A survivability model in wireless sensor networks

Abstract

In this paper, we present a survivability evaluation model and analyze the performance of Wireless Sensor Networks (WSNs) under attack and key compromise. First, we present a survivability evaluation model of WSNs by representing the states of WSNs under attack. Here, the survivability of WSNs is expressed as a continuous time Markov Chain to describe the status of real WSNs. We propose a threshold condition to trigger the transition between the states of WSNs in the face of DoS attack. According to the threshold conditions, the survivability model is capable of taking action(s) and enables the formalization of the effect of action(s). We present an analytical model and numerical analysis to show the feasibility of our work in terms of availability, survivability, and cost.Second, we present a survivability model and evaluation of WSNs under key compromise. Sensor nodes in a network can be captured by an adversary. The compromised node can be identified by one of the intrusion detection mechanisms. After detecting the compromised node, the key in the node has to be revoked. In this paper, we present the effect of key compromise and key revocation in WSNs. We additionally present a survivability evaluation model using the SHARPE simulation tool in terms of availability.