Impact of Transmission Power and Routing Algorithms in Designing Robust Wireless Sensor Networks

In wireless sensor networks, accurate and complete sensor data is important to the analysis of a phenomenon. Therefore, designing a robust sensor network is of high importance. We first derive the relationships between transmission power, energy consumption and robustness of the network using analytical modeling. Then, we validate our model by using linear programming to find the optimal routing algorithm that will optimize the robustness of the sensor network. The optimal robust routing algorithm is then compared with other popular routing algorithms used in sensor networks and we find that the robustness achieved varies with different routing algorithms. This demonstrates the need for a cross-layer design approach when designing robust sensor networks. In addition, we also prove that an energy-efficient routing algorithm that optimizes the lifetime of the sensor network does not optimize the robustness of the network. By using correlation analysis on the optimal routing algorithm, we design a robust geographic routing algorithm that only utilizes local neighborhood information to obtain nearly optimal robustness performance. Our algorithm also minimizes data losses in cases when sensor nodes fail.