Liu Donghui, Qian Juchao. Review of indoor location technology and its
application[J]. Navigation and timing, 2017, 4(03): 1-10.
[2] Yan Dayu, Song Wei, Wang Xudan, etc. A review of the development of indoor location
technology in China [J]. Journal of Navigation and positioning, 2019, 7(04): 5-12.
[3] Deng Zhongliang, Yin Lu, Tang shihao, etc. Review of key technologies for indoor location [J].
Navigation and timing, 2018, 5(03): 14-23.
[4] Liu Jingbin, Zhao Zhibo, Hu Ningsong, etc. Summary and prospect of indoor high precision
positioning technology [J] . Journal of Wuhan University Sciences (information science
edition),2022,47(07):997-1008.
[5] Deng Zhongliang. Demonstration of key technologies and services for high-precision indoor
and outdoor positioning and navigation in cities[J]. China's scientific and technological
achievements,2015,(5):37-40.
[6] Zhang Shengjun, Lin Ruolin. The present situation of indoor positioning technology [J].
Mapping and spatial geographic information, 2018, 41(7): 128-131.
[7] Wang Congchao. Research on key technologies for wireless indoor location [D] Southeast
University,2018.
[8] Zhang ran, song Lailiang, ran Longjun. Particle swarm optimization based ultra-wideband
location algorithm in non-line-of-sight environment [J]. Sensors and
microsystems.2017,36(09):117-120,124.
[9] Yang Ziyang, Wu Caizhang, Zhang Chi. Hybrid UWB indoor location algorithm based on
Chan and improved UKF [J]. Combined Machine Tool and automatic machining
technology,2020(12):65-69.
[10] Chen L, Kuusniemi H, Chen Y W, et al. Information filter with speed detection for indoor
bluetooth positioning. Proceedings of 2011 International Conference on Localization and
GNSS, 2011:47-52.
[11] Varsamous M, Antonakopoulos T. A bluetooth smart analyzer in iBeacon networks[C].
Proceedings of 2014 IEEE Fourth International Conference on Consumer Electronics, 2014,12:
288-292.
[12] Jiang Dexiang, Hu Mingqing, Chen Yiqiang, etc. Adaptive Bluetooth location method based
on kernel ridge regression [J]. Computer Application Research, 2010, 27(9): 3487-3489,3492.
[13] Manandhar, D.; Okano, K.; Ishii, M.; Kogure, S.; Maeda, H. Development of ultimate seamless
positioning system based on QZSS IMES. In Proceedings of the International Technical
Meeting of the Institute of Navigation, Savannah, Georgia, 16–19 September 2008.
[14] Manandhar, D.; Kawaguchi, S.; Torimoto, H. Results of IMES implementation for seamless
indoor navigation and social infrastructure platform. In Proceedings of the 23rd International
Technical Meeting of the Satellite Division of the Institute of Navigation, Portland, OR, USA,
21–24 September 2010.
[15] Fujii, K.; Sakamoto, Y.; Wang, W.; Arie, H.; Schmitz, A.; Sugano, S. Hyperbolic Positioning
with Antenna Arrays and Multi-Channel Pseudolite for Indoor Localization. Sensors 2015 , 15,
25157–25175.
[16] Fujii, K.; Yonezawa, R.; Sakamoto, Y.; Schmitz, A.; Sugano, S. A combined approach of Doppler
and carrier-based hyperbolic positioning with a multi-channel GPS-pseudolite for indoor
localization of robots.In Proceedings of the 2016 International Conference on Indoor
Positioning and Indoor Navigation (IPIN),Alcala de Henares, Spain, 4–7 October 2016; pp. 1–7.
[17] Sakamoto, Y.; Ebinuma, T.; Fujii, K.; Sugano, S. Doppler pose estimation using multiple IMES
transmitters for indoor localisation. J. Locat. Based Serv. 2014, 8, 36–53.
[18] Niwa, H.; Kodaka, K.; Sakamoto, Y.; Otake, M.; Kawaguchi, S.; Fujii, K.; Kanemori, Y.; Sugano,
S. GPS-based indoor positioning system with multi-channel pseudolite. In Proceedings of the
2008 IEEE International Conference on Robotics and Automation, Pasadena, CA, USA, 19–23
May 2008; pp. 905–910.
[19] Gan, X.; Yu, B.; Chao, L.; Liu, S. The development, test and application of new technology on
Beidou/GPS dual-mode pseudolites. In Proceedings of China Satellite Navigation Conference
(CSNC) 2015 Proceedings, Xian,China, 13–15 May 2015; Springer: Berlin/Heidelberg, Germany,
2015; Volum，340, pp. 353–364.
[20] [9] Huang L , Gan X , Yu B , et al. An Innovative Fingerprint Location Algorithm for Indoor
Positioning Based on Array Pseudolite[J]. Sensors (Basel, Switzerland), 2019, 19(20).
[21] Gan, X.; Yu, B.; Heng, Z.; Zhu, R.; Li, Y. Pseudolite cellular network in urban and its high
precision positioning technology. In Proceedings of the China Satellite Navigation Conference
(CSNC) 2017 Proceedings, Shanghai, China,23–25 May 2017; Springer: Singapore, 2017;
Volume 437, pp. 313–324.
[22] Taejo of Goryeo, Zhang Xianzhou. GPS single epoch ambiguity search based on improved
particle swarm optimization algorithm [J]. Science of surveying and mapping,2013,38(6):157-
159.
[23] Zhai Yanrong, Huang Huan, Zhang Shen, etc. . Improved Particle swarm optimization for
TDOA location [J]. Sensors and Microsystems,2013,32(4):145-148,152.