Exploiting Collaborative Behavior for User QoE Enhancement in Device-to-Device Networks

Abstract

The emergence of device-to-device (D2D) communication has become crucial in the evolution of 5G systems, enabling efficient distributed communication. D2D communication optimizes spectrum and energy utilization, reduces transmission latency, and alleviates base station traffic by establishing direct connections between nearby devices. However, in large and intricate D2D networks, optimizing the number of reliable links poses a significant challenge given the impact on communication quality. This paper introduces the computation-enhanced fireworks algorithm (CEFA) as a means to enable collaborative communication among devices, facilitating informed decision-making for communication paths to benefit network users. Our approach takes into account the communication behavior of nodes to assess link quality and its direct influence on user's quality of experience (QoE). A graph-based communication system is utilized, where devices act as nodes and the graph edges indicate the connection between them. Additionally, we employ a weighted graph to gauge connection reliability, with edge weights representing the mean opinion score (MOS) value. Optimizing the number of reliable links enhances connection quality and ensures QoE. Through CEFA, nodes can effectively and collaboratively reorganize the graph to facilitate communication. The scheme utilizes an elite selection strategy and mutation for communication diversity coordination, along with a mapping operator for media content sharing among devices. A fitness function is designed to evaluate the effectiveness of the proposed scheme, and extensive simulations demonstrate its efficiency even with increased nodes and communication complexity.