Stable femtocells cluster formation and resource allocation based on cooperative game theory

Katty Rohoden, Rebeca Estrada, Hadi Otrok, Zbigniew Dziong

Research output: Contribution to journalArticlepeer-review

9 Scopus citations

Abstract

In this paper, we address the problem of forming stable groups of femtocells that can reduce the complexity of the resource management and enhance the subscribers’ satisfaction while guaranteeing the service to nearby public users. In a macro-femtocell network, the resource management becomes a very challenging task as the number of deployed femtocells increases. Several strategies of clustering have been proposed to allocate resources in a distributed manner. However, forming stable clusters of femtocells is yet to be addressed. We propose a distributed cluster-based resource allocation framework that consists of three components: (1) a base station selection algorithm for public users that guarantees them a high data rate, (2) a coalition game, where femtocells are grouped into stable clusters to reduce the resource allocation complexity, and (3) a fair resource allocation using the Shapley value to compute the payoff of each cluster member based on Particle Swarm Optimization algorithm. The ε-core concept from game theory is used as the stability criteria to form the clusters. A performance comparison is carried out between the proposed solution and two benchmark models: a centralized approach and a distributed approach with non-stable group formation. Simulation results show that our framework indeed increases the network throughput, provides higher subscribers satisfaction, and higher Jain fairness index for the distribution of resources among the existing users in the femto-tier.

Original languageBritish English
Pages (from-to)30-41
Number of pages12
JournalComputer Communications
Volume134
DOIs
StatePublished - 15 Jan 2019

Keywords

  • Clustering
  • Femtocell networks
  • Game theory
  • OFDMA
  • PSO

Fingerprint

Dive into the research topics of 'Stable femtocells cluster formation and resource allocation based on cooperative game theory'. Together they form a unique fingerprint.

Cite this