Secure application-centric service authentication with regression learning for security systems in smart city applications

Smart city applications rely on different security paradigms for meeting the user demands and authenticated service disseminations. Diverse applications require different security modifications for improving the smart city contract-level application support. The challenging task is security adaptability and its improvements for smart city scenarios. In this article, a Secure Application-Centric Service Authentication (SACSA) is introduced for leveraging end-to-end authentication. This scheme introduces group key-based authentication for securing services in an end-to-end manner. The proposed scheme administers security using batch keys to improve the sharing efficiency of different services. The security and service time rely on the application type and distinct intervals, providing less complex and time-consuming security. In this process, blockchain is applied to perform the grouping, key generation and authentication recommendation in collaboration with the regression learning. Through this learning, batch consecutiveness is identified for improving application security. In the proposed scheme, authentication and key generation are performed using the Merkle Hash tree to prevent replication and decrease distribution. The proposed scheme's performance is analysed using the metrics authentication time, complexity, service failure, and service latency. Thus, the SACSA system maintains system security with minimum authentication time, complexity, service failure, and latency of 9.45%, 7.75%, 9.2%, and 9.39%, respectively.