ENERGY-EFFICIENT INTRUSION DETECTION IN WSN: LEVERAGING IK-ECC AND SA-BILSTM
Keywords:
Smish Activated-Bidirectional-Long Short Term Memory (SA-BilSTM), Identity-Key-based Elliptic Curve Cryptography (IK-ECC), Ring Sign-based Elliptic Curve Digital Signature Algorithm (RS-ECDSA), Skew Tent-based K-means (ST-KM), False Alarm rate (FAR)Abstract
Wireless Sensor Networks (WSNs) play a vital role in numerous applications. This paper proposes an energy-efficient intrusion detection framework for WSNs. Initially, nodes are registered to the network. During registration a unique key is generated for each node. Then, the nodes are formed into cluster and for each cluster CH and SCH are selected. Thereafter, the path is created between the clustered nodes and the CH. After that, data transmission begins. To attain secure data transmission, the unique signature is generated for each data using RS-ECDSA and then the data is encrypted by IK-ECC. The encrypted data is transmitted to the sink node (SN), which acts as the gateway to base station (BS). In the SN, subnet masking and batch verification is performed. After that, IDS using Bi-LSTM is employed this predicts whether the data is attacked or not. The experimental result stated that the proposed method withstands energy efficient as compared to existing methodologies.
