On Semantic Caching and Query Scheduling for Mobile Nearest-Neighbor Search

Location-based services have received increasing attention in recent years. In this paper, we address the performance issues of mobile nearest-neighbor search, in which the mobile user issues a query to retrieve stationary service objects nearest to him/her. An index based on Voronoi Diagram is used in the server to support such a search, while a semantic cache is proposed to enhance the access efficiency of the service. Cache replacement policies tailored for the proposed semantic cache are examined. Moreover, several query scheduling policies are proposed to address the inter-cell roaming issues in multi-cell environments. Simulations are conducted to evaluate the proposed methods. The result shows that the system performance, in terms of cache hit ratio, query response time, cellcross number and cell-recross number, is improved significantly.