LSTM-FFAM: utilization of long-short term memory recurrent neural network with feed forward attention mechanism for detection of replay attacks in wireless sensor networks

LSTM-FFAM: utilization of long-short term memory recurrent neural network with feed forward attention mechanism for detection of replay attacks in wireless sensor networks

연구 분야: Infrastructure

학회: International Journal of Information Technology

Wireless sensor networks (WSNs) help fight many security threats because they have limited hardware and lack infrastructure. The biggest danger comes from the reverse attack, a denial of service (DoS) attack that targets the network layer. Thus, the given paper introduces a deep learning method to spot and stop reverse attacks. The proposed approach combines long-short term memory (LSTM) with feed forward attention mechanism (FFAM) for detection of replay attacks in WSN’s. Replay attacks involve catching and sending back sent objects after some time. In a replay attack, spoofed packets traverse the path from the sensor node to the base station, resulting in simulated propagation time and spoofed signal strength. These effects lead to incorrect estimation of the receiver’s location and distance relative to the control signal’s arrival time. The performance of the proposed approach is evaluated and compared with existing recent studies based on metrics such as false acceptance rate (FAR) and Accuracy (%).