MODELING LONG RUNNING QUERIES IN WIRELESS SENSOR NETWORKS WITH MULTIPLE BASE STATIONS

Abstract

Many sensing based applications are facilitated by Wireless Sensor Networks (WSN). Even though there are number of WSN based applications, all of them have one common inherent limitation which is energy limitation of the sensors. Most of the time, modern WSN based applications use a shared WSN so that it can be used by many users for di erent purposes. Inability to bear the di culty of deploying separate WSN per user is also a catalyst for the establishment of shared WSN. Energy limitation in sensor node becomes more important in shared WSN because energy consumption level will be increasing with the user count. Also in a shared WSN, most probably users are accessing the sensor through multiple base stations rather than using a single centralized base station. When it comes to the continuous sensing applications, users need to monitor the sensor eld for a particular duration while utilizing the energy of the sensor node e ectively. Lifetime based long running queries play a major role with respect to that. Also, in a shared WSN, there is a possibility to share the results among di erent users rather than retrieving them multiple times from the sensor node. This can be done if two or more users demand the same data item. WSN research community has introduced many solutions for continuous sensing applications. However, those solutions have been designed for a WSN with a single base station. Also the sensing rate is de ned at the user level without giving the exibility of adjusting the sensing rate based on the amount of energy available in the sensor node. This research introduces a model to deploy lifetime based long running queries in a multi user WSN with multiple base stations. Moreover, the proposed model o ers the exibility to adjust the sensing rate depending on the energy level of the sensor node. To develop the solution, connection between base stations are used. Results obtained at the evaluation stage show that the proposed model is capable of saving energy while o ering the exibility to decide the sensing rate based on the energy of the sensor node. Connection between base stations and attribute sharing is the core concept in this model.