An Efficient Intrusion Detection System and Data Security Using Kcccr-gru and Rfiecc With Blockchain for Iot

The Internet of Things (IoT) has transformed numerous sectors by facilitating smart devices for collecting and exchanging data. Nevertheless, owing to insufficient data protection, numerous IoT systems face security risks. To overcome these issues, an efficient intrusion detection system and data security using KCCCR-GRU and RFIECC with Blockchain (BC) for IoT is proposed in this paper. The router details are entered during router registration. By using Unicode Transformation Format 32-Universally Unique Identifier (UTF32-UUID) and Entropy Naor Reingold Pseudo Random Number Generator (ENR-PRNG), the Service Set Identifier (SSID) and router Personal Identification Number (PIN) are generated, respectively. Further, for both router registration and the initialized IoT sensor node, public and private keys are generated. By using Exponential Public Keys-MenezesQu Vanstone (EPK-MQV), a secret key is created. Then, the SSID and PIN are combined by the Exclusive OR (XOR) operation and then hashed with Exclusive NOR (X-NOR) Inversion-HAVAL (XNORI-HAVAL). After that, the resultant hashcode is stored in the BC. Simultaneously, the SSID is stored as nonce in the BC. During transmission, the hashcode of the connected router of the IoT sensor node is verified with the hashcode in the BC. If the verification is successful, then the device is authenticated by utilizing the Exponential Public Keys-MenezesQu Vanstone-Message Authentication Code (EPK-MQV-MAC). Thereafter, by using an IDS, data from the IoT node is collected and tested. The dataset undergoes preprocessing in which features are extracted. These features are selected utilizing Linear Scaling-based Walrus Optimization Algorithm (LS-WaOA) and further classified with Kendall Correlation Coefficient Concatenated Rectified Linear Units Gated Recurrent Unit (KCCCR-GRU). Lastly, using Robust Frobenius Isogenies Elliptic Curve Cryptography (RFIECC), non-attacked data is encrypted. Thus, the proposed approach achieves a high level of data security with 98.16667% accuracy. The outcome of the research in terms of societal benefits include increased security and detection of malicious nodes for a safer computing environment.