A Novel Multi-Party Authentication Scheme for FCN-based MIoT Systems in Natural Language Processing Environment

Abstract

Natural language processing(NLP) assists to increase the efficiency of human andMultimedia Internet of Things(MIoT) interaction. Notably, large-scale NLP tasks can be offloaded from a cloud server to fog nodes closer to a mobile terminal device for lower response latency. But communication security is ongoing issues that need to be addressed. Effective mutual authentication among multiple entities is essential to ensure the security of MIoT systems based on a dynamicFog Computing Network(FCN). However, the existing schemes are unsuitable for the dynamic FCN due to the security vulnerabilities such as the linkable sessions. To solve this problem, anAnonymous Multi-Party Authentication(AMPA) scheme is proposed to address the challenges of secure FCN-based MIoT communications in this paper. The proposed scheme uses a bilinear pairing operation to realize the authentication between the fog nodes and cloud server and to establish the group key. Besides, the scheme allows cloud-authenticated terminal devices to be added to the FCN and reduces the need for the resource-limited terminal device to perform many authentication protocols. The security analysis is carried out to demonstrate that AMPA scheme can meet various safety requirements. Performance evaluations shown that the proposed AMPA scheme achieves satisfactory performance.