A Review on Task Offloading using Meta-Heuristic Algorithms on Fog Computing

Fog computing offers low latency services at a lower cost as opposed to the expensive cloud resources having long delays and response time. However, the resources on fog computing are limited, therefore an effective task offloading algorithm is essential. Traditional task offloading algorithms mostly struggle to deal with heterogeneous resources and large solution spaces, therefore a meta-heuristic algorithm with the right mix of exploration and exploitation capabilities is desired to offload tasks for better quality of service. This review presents an in-depth study of task offloading techniques utilizing meta-heuristic algorithms based on the types of tasks, offloading ratios, different architectures, constraints, and mobility of fog nodes. The meta-heuristic algorithms are categorized into single meta-heuristic algorithms, hybrid algorithms, and hybrid meta-heuristic algorithms with critical evaluation. It is observed that delay, energy consumption, and waiting time are the most investigated (50%) optimization metrics by the researchers in the selected studies. Further, it is revealed that random tasks are generated by majority of the researchers in MATLAB for evaluation of their proposed algorithms. The open issues, obstacles, and future directions are discussed at the end.