Research On Continuous Probabilistic Skyline Queries For Moving Objects

As the rapid development of wireless communication and positioning technologies, as well as much popular with location-based wireless handheld and car mass devices, the Location-Based Service has become a hot topic research in the field of mobile computing. Skyline query is an important operator of LBS, which aims to find a subset of interesting points that are not dominated by any other points. The moving objects with uncertainty make the dominant relationship of data points instable, which affects skyline operator. In this paper, skyline queries for moving objects with uncertainty under unconstrained space and road network environment are studied. The main work can be described briefly as follows:Firstly, based on the existed models of uncertain moving object, the uncertainty-point model in unconstrained space environment and the uncertainty-interval model in road network environment are choosed. The expressional methods of dominant probability between two data points are defined according to the above models, and then it is proposed how to compute the dominant probability and skyline probability by differential element method.Secondly, the influence of spatial (road network) positions of data points on dominance probability is investigated, which provides an indication that how to find changes for dominant probability. Two types of events affecting p-Skyline are defined to track and update p-skyline set continuously instead of re-computing the whole dataset each time.Finally, the dynamic incremental algorithms are presented based on above definitions. By tracking and calculating the events, the operation of continuous updated p-skyline can be achieved, which reduced search steps and system overhead, and the efficiency is improved as well. There is no algorithm to handle continuous probabilistic skyline queries for uncertain moving object, so we proposed the static algorithms to compare with the dynamic incremental algorithms. The experiments having positive results show effectiveness of the proposed algorithms.