Study On Key Localization Algorithms Of Mobile Terminal

With the development of electronic, Internet and mobile communication technology, intelligent transportation information engineering with position information as the core technology is supported to extend continuously in the depth and breadth, requirements of LBS(Location Based Service) in public traffic travel are also more broad and clear. Many countries Governments in the world are also give specific mandatory demands to improve the efficiency of government administration. This further leads to a wide application of LBS in more industries and national life field. So, LBS also shows a great application prospect and commercial value, it is important and necessity to accurately estimate the location of mobile terminal. This has attracted wide attention and researching at home and abroad.The research studies the status and existing problems in mobile terminal positioning. The core of researching is focusing on improving the terminal position estimation accuracy and giving the following several innovative points:(1) Analyzes several classic coordinate calculation algorithms based on TDOA(Time Difference of Arrival), simulates and gives comparative analysis results in WLAN environment from the applicability, convergence and accuracy several aspects. For the steepest descent method is adaptability and with a higher convergence speed with a large error in initial, and Levenberg-Marquardt iterative method has a higher convergence speed while the estimates closer to the true value, this paper gives a hybrid location method fusion the two methods’ advantages. With data sampling using handheld terminal and simulation analysis, this algorithm is verified has a higher positioning accuracy, good applicability and strong robustness.(2) Calculates and simulates geometric dilution of precision(GDOP) of three typical layouts, respectively are terminal located in the central, at any position on the circumference and any position on the symmetric line. Reduces a conclusion: the GDOP value appears a certain regularity with the position relationship between terminal and base stations, this will be a good guiding significance for base station layouts implementation in engineering. This paper also provides a layout optimization model using barrier function method and simulates annealing method with the condition of base stations’ location relationship. Through data sampling using vehicle terminal and simulation analysis, the layout optimization model can effectively reduce the GDOP values of positioning region and overall improve the vehicle terminal positioning accuracy in the region.(3) Studies the NLOS identification algorithm and error suppression algorithm of several typical signal statistical distributions and statistical characteristics. Using the information of TDO A measurement variance in NLOS conditions is larger than LOS conditions, gives a NLOS identification method in TDOA algorithm established a contrast model and discriminant threshold of actual measurement variance and ideal LOS measurement variance. Also gives a Taylor optimization method adding NLOS error compensation based on Taylor series expansion method. Simulates the two methods in a cellular communication network, concludes the result that demonstrates the validity of the N LOS identification method and the positioning precision of Taylor adding error compensation be improved significantly compared with no error compensation. This paper also gives a kind of GPS/TDO A hybrid location algorithm with data fusion based on adaptive federal Kalman filter. N LOS error suppression based on extended Calman filter are used respectively in GPS pseudo range positioning and the cellular TDOA positioning. With data sampling using handheld terminal and system simulation, it is showed that after the NLOS error suppression and data fusion, the positioning errors of GPS and TDOA are obviously decreased, and the accuracy of GPS+TDOA hybrid location algorithm is obviously higher than separate positioning accuracy.(4) Introduces GPS carrier phase observation model and fuzzy degree solution principle using integer least square method. Analysis several common integer ambiguity resolution algorithms. According to the double difference of strong correlation lead to LAMBDA float ambiguity resolution failure, this paper gives a new method using Gauss transform on LAMBDA float ambiguity for decorrelation to improve the ambiguity integer solution accuracy and searching efficiency. According to the situation of baseline pseudo distance and less epoch may be measured, puts forward a method of using baseline pseudo distance constraints to further optimize LAMBDA. In this method, improving the co factor matrix of float ambiguity, reducing the correlation of ambiguities and ambiguity search space are happened to improve the ambiguity resolution accuracy, stability and success rate. The experiment simulation proves the validity of the two optimization algorithm.