Defending Sensor Worm Attack Using Software Diversity Approach

Recently, the sensor worm attack has been identified to be one of the serious threats to the wireless sensor networks. However, sensor nodes do not have complicated hardware architectures or operating systems to protect program safety. Sensor worms can exploit the vulnerabilities of sensor nodes, such as the vulnerability of the buffer-overflow. In this paper, we utilize a software diversity approach to defend the sensor worm attacks. A Worm Attack DEfence (WADE) problem, which is to minimize the total number of defective edges with limited software versions, is defined in this paper. To solve the WADE problem, we present a role-base graph coloring scheme. Simulation results illustrate the efficiency and efficacy of our approach.