On Maximizing Lifetime of a Sensor Cluster

We consider the energy consumed in radio transmission of a set of sensors forming a data gathering wireless network. Our objective is to enhance the lifetime of such networks by exploiting three system-level opportunities. Firstly, the number of bits to be transmitted can be reduced by taking advantage of the redundancy induced by spatio-temporal correlation in sensor data. Secondly, channel coding allows us to reduce transmission energy at the cost of increased transmission time. Thirdly, sensor nodes can be expected to operate collaboratively, allowing optimal management of distributed energy resources. Our main contribution lies in providing a framework to merge these ideas for energy conscious networking. We pose the problem of maximizing network lifetime as an optimal scheduling problem. We first consider a special case where data rate is linearly proportional to received signal power. In this scenario, we investigate both static and dynamic scheduling strategies. The optimal static schedule turns out to have a very simple form. For the dynamic case, we obtain an integer linear program formulation to find the optimal strategy. We then propose an efficient algorithm that exploits the special nature of the problem setting to quickly find the optimal solution. Finally, we consider the general case where data rates and signal power need not be linearly related and propose an algorithm to find the optimal transmission times subject to the deadline constraint imposed by the system