Robust Wireless Sensor Networks for Transmission Line Monitoring in Taiwan

Based on topological and geometrical data of an actual national power grid, this paper first presents the statistical characteristics of a transmission line monitoring wireless sensor network (WSN). We have discovered that the length of a transmission line follows a Weibull distribution. The number of towers in a transmission line has a Negative Binomial distribution, and the distance between two adjacent towers is Gamma distributed. Using this topology information, we further study the robustness of a WSN for transmission line monitoring. Compared to controlling node degree, we have found that robustness can be better achieved by providing multiple node-disjoint paths between a communication node and the control center. We have developed an algorithm to find the minimum communication range in achieving a desired robustness. Through extensive numerical results, we have confirmed that it is possible to guarantee robustness against a single node failure as long as we can provide a communication range of at least 1.09 km.