Swarmio Media

• Led the design of the Latency Optimized Edge Cloud solution tailored for edge-user allocation in esports.• Proposing a sub-optimal solution to the multi-objective optimization Reinforcement Learning, and proposed modifications on Q-learning to optimize for fairness: action suppression, Quadruple Q-learning, and Local Response Normalization with Softmax.• Built Python-based scripts to collect the first Cloud Gaming Client-Server Delay Dataset (CGCSDD) from 189 users connecting to 11 servers worldwide, as well as python-based simulations of the RL environment tested the proposed model• Designed a CNN based model that estimates delay between network nodes and obtained accuracies up to 96%.