Design and evaluation of a query-based jamming detection algorithm for wireless sensor networks∗

Jamming attacks, which are the main cause of data corruption and network blockages in wireless sensor networks (WSNs), are one of the most serious threats for WSNs. This type of attack not only blocks the ongoing communication in the network, but also causes the wireless nodes to exhaust their energy much earlier than expected. A countermeasure must be deployed against jamming attacks, especially in military and medical applications in which security breaches cannot be tolerated. In this paper, we have designed a new query-based jamming detection algorithm (QUJDA) to detect jamming attacks occurring in WSNs. QUJDA is an attack detection mechanism, which uses an anomaly-based approach and operates in a distributed manner. It separates attacking cases from natural network conditions by the help of packet delivery ratio, bad packet ratio, and the amount of energy consumption parameters. QUJDA enables sensor nodes to operate with their neighbors in a collective sense in order to achieve higher detection rates. QUJDA was evaluated and analyzed over 3 parameters: detection rates, false-positive rates, and communication overheads. According to the simulation results obtained, all critical jamming attacks can be detected with 97% or above detection rates, along with 0.95% or lower false positive rates. In addition, QUJDA only charges about 13% extra communication overhead to the network message traffic of the sensor nodes.