Association Control in the 5G Network with Cloud Radio Access Network Architecture

Abstract

With the popularity of mobile devices and the advance of Internet of Things the fifth generation mobile network (5G) requires to meet large amount of service requests in the heterogeneous networks For this purpose flexible network architectures with dynamic resource management such as Cloud Radio Access Network (C-RAN) and Software-Defined Networking (SDN) are widely studied in recent years In C-RAN a novel mobile network architecture Baseband Units (BBUs) are centralized virtualized and shared among Remote Radio Units (RRUs) to optimize the resources utilization of base stations Compared to the traditional networks SDN separates control plane and data plane to provide a more flexible and programmable network architecture In this centralized architecture services can be optimally managed in a global view However existing association control schemes are not designed for the centralized network architecture Besides performance issues occur when the amount of devices increase significantly with limited wireless resources With the large amount of devices severe collisions occur under one base station and load imbalance among base stations happens without a centralized association control At the edge between cells ping-pong effect may cause unnecessary handovers and signal overload Therefore how to design an efficient association control scheme under 5G centralized network architecture to handle the overloaded demands is required to be studied The idea of grouping devices is proposed in this dissertation to reduce the computation from the number of devices to the number of group in the association schemes The collisions at a base station can be reduced by the scheduling in the group When groups of devices move among base stations SDN provide global view to find optimal coverages of base stations to increase throughput At the edge of cell the execution of handovers are postponed to avoid unnecessary handovers This dissertation proposes three schemes to handle the issues mentioned above: collisions load balancing and ping-pong effect Mathematical analysis and simulation experiments are conducted to evaluate the performances of these proposed schemes The results show that the impacts of the issues are reduced with little delay when the demands are significantly higher than the resources that system can supply The contribution of this dissertation is to provide some suggestions to operators at the early stage of 5G deployment

Date of Award	2019
Original language	English
Supervisor	Meng-Hsun Tsai (Supervisor)