Online Games Servers Placement in Fog Computing: an Hybrid Bio-inspired Approach
Résumé
Fog computing is a promising paradigm to enable online games at the rim of the network, especially for First Person Shooter games with a restrict deadline. In this context, server placement is a fundamental issue. The latter concerns where game services providers should place their servers in order to meet players expectations in terms of quality of experience while maximizing their profit. In this paper, we investigate the server placement problem, at the fog landscape, to minimize the allocation cost and the potential penalty in terms of resources due to the limited fog capacity and deadline violation. We formulate this problem as an integer nonlinear programming. As the latter is time consuming, we propose an hybrid bio-inspired heuristic composed of Grouping Genetic Algorithm and Particle Swarm Optimization. Performance results demonstrate that our proposal can generate solutions at reduced time with a gap to optimal less than 5%.