Fault-tolerant scheme for cascading failure of scale-free wireless sensor networks

Abstract

The fault-tolerance control in scale-free wireless sensor networks against cascading failure is analysed in this paper. Based on the node load redistribution and constant capacity of the scale-free wireless sensor networks, a new load distribution model is proposed under a single random node failure. By the probability generating function method, the relationships between the load and the largest connected component after cascading failures is given. The influence of the single random node failure load distribution parameters and topological parameters on the size of the cascading failure is analysed, and then the critical load value of the cascading failure in the power law load distribution is obtained. The scale of cascading failure is proportional to the degree distribution coefficient and power exponent, and it can be controlled by the critical load value. Once the network load parameter is beyond the critical load value, a random node failure will cause the whole network cascading failure. In this paper, the load redistribution rule is proposed, which can enhance the cascading failure critical load value fault-tolerance accuracy. Meanwhile, the degree distribution and the exponential correlation parameters can effectively compensate for the influence of the cascading failure and enhance the fault-tolerant ability in scale-free wireless sensor networks.