IoT Data Prefetching in Indoor Navigation SOAs

Abstract

Internet-based Indoor Navigation Service-Oriented Architectures (IIN-SOA) organize signals collected by IoT-based devices to enable a wide range of novel applications indoors, where people spend 80--90% of their time. In this article, we study the problem of prefetching (or hoarding) the most important IoT data from an IIN-SOA to a mobile device, without knowing its user’s destination during navigation. Our proposed Grap (Graph Prefetching) framework structurally analyzes building topologies to identify important areas that become virtual targets to an online heuristic search algorithm we developed. We tested Grap with datasets from a real IIN-SOA and found it to be impressively accurate.