H-RAN: An Approach Toward Cloud-RAN Load Balancing

Abstract

The centralized cloud radio access network (CRAN), change the BS architecture to a large extent, in term of associated cost and coverage. The main limitations of C-RAN architecture is that it has a centralized processing unit, which increases the network latency by increasing the service waiting time. The proposed Heterogeneous Radio Access Network (HRAN) consist of some Edge node and local processing unit (LPU) which resides in remote radio head (RRH). This introduction of enhanced RRH (eRRH) node increases the capacity of the network and decrease the network latency, on the other hand its also share the computing and controlling task of the centralized BBU. The proposed LPU unit consists of a number of small VB which generally handle real-time computing and processing task.In the first section of this paper, we have proposed a new RAN architecture and all the load balancing condition are represented through a detail mathematical model. The paper also present a load-aware dynamic BS switching algorithm (LDBSS) for power optimization and load balancing. The simulation work is carried out with the help of CloudSim 3.0.3 simulator by considering equal number of packet arrival for both C-RAN and H-RAN. The simulated result shows that with the introduction of the middle Edge layer and LPU at the RRH reduces the waiting time and blocking probability of the network as compared the C-RAN architecture.