Profit Maximization for Service Placement and Request Assignment in Edge Computing via Deep Reinforcement Learning

With the integration of Mobile Edge Computing (MEC) and Network Function Virtualization (NFV), service providers are able to provide low-latency services to mobile users for profit. In this paper, we study the problem of service instance placement and request assignment in an MEC network for a given monitoring period, where service requests arrive into the system without the knowledge of future arrivals. Each incoming request requires a specific service with a maximum tolerable service delay requirement. The problem is to maximize the profit of the service provider by admitting service requests for the monitoring period, which can be achieved by preinstalling service instances into cloudlets to shorten service delays, and accommodating new services by removing some idle service instances from cloudlets due to limited computing resources. We then devise an efficient deep-reinforcement-learning-based algorithm for this dynamic online service instance placement problem. We finally evaluate the performance of the proposed algorithm by conducting experiments through simulations. Simulation results demonstrate that the proposed algorithm is promising.