A framework for managing the selection of spatiotemporally relevant information providers

The development of future pervasive sensor-enabled systems, where information is distributed on-demand across heterogeneous networks, highlights the necessity for an efficient framework to determine the relevancy of provided information with respect to one's needs. This paper considers the problem of selecting the most “spatiotemporally” relevant providers in order to meet a user's information needs over a time period of interest. Initially, a definition and a measure of spatiotemporal relevancy is developed to measure the degree of relevancy of sensory information with respect to both its spatial and temporal characteristics. Based on these, the selection of the most relevant set of providers under budget constraints is expressed as an integer programming optimization problem and a two-level dynamic programming (DP) algorithm is proposed to solve it optimally. Moreover, a number of alternative methods are proposed in order to accelerate the provider selection process by making approximations either to the overall optimization problem formulation or the relevancy calculation method itself. Finally, the performance of the proposed methods are examined both analytically and by simulation for a number of provider scenarios.