QoE-based dynamic resource allocation for heterogeneous smart distribution grids

In the realm of conventional smart distribution grid resource allocation, the prevalent issue resides in its narrow focus on base station capacity, striving to optimize resource allocation for base station communication while disregarding the genuine requirements on the user side. This inadvertently leads to excessive squandering of wireless resources, despite already fulfilling the fundamental service demands of terminal operations. This article, while taking into account the capacity of base stations, introduces an innovative approach by amalgamating the terminal operations concerning data rate, latency, and packet loss rate. Through the construction of a Quality of Experience (QoE) evaluation framework, a scenario is realized within which user experience requirements are ensured by terminals in various practical settings of smart distribution grids, without wireless resources being needlessly dissipated. In this article, the dynamic resource allocation is tackled using Deep Q-Networks (DQN), while the reward function is formulated based on QoE. The simulation results, which track the accumulation of reward values throughout the entire operational process, provide substantial validation for the effectiveness and practicality of the ultimately formulated dynamic resource allocation scheme.