Study of Simulation Platform for BDS/INS/CNS Deep Integration Navigation

In this paper, INS/CNS (CINS) is integrated into a module, and then CINS and BDS are further combined to form a deep integrated BDS/INS/CNS navigation system, which can significantly improve the navigation, positioning, and attitude measurement accuracy under high dynamic and strong interference conditions. It has broad application prospects for specific users such as high-altitude long-endurance Unmanned Aerial Vehicle (UAV) and high-maneuvering glider. In order to verify and analyze the algorithm and performance of the deep integrated navigation system, the design and implementation of BDS/INS/ CNS deep integrated navigation simulation platform are presented and the overall architecture, information flow, and the composition of each subsystem of the simulation platform are introduced. ,e simulation results show that, under high dynamic conditions, the position accuracy of the BDS/INS/CNS deep integrated navigation system is better than 1m, the speed accuracy is better than 0.1m/s, and the overall performance is better than the BDS/INS deep integrated navigation system. It also verifies the availability of the simulation platform, which has guiding significance for the next design of BDS/INS/CNS deep integrated navigation prototype.


Introduction
Global Navigation Satellite System (GNSS) and Inertial Navigation System (INS) have good complementary advantages. erefore, GNSS/INS integrated navigation is called "golden combination" [1][2][3][4][5]. According to different observation information, it can be divided into three combination modes: loose combination, tight combination, and deep combination [1,6,7]. GNSS/INS deep combination is a deep and hardware-based combination method, which can realize the fusion of I/Q signals output by GNSS receiver correlators and INS navigation parameters and has excellent navigation performance in high dynamic, strong interference, weak signal, and occlusion environment [2,8], which is the main development direction of GNSS/INS integrated navigation in the future. e algorithm research in this area has also been widely concerned [5,[9][10][11]. e Celestial Navigation System (CNS) can determine the attitude of the aircraft through real-time sensitive star map, stellar extraction, star map recognition and tracking, attitude calculation, and other operations, with an accuracy of angular second [12]. INS/CNS integrated navigation can meet the need of long-endurance and autonomous navigation [13]. Now, China's BeiDou Navigation Satellite System (BDS) has launched 41 BeiDou satellites, which can provide services along the "belt and road" by the end of 2018 and achieve global services around 2020. At present, INS and CNS have been widely installed on some high-altitude aircraft.
e research on GNSS/INS/CNS combination at home and abroad mainly focuses on the federal filter fusion processing and the improvement of related algorithms [14][15][16][17]. Loose combination is still the main mode, and the combination information has not been fully utilized. e BDS/INS/CNS deep integrated navigation is to combine INS/CNS into a mode (CINS). e deep combination of CINS and BDS can improve the accuracy of navigation and positioning and can be applied in the fields of unmanned reconnaissance aircraft and high-mobility gliding aircraft in the air. It has broad application prospect. e deep combination of BDS/INS/CNS is the fusion of INS information and CNS information with the tracking level of the BDS receiver. It involves the internal programming of the BDS receiver. It is difficult to carry out physical or semiphysical simulation test in the application environment of high dynamic and strong interference, and many factors are taken into account [18,19]

Deep Integrated Navigation System
e concepts of GNSS/INS deep combination is derived from the "vector tracking" proposed by Spilker [8]. Its main idea is to generate new navigation parameters by using the residual error of correlator and to predict the BDS signal tracking loop parameters by using updated navigation parameters [20]. At present, there are generally two design systems [1,2,21,22]. One is the cascade deep combined navigation system. e measurement value of integrated navigation filter is the baseband I/Q information processed by the baseband signal preprocessing filter. e other is the centralized deep integrated navigation system. e measured value of the integrated navigation filter is the difference between the baseband I/Q information and the inertial navigation estimation I/Q information. From the perspective of information fusion optimal criteria, the centralized deep integrated navigation system is the best in improving navigation performance. Here, a centralized deep integrated navigation system is taken as an example for simulation analysis. In order to make full use of the CNS information on the aircraft, the deep composite module is extended to the BDS/INS/CNS deep combination mode, and the INS/CNS combined module (CINS) is further combined with BDS. e CINS architecture is shown in Figure 1. e BDS/INS/CNS deep integration architecture is shown in Figure 2.
As can be seen from the structure diagram, the combined model integrates CINS navigation parameter estimation and BDS signal tracking and can simultaneously track and process all visible BeiDou satellite signals. e integrated navigation system can make full use of the high precision of pose measurement and navigation of CINS, provide accurate dynamic information, enhance the adaptability of BDS receiver to dynamic stress, and reduce the bandwidth of BDS signal tracking loop. At the same time, CINS is corrected by using the output information of navigation filter so that the racking loop of the BDS receiver can only track the Doppler frequency shift error caused by the CINS calculation error, crystal vibration of BDS receiver, and external noise, which effectively improves the tracking accuracy of BDS signal. When the BDS signal is disturbed or blocked, the CINS measurement value can meet the requirements of seamless navigation. Meanwhile, the integrated navigation system can predict the aircraft's dynamic information, continuously predict the Doppler and phase shift information of the BDS signal, and significantly reduce the recapture time of the BDS receiver.

