Semi-Clustering of Victim-Cells Approach for Interference Management in Ultra-Dense Femtocell Networks

Network densification is seen as a necessary trend in the evolution toward 5G networks to cope with future high-traffic demands. In particular, the use of femtocells seems to be the most feasible solution for hotspot areas where high traffic is concentrated. However, femtocell densification will face a number of technical challenges. Among others, the co-tier interference arises as an old acquaintance with a new significance. In point of fact, the interference characteristics and the role of aggressor and victim depend on a large extent on the density scenario, especially when femtocells share the same channel and operate in a Closed Subscriber Group (CSG) mode. In this regard, identification of victim and aggressor femtocells is a prerequisite to design an effective interference management scheme for ultra-dense femtocell networks (UDFNs). This paper implements a new approach for radio resource utilization and management for victim femtocells in an ultra-dense network environment. The proposed semi-clustering of victim-cell (SCVC) approach focuses mainly on users' status whether critical or non-critical to categorize victim femtocells and their aggressors. After identifying the victim femtocells, the SCVC scheme semi-cluster each victim femtocell based on the status of each user. Then, it smartly allocates the appropriate radio resources in a dynamic manner, which allows ranging from shared use of resources for critical users to the frequency reuse one for non-critical users, thereby ensuring minimum co-tier interference and enhancement of spectrum efficiency. Simulation results show that our approach outperforms one of the prominent related schemes, namely, femtocell cluster based resource allocation (FCRA), in terms of the mean throughput of critical users, the average capacity of victim femtocells, and the percentage of resource utilization by an approximate average gain of 185%, 64%, and 31%, respectively. © 2017 IEEE.