A new hybrid algorithm on TDOA localization in wireless sensor network

A hybrid algorithm for TDOA localization is proposed in this paper. It has well combined the advantages of genetic algorithm and quasi-Newton algorithm. The hybrid algorithm has sufficiently displayed the characteristics of genetic algorithm´s group searching and quasi-Newton method´s local strong searching. At the same time it effectively overcomes the problem of high sensitivity to initial point of quasi-Newton method and shortcoming of genetic algorithm which reduces the searching efficiency in later period. The experimental results show that if the parameters are assumed reasonably the hybrid algorithm has extremely stability, higher localization rate and localization precision than genetic algorithm and quasi-Newton algorithm.