A Survey on Intelligent Attack Mitigation and Latency Optimization in High-Speed Edge Networks using Deep Learning and Bio-Inspired Algorithms

As the Internet connected devices and cyber– physical systems (CPS) are exponentially increasing, edge computing has become a crucial enabler to support real time processing in high-speed networks (HSNs). However, integrating edge computing into broadband infrastructure is a challenging problem and that includes mitigating security risks such as DDoS and time-delay attacks and keeping ultra-low latency. Explored for this survey are some recent technologies on how to detect attacks and optimize latency on the edge, focusing on forms of deep learning and bioinspired approaches. It analyzes the lack of effectiveness in traditional mitigation strategies and represents new means, based on the use of ensemble learning, dimensionality reduction and hybrid models. Distributed edge infrastructures are specially emphasized on real time implementation, computational overhead and security. This paper attempts to contribute valuable insight for researchers developing robust, scalable, and intelligent edge-based solutions in high-speed networking environments by consolidating recent methodologies and their outcomes.