A taxonomy of AI techniques for 6G communication networks: Intelligent transportation system

From the past few decades, Artificial Intelligence (AI) becomes the most powerful technology in solving many challenging problems such as route management, congestion control, topology management, security, and privacy of the communication network. Initially, it was accepted for network monitoring and diagnosis [1], but later, it is being used to optimize the network topology of large scale distributed systems using heuristic algorithms to produce an optimal solution. Non-optimal solutions were otherwise considered impossible to solve due to the com- binatorial explosion problem [2]. The aforementioned implementations of AI were considered as its initial usages during the pre-2000 era. But, since its inception, affordable computing was made possible by a large number of Graphics Processing Units (GPUs) and Compute Unified Device Architecture (CUDA) cores diversified the role of AI. Instead of simply designing and optimizing the network topologies, AI can now be used to monitor and secure the network also. Hardcoded algorithms can be used to monitor the network to determine whether any malicious activity in the network occurs. It can also monitor the network load and traffic which helps to prevent the Denial-of-service (DoS) attacks. But, since the network-related algorithms are always predefined, so by applying the AI techniques to such algorithms, we can prevent many attacks on the network. This continuous updates in network algorithms using AI techniques can remove various vulnerabilities in the networks.