네트워크 가상화 환경에서 MVNO 수익 극대화를 위한 TD3 기반 대역폭 동적 가격 결정 기법

초록

With 5G network development, network virtualization has seen increasing use, as it enables operation adapted to diverse requirementssuch as video streaming, autonomous driving, and smart factories. In this context, dynamic bandwidth pricing that reflects real-timedemand-supply fluctuations is increasingly important for improving resource efficiency. However, conventional static, rule-based pricinglimits to adapt to changing network states and diverse service needs. To overcome this issue, We propose a reinforcement learning baseddynamic pricing approach that sets prices from real time network information. In this paper, We formulate the interaction between amobile virtual network operator (MVNO) and users as a Stackelberg game, where the MVNO posts prices and users purchase bandwidthto maximize utility. We train the leader’s pricing policy with a Twin Delayed Deep Deterministic Policy Gradient (TD3) agent adaptedto the network environment. To demonstrate the method's superiority, We compare it with a Deep Q-learning (DQN)-based dynamicpricing model and a fixed price policy. The method acquires the highest average MVNO reward and lower variability in user utility.

키워드