Effective Load Balancing and Load Sharing in Multi-access Edge Computing for Extreme Network Congestion

Abstract

A multi-access edge computing (MEC) plays a vital role in dealing with the exponential growth of data traffic to avoid network congestion. This is also essential for fulfilling the user requirements of computation resources. MEC network brings cloud functionalities to the edges that are in close proximity to the devices supporting multiple access points. Mobile devices that have resource constraints can save energy and enrich users’ experience using the computation load balancing technique. This paper provides a comprehensive approach to load balancing and load sharing in a MEC environment. We proposed two algorithms for the MEC framework. First is the proactive load rationalization technique that accounts for the load as a function of chaos resolution through user load prediction. The second algorithm is used for the distribution of data across servers in a given geographical region which aims to achieve an equilibrium that ensures no server is overworked while other servers are left idle. Finally, a fine-grained computation-based MEC framework is developed using our proposed algorithms that is capable of load balancing and sharing in extreme network congestion. The ultimate goal is to provide optimal resource services to each user while balancing the load among the MEC servers. The simulation results demonstrate the efficiency of our methods in balancing load in the real-time scenario. Moreover, it has the ability to produce minimal load overhead during load failure.