The Simulation Platform of BDS/INS/CNS
Deep-Integrated Navigation e BDS/INS/CNS deep integrated navigation simulation platform is mainly composed of the following parts: trajectory generator, IF Signal simulator of BDS, BDS receiver baseband signal processing, IMU signal generator, star sensor simulator, INS mechanical arrangement, and integrated navigation filter. UKF filtering algorithm is adopted. e overall structure of the platform is shown in Figure 3.

Trajectory Generator.
Trajectory generator is the basis for the deeply BDS/INS/CNS integrated navigation simulation platform [1], which can simulate the aircraft's various kinds of sports, such as constant speed, acceleration, turn, and dive. First of all, it is use to input information, including the position, velocity, and acceleration, and simulate IMU output, star sensor output, BDS satellite navigation message, pseudo-orange and the ionosphere, and troposphere delay correction information. Secondly, as the reference value and evaluation criteria, the simulation results are analyzed to verify the correctness and superiority of the BDS/INS/CNS deep combined navigation theoretical model and algorithm. e commonly used method in trajectory simulation is the fourth-order Runge-Kutta method to solve a set of trajectory differential equations, and the specific simulation process is shown in [13].

IMU Signal Generator.
Based on the error model of INS components, the gyroscope and accelerometer simulators are built to simulate the angular and linear acceleration information output. In order to simplify the analysis, it is assumed that the inertial alignment has been completed and the installation error is effectively compensated.

Gyro
Simulator. Given the error information of the gyroscope, its output is

Mathematical Problems in Engineering
where

Accelerometer
Simulator. e output of accelerometer is where ∇ b a represents the error of accelerometer and f b represents the acceleration of the carrier relative to the inertial coordinate system and is expressed as follows: where the relevant parameters are referenced in literature [1,19,21].

Digital Model of Intermediate Frequency
Signal. e signal of B1 is composed of the "range-finding code + navigation message" orthogonal modulation of the two branches of I and Q on the carrier wave [19,23]. Here, the B1I signal is taken as an example for analysis. When the B1I signal transmitted by the j visible BDS satellite passes through the space environment, its analytical expression is where P represents the power of the j visible satellite signal received by the receiver, T p represents the transmission delay from the satellite to the receiver, δt iono represents ionospheric delay, δt tropo represents tropospheric delay, δt SV represents satellite clock, φ 0 represents initial phase, and n(t) represents received noise. In order to facilitate the subsequent baseband signal processing, the received BDS signal need to be mixed frequency processing, the signal down conversion to intermediate frequency, and then sampling processing. Let the local signal be After BDS signal is mixed with frequency, the low-pass filter is expressed as follows: where δt r represents the clock error of the receiver. Let Equation (8) is further simplified as follows: where ω IF represents the desired angular frequency of intermediate frequency signal. When the aircraft is moving at high speed, the change of carrier phase IF caused by Doppler frequency shift should be considered. erefore, the intermediate frequency signal including all visible satellites can be expressed as follows:

3.3.2.
Simulation. e simulation structure of BDS IF signal is shown in Figure 4, it mainly includes the following functional modules: one is the BDS navigation message simulation module compiling down navigation messages based on BDS ephemeris data and the navigation message coding format and error model defined in the file of BDS-SIS-ICD-B1I-1.0; the other is the parameter setting module of the signal state, including the setting of error model parameters, such as simulation time, ionospheric delay, tropospheric delay, and intermediate frequency signal parameters. irdly, BDS IF signal generation module simulates multiple BDS satellites for signal synthesis according to the number of visible satellites in the simulation period. e IF signal simulator is reserved for GPS and Galileo constellation, and the navigation message can be generated according to the corresponding ICD file. Combined with the carrier frequency of different constellations, it can be extended to the IF signal under the corresponding constellation.
Transmission time calculation, various error simulation, signal sampling, and quantitative reference [23]. Different from the GPS signal C/A code, the B1I signal distance code (C B1I ) code speed is 2.046 Mcps, and the code length is 2046. e C B1I codes are generated by two linear sequences, G1 and G2 mode 2, and the truncated 1 code fragment that generates the balanced Gold code. G1 and G2 sequences are generated by two ll-level linear shift registers, and their generating polynomials are G1(X) � 1 + X + X 7 + X 8 + X 9 + X 10 + X 11 , G2(X) � 1 + X + X 2 + X 3 + X 4 + X 5 + X 8 + X 9 + X 11 .

