Radio resource coordination and scheduling scheme in ultra-dense cloud-based small cell networks

Zengxian Chen, Sai Zou, Yuliang Tang, Xiaojiang Du, Mohsen GuizaniEURASIP J. Wirel. Commun. Netw 2018 · DOI

Introduction

The transition to 5G and beyond necessitates the deployment of Ultra-Dense Networks (UDNs) to accommodate a 1000-fold increase in data traffic. However, the proximity of numerous small cells introduces significant challenges, primarily inter-cell interference and excessive signaling overhead. Traditional centralized management via Cloud-RAN often suffers from high latency and signaling congestion when handling real-time local adjustments. This paper proposes a hybrid Radio Resource Coordination and Scheduling (RCS) scheme that leverages both cloud and fog computing to optimize network performance.

System Architecture

The proposed architecture integrates Cloud-RAN with localized fog computing to create a two-tier management system. The cloud level, consisting of a centralized Base Band Unit (BBU) pool, handles global resource planning. Meanwhile, fog devices are deployed closer to the network edge to manage clusters of small cells, providing low-latency responses to local traffic fluctuations, often referred to as the "tidal effect."

Fig. 1: System network model illustrating the hierarchy of Cloud, Fog, and Small Cell layers.

The Hybrid RCS Scheme

The RCS scheme operates through a coordinated effort between the two tiers:

  1. Global Tier (Cloud): The BBU pool utilizes a global interference graph to pre-allocate Physical Resource Blocks (PRBs) using a flexible graph-coloring approach, ensuring that adjacent cells minimize frequency overlap.
  2. Local Tier (Fog): Fog devices monitor real-time traffic. When a cell requires additional resources, the fog device facilitates dynamic "borrowing" or "transferring" of idle PRBs from neighboring cells.
Fig. 2: Function block diagram showing the interaction between global resource allocation and local coordinated scheduling.

To ensure stability and efficiency, the system employs a Bayesian scoring model to predict the interference impact of resource transfers. Furthermore, a "PRB Value" is calculated to ensure that any resource transfer results in a net gain for the network throughput, accounting for the potential rate loss of the donor cell.

Performance Evaluation

The scheme was evaluated in a simulated 600m x 600m environment containing 64 small cells (177 cells/km²). Key findings include:

  • Throughput Gains: The hybrid approach increased system throughput by 33% compared to traditional graph coloring and 20% compared to clustering-based algorithms.
  • Signaling Efficiency: By handling local adjustments at the fog level, signaling overhead was reduced by approximately 50%.
  • QoS Stability: The system maintained superior Quality of Service (QoS) even as user density increased, outperforming traditional methods which typically saturate at lower user counts.

Conclusion

By splitting the workload between centralized cloud processing and localized fog computing, the proposed RCS scheme effectively mitigates interference and signaling bottlenecks in UDNs. Future research will address constraints in backhaul capacity and the complexities of high-speed user mobility.

Summary updated

BibTeX
@article{chen2018radio,
  author = {Zengxian Chen and Sai Zou and Yuliang Tang and Xiaojiang Du and Mohsen Guizani},
  title = {{Radio resource coordination and scheduling scheme in ultra-dense cloud-based small cell networks}},
  journal = {EURASIP J. Wirel. Commun. Netw.},
  year = {2018},
  doi = {10.1186/s13638-018-1150-1},
}
Figure