AI Based Network Slicing In 5G Network

Abstract

5g is the future of mobile communications. Upcoming networks needed high quality standards with reliability and very low latency. The main feature of upgrading into 5g networks is to have very high end user connectivity. Optimizing network slicing can be more helpful to achieving the remarkable performances in 5g networks. Network slicing divides the physical network into several logical networks in order to support the variety of developing applications with higher performance and flexibility needs. Recently data driven decisions are more efficient and useful for different industries, the same can be applied to communications industries also. Due to the use of different applications, an enormous number of mobile phones have created a huge amount of data, which creates incredible difficulties and significantly affects the effectiveness of network slicing. In this paper we have proposed an AI based Model which can handle the network slicing very efficiently and is also capable of handling load efficiency and network availability. Using various 5g slicing datasets our model has been trained and capable to predict network slices for unknown device types by analyzing the incoming traffic. It also handles the situation of load balancing in the case of a network failure.