Performance Evaluation in Single or Multi-Cluster C-RAN Supporting Quasi-Random Traffic

Published online: May 18, 2020
Full Text: PDF (5.82 MiB)
DOI: 10.24138/jcomss.v16i2.1013
Authors:
Iskanter-Alexandros Chousainov, Ioannis Moscholios, Alexandros Kaloxylos, Michael Logothetis

Abstract

In this paper, a cloud radio access network (C-RAN) is considered where the remote radio heads (RRHs) are separated from the baseband units (BBUs). The RRHs in the C-RAN are grouped in different clusters according to their capacity while the BBUs form a centralized pool of computational resource units. Each RRH services a finite number of mobile users, i.e., the call arrival process is the quasi-random process. A new call of a single service-class requires a radio and a computational resource unit in order to be accepted in the C-RAN for a generally distributed service time. If these resource units are unavailable, then the call is blocked and lost. To analyze the multi-cluster C-RAN, we model it as a single-rate loss system, show that a product form solution exists for the steady state probabilities and propose a convolution algorithm for the accurate determination of congestion probabilities. The accuracy of this algorithm is verified via simulation. The proposed model generalizes our recent model where the RRHs in the C-RAN are grouped in a single cluster and each RRH accommodates quasi-random traffic.

Keywords

Cloud, radio access, call blocking, product form, quasi-random
Creative Commons License 4.0
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.