Network Slicing in 5G Edge Networks with Controlled Slice Redistributions

# 72






Abstract

5G communications has led to the emergence of new services and applications that are broadly categorized as eMBB, uRLLC, and mMTC. These new services and applications have diverse and stringent quality-of-service (QoS) requirements that range from high bandwidth to ultra-low latency and high reliability. Network operators can no longer support such diverse requirements using the traditional ‘one-size fits all’ network architecture. Network slicing is one of the prominent solution approaches for differentiated service provisioning by creating multiple logical networks. Each logical network is customized based on the application-specific QoS requirements. In this talk, we will introduce the concept of network slicing in 5G. Then we will discuss the issues with network slice redistributions in 5G edge networks that motivate us to revisit the network slicing problem. We pose it as a constrained optimization problem, which is NP-hard. Subsequently, we will present our proposed solution approach that takes into account the active slice redistributions in network slicing.

Samaresh Bera, IISc Bangalore

Dr. Samaresh Bera is a C.V. Raman Post-Doctoral Fellow at the Department of Electrical Communication Engineering, IISc Bangalore, India. Before that, he was a Visiting Researcher at the British Telecom India Research Centre (BTIRC), IISc Bangalore, India. He also worked as a Research Engineer at INRIA Sophia Antipolis, France in a joint collaboration with Nokia Bell Labs, France. He obtained his PhD and MS degrees from the Indian Institute of Technology Kharagpur, India in 2020 and 2015, respectively. He was recognized as a young researcher to attend the 5th Heidelberg Laureate Forum, Germany in 2017. He also received the IEEE Richard E. Merwin scholarship for his outstanding contributions to the IEEE in 2017. His current research interests include software-defined networks, network function virtualization, and 5G network slicing.