LiteWS: A Web Service Enhancing Multi-User Queries in Data Intensive Sensor Networks

Internet-based data intensive sensor networks (DISNs) are sensor networks wherein large volume of different types of sensory data are sensed and generated from the physical world, and queried by multiple users simultaneously. One critical issue in Internet-based DISNs is how to support multiple user queries simultaneously in an efficient manner, in terms of query response time, query loss ratio and sensor node energy consumption. In this paper, we formulate and study multi-user data query problem in DISNs and present our solution. Specifically, we present our design, implementation, and evaluation of LiteWS, a web service system aiming to enhance multi-user queries in Internet-based DISNs. We propose a simple caching technique and give an analytical model of how to reduce query loss ratio in our system. Through extensive experiments based on Crossbow IRIS motes, we evaluate the system performance of LiteWS under both correlated and uncorrelated query traffic. We show the performance with data caching is much better than the one without; particularly, the average query response time of LiteWS is improved 5 to 10 times and the query loss rate is improved 2 times.