An improved K-th optimal path searching algorithm for mobile phone location

Abstract Mobile phone location data has the characteristics such as sampling is random, time interval is long and uncertain, there is greater limitations using the traditional K-th optimal path searching algorithm to search path. We propose an improved random probability K-th optimal path searching algorithm (K-OPSA). This method takes several user's mobile location data as the research object, the searching area is limited by using the modified A * algorithm, combined with random probability statistics to obtain the history probability of the road links, integrated K-th optimal path and improved Dijkstra optimal path algorithm to select the optimal road links which covered by the searching area, find out several optimal paths as reference between OD cells. And then describes the design idea, organizational structure and operational process in detail. Analyze the time and space complexity of the algorithm through practical verification, this method can improve the search space and time cost effectively, so as to solve the problem that there are several random paths and the accuracy of the map matching in traditional road network.