Dynamic Clustering and User Association in Wireless Small Cell Networks with Social Considerations

TL;DR

This paper introduces a social network-aware user association method for wireless small cell networks with D2D communication, optimizing network performance by considering social ties and physical proximity through a dynamic clustering and matching game approach.

Contribution

It proposes a novel social-aware user association framework that incorporates social relationships into the matching process in small cell networks with D2D, improving throughput performance.

Findings

Achieves up to 31% throughput improvement at the 95th percentile.

Effectively incorporates social ties into user association decisions.

Demonstrates the viability of social-aware clustering and matching in wireless networks.

Abstract

In this paper, a novel social network-aware user association in wireless small cell networks with underlaid device-to-device (D2D) communication is investigated. The proposed approach exploits social strategic relationships between user equipments (UEs) and their physical proximity to optimize the overall network performance. This problem is formulated as a matching game between UEs and their serving nodes (SNs) in which, an SN can be a small cell base station (SCBS) or an important UE with D2D capabilities. The problem is cast as a many-to-one matching game in which UEs and SNs rank one another using preference relations that capture both the wireless aspects (i.e., received signal strength, traffic load, etc.) and users' social ties (e.g., UE proximity and social distance). Due to the combinatorial nature of the network-wide UE-SN matching, the problem is decomposed into a dynamic clustering problem in which SCBSs are grouped into disjoint clusters based on mutual interference. Subsequently, an UE-SN matching game is carried out per cluster. The game under consideration is shown to belong to a class of matching games with externalities arising from interference and peer effects due to users social distance, enabling UEs and SNs to interact with one another until reaching a stable matching. Simulation results show that the proposed social-aware user association approach yields significant performance gains, reaching up to 26%, 24%, and 31% for 5-th, 50-th and 95-th percentiles for UE throughputs, respectively, as compared to the classical social-unaware baseline.