Edge Cloud Capacity Allocation for Low Delay Computing on Mobile Devices

Newly-emerged edge computing paradigm advocates deploying one or several server-class machines at the edge of network to augment computing ability of mobile devices and save network bandwidth simultaneously. Compared with conventional mobile cloud computing, mobile edge computing is constrained in computing capacity, especially when serving a large number of mobile users. To address the above concerns, we formulate the Edge Cloud Capacity Allocation (ECCA) problem in this paper, where a certain amount of resources should be explicitly allocated to each edge cloud for optimal QoS provisioning. We firstly introduce two algorithms, namely, GACA (genetic algorithm based capacity allocation algorithm) and PSOCA (particle swarm optimization based capacity allocation algorithm) for the ECCA problem. Due to the unsatisfactory convergence speed of the previous two algorithms, we then propose IMMF (improved max-min fairness algorithm), using a combination of heuristic and mathematical optimization to solve the ECCA problem. Computational simulations confirm the effectiveness of IMMF.