(12)
An adaptive step size calculation method is adopted to precisely control the Doppler signal. e criteria are e baseband signal processing algorithm of the BDS receiver refers to the design idea of the GNSS software receiver to realize the acquisition, tracking, demodulation, and loop estimation of the BDS signal direct input navigation filter for subsequent combination calculation, which is not detailed here.

Star Sensor
Simulator. At present, the main equipment of astronomical observations for the aircraft is the star sensor. erefore, the CNS data simulation is to simulate the altitude angle and azimuth angle output by the star sensor in the carrier system near the latitude and longitude of the aircraft. With the optimal estimation method, the estimation  Mathematical Problems in Engineering of the platform error angle is realized, which simulates the star sensor to achieve "star light, error out" [12,13], and the effect of the data using the new FK5 star catalog is dated. e measurement accuracy of the star sensor has reached angular second level (1σ), and the error will not accumulate with time. erefore, the measurement error of the star sensor can be considered as a white noise process with the zero mean value. e star sensor simulation process is shown in Figure 5.

INS Mechanical Arrangement.
INS mechanical arrangement is to select an appropriate mathematical model according to the observation information of INS to calculate the navigation positioning parameters such as the speed, position, and attitude information of the carrier. Specifically, the rotation angular velocity of the vector relative to the inertial coordinate system measured by the gyroscope is used to calculate the transformation matrix of the vector coordinate system to the navigation coordinate system. e transfer matrix is used to transfer the acceleration measured by the accelerometer from the inertial space to the navigation coordinates, compensate for gravity and Coriolis acceleration, and obtain navigation and positioning information through integration [1]. e corresponding data flow is shown in Figure 6. e mechanical arrangement equation in the local horizontal coordinate system can be expressed as follows: where Ω is the antisymmetric matrix of the angular velocity vector ω, 3.6. Filtering Algorithm. Among these nonlinear filtering methods, the UKF is widely used due to its elimination of the cumbersome derivation and low-computational complexity [2,8,19].

Performance Simulation Analysis
In order to verify the performance of the BDS/INS/CNS deep combined navigation simulation platform, the  Mathematical Problems in Engineering reference moment was selected at 0 : 0 : 0 second on September 9, 2018, the satellite altitude cutoff angle was set at 15 degrees, the sampling frequency of the BDS signal was 30 MHZ, the IF was 7 MHZ, and the predetection integral was 1 ms. And the combined cycle was 1 s. e device characteristic parameters of simulated tactical INS is shown in Table 1, and the update frequency is 100 Hz. Aircraft speed is 1000 m/s, accelerating to 100 m/s 2 , and the attitude angle is 0°. For the inertial position information of longitude  Figure 6: Data flow arrangement of mechanics in the local horizontal coordinate system.   Table 2.
From the above simulation results, it can be seen that  (3) e position accuracy of the BDS/INS/CNS deep combined navigation system is better than 1 m, the Up position error RMS is 0.4161, the velocity accuracy is better 0.1 m/s, and the Up velocity error RMS is 0.05. e accuracy of position and velocity is better than that of the BDS/INS deep integrated navigation system. (4) e accuracy of the position and velocity in the East is better than that in the north, which is caused by the poor observability of the carrier northward during simulation.

Conclusions
BDS/INS/CNS deep combined navigation has broad application prospect and is the focus of current research. Physical and semiphysical simulation experiments are expensive and difficult to be carried out, while the construction of full digital simulation platform can establish and provide physical models and working parameters close to the actual situation. erefore, it provides the basis for the project demonstration and engineering design with strong economy and adaptability.
is paper presents the design and implementation method of BDS/INS/CNS deep integrated navigation simulation platform and introduces the architecture and information flow of the simulation platform in detail. e simulation analysis based on this platform shows that, under the background of high dynamic application, the position accuracy of the BDS/ INS/CNS deep combined navigation system is better than 1 m, and the velocity accuracy is better than 0.1 m/s, which verifies the correctness and effectiveness of the BDS/INS/CNS deep combined navigation platform constructed in this paper. e research on BDS/INS/CNS deep combined navigation accuracy under interference will be carried out in the future. With the further development of artificial intelligence and neural network theory, the centralized deep combination fault detection algorithm will be further studied. is research content has guiding significance for the next step of the design of BDS/ INS/CNS deep combined navigation prototype. e related achievements of the subject can be widely used in the precision guidance weapons of various services and arms and have broad application prospect in the civil field (civil unmanned aerial vehicle navigation, control, etc.). Some algorithm models designed in this paper still need to be further verified by actual measurement data.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.