SECURE KEY AGREEMENT AND AUTHENTICATION MECHANISM WITH IDS FOR FOG COMPUTING

Abstract

Cloud computing technology was first introduced by Amazon in 2006 and has provided customers with high-quality services through the Internet. However, the growing number of IoT and mobile devices has led to challenges such as latency, power consumption, and network strain. To address these issues, Cisco developed fog computing, which enhances cloud capabilities by processing data closer to the network edge. Despite its advantages, security concerns remain a critical challenge within the fog computing paradigm. The issue with secure authentication in fog often involves balancing the need for robust security measures with the constraints of limited resources and varying trust levels among distributed devices. Moreover, the security protocols are weak against unknown threats, leaving the system vulnerable to potential attacks. To address these issues, the study proposes an online sequential random vector functional link (OS-RVFL) neural network-based authentication approach. This method adapts to security threats in real time, improves authentication resilience, and efficiently uses available resources. This intrusion detection model is integrated into the proposed lightweight authentication protocol to enhance the overall security framework in the fog environment. This ensures a dynamic response to emerging threats while maintaining a low resource footprint across the fog computing environment. The proposed system can also learn and update in realtime without depending on the previous training data batch. The security of the proposed scheme has been rigorously analyzed using the realor-random model, which provides formal proof of its robustness. Furthermore, the scheme has been verified using the widely accepted AVISPA tool.