FURI | Fall 2026, Needs Review

Computational Analysis of an Improved Reinforcement Learning Queueing Algorithm

Memory complexity is a major limiting factor when implementing reinforcement learning (RL) algorithms to service large queueing systems. This research reports the results of a new RL algorithm, with memory complexity 2^N, where N is the number of queues in the system, which is a significant improvement on the previous minimum memory complexity found in the literature, U^N, where U is the queue buffer size. Such a reduction could significantly lower the cost of running servers in data centers, thus potentially lowering the cost of commonly used internet services.