A fuzzy-based en-route filtering scheme in sensor networks

Most of the sensor networks use a symmetric protocol since sensor networks are comprised of sensor nodes with restricted hardware. Sensor networks with symmetric cryptography contain a global key stored on each sensor node and may be deployed in a hostile environment. When sensors nodes are compromised, an attacker can inject false sensing reports or false Message Authentication Codes into a legitimate report. A probabilistic voting-based filtering scheme has been proposed to combat these threats from compromised nodes. However, this scheme has the problem that it cannot re-establish a session key when some nodes of the source cluster or some intermediate cluster head have been compromised. The scheme cannot also control the position of verification nodes to minimize energy consumption as topology changes. Therefore, we propose a fuzzy-based en-route filtering scheme to deal with these problems. Through performance analysis and simulation, our result shows that the proposed scheme is much more efficient than the probabilistic voting-based scheme in many cases.