Adaptive Packet-Combining to Counter Traffic Analysis in Wireless Sensor Networks

Wireless Sensor Networks (WSNs) have become an attractive choice for many applications that serve in hostile setup. The operation model of a WSN makes it possible for an adversary to determine the location of the base-station (BS) in the network by intercepting transmissions and employing traffic analysis techniques such as Evidence Theory. By locating the BS, the adversary can then target it with denial-of-service attacks. This paper promotes a novel strategy for countering such an attack by adaptively combining packet payloads. The idea is to trade off packet delivery latency for BS location anonymity. Basically, a node on a data route will delay the forwarding of a packet until one or multiple additional packets arrive and the payloads are then combined in a single packet. Such an approach decreases the number of evidences that an adversary will collect and makes the traffic analysis inclusive in implicating the BS position. Given the data delivery delay that will be imposed, the proposed technique is to be adaptively applied when the BS anonymity needs a boost. The simulation results confirm the effectiveness of the proposed technique.