Revenue Maximization of a Slice Broker in the Presence of Byzantine Faults

Abstract

Multi-Access Edge Computing (MEC) and network slicing are vital for advancing the Fifth Generation (5G) of cellular systems. MEC provides context awareness and reduces the latency for communication. Network slicing allows the division of a single network into multiple virtual networks so that different services can be provided. A slice broker is a business entity that buys the resources from the infrastructure providers and sells them to the tenants. A tenant sends a request for resources for different slices. In this work, we formulate the slice allocation problem to increase the revenue for the slice broker. We formulate a dynamic demand model based on the set price changes. We consider the profit maximization of a slice broker in the presence of Byzantine faults. Moreover, we propose the Comparative Gradient Elimination (CGE) method in Federated Learning (FL) for revenue maximization of the slice broker. Simulation results show that our proposed method outperforms the reference solution.