Resource Management in Wireless Sensor Networks Hosting Multiple Applications

Sensor grids comprise of wireless sensor networks (WSNs) and the computing grid. In a sensor grid real time information about a phenomenon being monitored can be obtained from a wireless sensor network and processed in a computing grid. With the advancement of sensor technology and with the aim to increase the cost efficiency of wireless sensor networks, multi-purpose wireless sensor networks are gaining popularity over wireless sensor networks dedicated to a single application. When a WSN is shared by more than one application, management of sensor resources becomes a challenging task. Resource management in a WSN comprises of two major functions: Allocation and Scheduling This research presents a simulation based investigation of allocation and scheduling on wireless sensor networks hosting multiple applications. This thesis proposes scheduling algorithms that combine the knowledge of the application and the system in order to make a scheduling decision. Simulation experiments demonstrate that such algorithms provide a better performance than the knowledge free First Come First Served (FCFS) algorithm. Simulation experiments demonstrate that communication delays play a significant role in determining the overall mean response time in a WSN and a scheduling algorithm that allocates a higher weight to higher hop distances from the cluster head provides the best performance.;A number of allocation techniques are also proposed in this thesis. The aim of any allocation algorithm is to allocate sensors to application requests in order to enhance the network lifetime of the wireless sensor network. Allocation algorithms that use varying degrees of information about the state of the sensor nodes and the network are proposed in this research. Both static and dynamic allocation techniques are proposed in this research. Experiments demonstrate that by using information about the energy consumption at the CPU component and the radio component of the sensor nodes, and making dynamic allocation of sensor nodes to application requests in the wireless sensor network, lifetime of the WSN can be increased.