| In recent years, with the increase in demand forlocation-based services LBS (Location Based Service), Galileo and Beidou satellite gradually launched into space, navigation and positioningresearch has attracted worldwide attention. Navigation and positioning studies have come out of the navigation and positioning service platform applications. This topic is based on the Fund of the Hong Kong Innovation and Technology (ITF)"Beidou satellite navigation system, marine location-based services to develop multiple sources of information. In this project, the final completion of the development of the system is an instant messaging distributed system. The research focuses on three aspects:client application-based distributed software system architecture, multi-processor-based concurrent data mining algorithm, the Beidou navigation and positioning as the core multimode navigation communications baseband algorithm research.1. The navigation of baseband signal processing research:the rapid processing of GNSS signals, Galileo phase tracking loop, interval kalman filtering algorithm.1) A bit-wise parallel algorithm base on the sign bit for real-time GPS software receiver. For the interest of low cost, short development cycle and versatility, developing GPS software receiver is hotspot technology nowadays. Considering the heavy computational load and the chip processing capacity which used in PDA or smart phone, it is necessary to improve the algorithm of real-time signal processing. By analyzing some existing parallel algorithms of correlation, an improved bit-wise parallel algorithm is presented which based on the sign bit. Through the simulation on time cost and positioning accuracy, it shows that the signal treatment speed of the new algorithm is increased at least2times than before while the loss of Carrier-to-noise-density ratio (C/N0) is less than1dB-Hz. These improvements have a very large significance for the navigation on the mobile devices which combined with the GPS software technology and GIS maps.2) Interval kalman filtering algorithm for high dynamic navigation and positioning. Kalman filtering algorithm for maneuvering target has a wide range of navigation and positioning applications, but there are still two shortcomings which make it can’t be the best filter. First, the precise maneuvering target model need to be known; second, a hypothetical model of system noise and observation noise is zero mean Gaussian white noise, and the statistical properties are known. In this paper, the effect of the deviation in the estimated statistical characteristics of the measurement noise and system noise has been analyzed from the perspective of time-varying noise. Based on interval arithmetic, according to a priori and a posteriori estimates range overlap interval estimates the probability of containing the true value is very high, a new interval kalman filtering algorithm has been proposed. Through the experimental comparison, the interval kalman filtering improves3dB than the traditional kalman filter to solve the movement with system noise and estimation errors of measurement noise.3) The research on the combination of discriminators in galileo LIF signal. Carrier phase tracking is the weak link of GPS. The presence of a pilot channel in Galileo LIF signal makes the choice of the phase discriminator is not restricted by navigation data. The traditional combination of discriminator in Galileo LIF signal carrier phase tracking loop is for the below. The data channel uses a discriminator which is insensitive to180°phase shift, when the pilot channel uses a discriminator which is not required to be insensitive to180°phase jumps. Then, the results of above are weighted. The discriminator for pilot channel-only has better performances than the discriminator for data channel-only. In this paper, utilizing the information of pilot channel makes the data channel can use the discriminator without restrictions of navigation data, a new combination of discriminators has been proposed. Through the simulation experiment under different thermal noise environments, the performances of the new and the traditional combinations of discriminators are tested with the objective of improving the carrier phase tracking accuracy. The results show that the new combination of discriminators has great advantages on restraining signal noise and preventing loss of lock.Therefore, the proposed algorithms to improve the performance of GNSS receivers in this thesis are based on theoretical investigation and experiments with the simulated and real GPS signals, and then show expectable results. It has a great theoretical and practical significance on the design of high performance and cost performance receivers.2. Parallel programs are mainly two models:one based on multi-processor, the other based distributed. Due to the huge size of database and increasing amount of computation, a number of parallel algorithms have already been proposed. In order to achieve high-performance parallel computing, it is necessary to reduce redundant computation and avoid too much communication between parallel tasks, and achieve load balancing. Parallel algorithms, based on multi-processor, are concerned about the distribution of parallel tasks to improve the utilization of computing resources. These methods often require a compromise in computation, communication, load balancing, etc. For the most effective control thread, we first introduce the thread pool into the data mining algorithms. This method does not require redundant communication or computation, but can achieve load balancing so as to fully utilize the computing resources.3. On the basis of the in-depth study, the topic develops an instant messaging server distributed system. The messaging middleware is used to connect a number of different servers. In order to achieve high-performance communications components, which focuses on analysis and design of high performance thread pool, message asynchronous communication, and independent of the routing algorithm based on the unlock. |
PDF Full Text Request