The Acquisition Optimization And Hardware Acceleration Of GNSS Receiver

GNSS receiver is an important part of the user segment of Global Navigation Satellite System(GNSS).GNSS receiver consists of high sensitivity receiver,high dynamic receiver and high precision receiver,which have different application scenarios and adopt different techniques.High sensitivity receiver is good at capturing low power signal.High dynamic receiver is suitable for the high dynamic scene,which have higher demand to the signal tracking technology.And high precision receiver usually adopts some advanced navigation solution in order to obtain better speed and location accuracy.This paper mainly studied the acquisition optimization and hardware acceleration of GNSS receiver.Acquisition optimization can improve the receiver sensitivity and positioning time,which is also one of the key technologies in receiver.Hardware acceleration algorithm can reduce signal processing delay,thus is helpful for the signal capture module,signal tracking module and navigation calculating module.In this paper,the main research contents and results are as follows:(1)The three parts of GNSS are described in detail,which respectively are space segment,control segment and user segment.Based on the three parts the analysis of four major global satellite navigation systems are introduced,which are GPS,GLONASS,GALILEO,BDS.(2)This paper briefly introduces the main framework of GNSS receiver,mainly analyzes the tracking module and capture module.Then,theoretical analysis of the parallel frequency capture algorithm and parallel algorithm for phase capture is given.Based on the analysis of processing gain in the parallel capture,safely draw the conclusion that the parallel capture algorithm can improve the efficiency of the capture time,but at the same time compared to the serial correlation capture will cause certain SNR gain loss.Then the paper analyzes the three factors that influence the SNR gain in AWGN,which are integration time,code phase shift and sampling rate.In fixed time and code fragment offset related cases,adopting the double code rate of the sampling rate can achieve the SNR gain that is close to the SNR gain extremum.(3)For the rest of the gain loss caused by the residual code phase offset,the residual code phase elimination algorithm and the PMF FFT used in the receiver are merged,which is called the new PMF_FFT_RRP algorithm.The false-alarm probability and detection probability of PMF_FFT_RRP is analyzed.Then safely draw the conclusion that this algorithm is not only simple,but also can keep less capture time the same as PMF_FFT.But unlike PMF FFT algorithm drastic fluctuation is detected when the the residual code phase changes,the performance of PMF FFT RRP algorithm in this case are quite stable and robust.At the same time the paper introduces an algorithm of precise frequency acquisition,which is compared to the coarse frequency acquisition according to simulation.(4)The paper introduces the FFT algorithm based on high-order approximate kernel and Cordic algorithm to achieve the higher requirements of signal processing real-time in GNSS receiver.FFT algorithm based on high-order approximate kernel use shifting and adding to achieve the original butterfly arithmetic and the multiplication with rotating factors.With multiple iterations of shifting and adding Cordic algorithm can achieve a variety of trigonometric functions,transcendental function and matrix decomposition.According to simulation we found that the above two types of hardware acceleration algorithm not only guarantee the same computing performance,but also save the operation time.