Dynamic Traffic Model With Optimal Gateways Placement in IP Cloud Heterogeneous CRAN

Haitham H M Mahmoud,M Saeed Darweesh,Tawfik Ismail

doi:10.1109/access.2020.3004421

Abstract

In this paper, topology design, optimal routing, and gateways placement selection algorithms are proposed in Heterogeneous Cloud Radio Access Network (C-RAN) with exploiting Free Space Optical (FSO) communication. The proposed network consists of two tiers; the lower tier concerns with clustering Remote Radio Heads (RRHs) based on traffic demands. The upper tier consists of transceivers along with the Cluster Heads (CHs) and gateways. Algorithms are proposed to achieve the lowest number of edges and the highest possible throughput based on the presented optimization problem. Moreover, route optimization and gateway selection problems have been considered in two different traffic scenarios; static (maximum occupancy) or dynamic (Partial occupancy). Furthermore, a disaster recovery algorithm is proposed in case of failure of one of the nodes or edges to maintain the continuous connectivity of the network at the expense of the traffic. Our simulation results show that the number of gateways in a dynamic traffic operation is reduced by 33% in node size of 25, 36, and 42. Also, this reduction is increased to 41% if the node size increased to 49.

Highlights

In the Cloud Radio Access Networks (C-RAN), the cellular fronthaul networks provide connectivity between Remote Radio Heads (RRHs) and Base Band Processing Units (BBUs)
The architecture is based on dual-tier C-RAN, where several RRHs are connected to the adjacent cluster head (CH) that acts as a router, and some of these Cluster Heads (CHs) are considered gateways that can transfer the data to BBUs
We investigate the best placement of the gateways and routing paths in RAN to achieve optimal throughput

Summary

INTRODUCTION

In the Cloud Radio Access Networks (C-RAN), the cellular fronthaul networks provide connectivity between Remote Radio Heads (RRHs) and Base Band Processing Units (BBUs). Topology design, optimal routing, and gateways placement selection algorithms are realized in this work with FSO technology to achieve the demand needs. The architecture is based on dual-tier C-RAN, where several RRHs are connected to the adjacent cluster head (CH) that acts as a router, and some of these CHs are considered gateways that can transfer the data to BBUs. In this paper, we investigate the best placement of the gateways and routing paths in RAN to achieve optimal throughput. A small selected fraction of CHs are acting as gateways with a connection to the BBUs. The traffic to/from the RRHs relays through multi-hop routes until reaching one of the gateways and to BBUs. The DT algorithm generates the distribution of RRHs and CHs in the form of clustered districts (see Figure 3). This work aims to maximize the total throughput of the network, whereas distinct minimum traffic from each RRH should be satisfied

FSO LINK RELIABILITY

DISTANCE CLUSTERING AND TRAFFIC GENERATION ALGORITHM

SIMULATION AND RESULTS

1: Identify the Grid boundaries

CONCLUSION AND REMARK