Small cells deployment for traffic handling in centralized heterogeneous network

Abstract

As the next phase of mobile technology, 5G is coming with a new vision that is characterized by a connected society, in which everything will be effectively connected, providing a variety of services and diverse business models that require more than just higher data rates and more capacity to target new kinds of ultra-reliable and flexible connection. However, next generation of applications, services and use cases will have extreme variation in requirements which in turn amplified the demand on the network resources. Therefore, 5G will require a whole new design that take into consideration efficient resource management and utilisation. An observation that was made throughout this research refers to the demand for more capacity, reduced latency, and increased density as common factors of many of the next generation use cases. This inescapably implies that the use of small cells is an ideal solution for next generation applications requirements, provided that the necessary storage and computing resources need to be distributed closer to the actual user. In this context, this research proposed an architecture of a centralised heterogenous network, consisting of Macro and Small cells with storage and computing resources, all controlled by a centralized functionality embedded within a gateway at the edge of the network. Compared to the basic network, the proposed solutions have been proven to provide overall system performance enhancement. This involves extending the system by adding small cells to serve dedicated services for User Equipment (UE) with dual connectivity from local server which reduces the overall system delay while increasing the overall system throughput. The added centralized mobility management was proven to be capable of tracing the mobility of the UEs within the system coverage, by keeping one connection with the main cell while moving between small cells resulting in enhancement to the handover delay by 11% without service interruptions. Finally, the proposed slicing model demonstrated the system’s ability to provide different levels of services to users based on different Quality of Service (QoS) requirements and to differentiate between various applications without affecting the performance of other services, benefiting from more flexible infrastructure than the traditional network. In addition, a 50% improvement in the performance was observed in terms of the CPU utilization. In such architecture, the required capacity can be added exactly where it is needed and when it is needed, coverage problems can be directly addressed, higher throughput, lower latency, and efficient mobility management can be achieved as a result of efficient resource management and distribution which is one of key factors in the deployment of next generation mobile network system.