Search K Nearest Neighbors on Air

Abstract

Abstract. While the K-Nearest-Neighbor (KNN) problem is well studied in the traditional wired, disk-based client-server environment, it has not been tackled in a wireless broadcast environment. In this paper, the problem of organizing location dependent data and answering KNN queries on air are investigated. The linear property of wireless broadcast media and power conserving requirement of mobile devices make this problem particularly interesting and challenging. An efficient data organization, called sorted list, and the corresponding search algorithm are proposed and compared with the well-known spatial index, R-Tree. In addition, we develop an approximate search scope to guide the search at the very beginning of the search process and a learning algorithm to adapt the search scope during the search to improve energy and access efficiency. Simulation based performance evaluation is conducted to compare sorted list and R-Tree. The results show that the utilization of search scope and learning algorithm improves search efficiency of both index mechanisms significantly. While R-Tree is more power efficient when a large number of nearest neighbors is requested, the sorted list has better access efficiency and less power consumption when the number of nearest neighbors is small.

On processing location based top-k queries in the wireless broadcasting system

“…Air index interleaving [2][3][4][5][14][15][16] is commonly used for reducing the tuning time at the expense of the increased access time. The basic idea is to interleave an index information with data objects on the broadcast stream.…”

“…When the best first algorithm tries to visit N1 after visiting N2, it has to wait for the next broadcast stream, since N1 has already been broadcast. By taking the sequential property of the broadcast stream into account, the appearance-first algorithm has been considered in [2,4] for improving the access time performance. The appearance-first algorithm sequentially visits the R-tree nodes in the order of their appearance on the broadcast stream, while filtering out the unqualified nodes according to the mindist-and minmaxdist-based heuristics.…”

“…Mobile users, equipped with hand-held devices, can now access a variety of valuable information from anywhere and at any time. In mobile computing environments, wireless broadcasting is considered to be an effective way for provisioning LBSs since it accommodates an unlimited number of mobile users simultaneously at a constant cost [2][3][4][5]16].…”

A Hilbert Curve-Based Distributed Index for Window Queries in Wireless Data Broadcast Systems

“…LDSQ in the wireless environment is that mobile users query the spatial objects dependent on their current location. Examples of LDSQs include querying local traffic reports and the nearest restaurants with respect to user's current location [7]. Because of its high scalability, wireless data broadcast is an efficient way to disseminate data to a large number of mobile users.…”

“…In [9], Zheng et al proposed the grid-partition index to support NN queries. The studies in [3,7] are specified for kNN queries. The Hilbert curve index [8] provides an index structure to support window queries and kNN queries.…”