Simulation-Based Evaluation of a Latency and Fairness-Aware Edge Computing Architecture for Competitive Online Gaming

Competitive online gaming requires ultra-low latency and fairness to ensure responsive and fair experiences of players. The traditional cloud-based designs often cannot meet such requirements due to long round-trip paths, network jitter, and congestion. Edge computing has become an interesting solution to the problem by locating it near users. However, many solutions in existence do not take into consideration practical considerations like latency spikes, player mobility and fairness in scheduling. This paper presents a latency aware and fair edge-computing architecture that has been specifically tailored to competitive online gaming. Distance-aware jitter modeling, mobility-aware latency estimation, multi-server implementation, and fairness-aware scheduling are implemented on the proposed system to reduce the average and tail latency. The performance is evaluated through a Monte Carlo simulation model in Python which is based on 200 iterations per scenario. Findings of simulation show that the proposed system saves up to 30\% and 99th percentile latency over a single-server system in a set of simulation experiments, and the Jain Fairness Index of the suggested system remains above 0.97 at different player loads. Despite the simulation nature of assessment, the trends that are observed are consistent with the actual edge-computing behavior, thus offering practical values of designing fair and low-latency gaming architecture.