Study On Positioning Algorithms Based On TDOA In Cellular Networks

With the development of cellular network and emergence of internet O2 O, an explosive growth is presented in exact location-based service App, and wireless location service has become one of the most popular mobile communication business at present. This thesis has conducted a study on TDOA-based location technology in cellular network.Firstly, this thesis summarizes the research background and status of the subject; and introduces the common location technology in cellular network and evaluation index of positioning performance; it mainly emphasizes on the principle and mathematical model of TOA, AOA and TDOA. Through the comparative analysis, TDOA technology is selected as the research emphasis of this thesis.Secondly, detail derivation is conducted on TDOA-based classic localization algorithm. Under different parameter condition, simulation comparison is carried out on Fang algorithm, Chan algorithm, Taylor series method and SI algorithm, and four algorithms are summarized. Due to the fact that accurate initial iterative value is hard to obtained by Taylor series method, collaborative algorithm which takes the location result of Chan algorithm as the initial value of Taylor series method has been conducted. The simulation shows that the positioning performance of collaborative algorithm is well behaved.Finally, this thesis focuses on the application of genetic algorithm in TDOA location. Two improved methods are proposed for the premature problem existing in SGA. On the basis of SGA, cosine type self-adaption operator and improved niche technology are introduced, so that the individual can dynamically adjust parameter in the process of population evolution, reaching the goal of maintaining population diversity and relieving premature in the final. Simulation shows that the accuracy of algorithm proposed in this thesis is higher than that of comparing algorithms, which can effectively alleviate the premature problem of SGA.