A novel dynamic service function chaining to enable Ultra-Reliable Low Latency Communication in NFV

Abstract

Network Function Virtualization (NFV) is a revolutionary technology that can make traditional network architectures more efficient and flexible by decoupling software from hardware. In recent years, the use of Ultra-Reliable Low Latency Communication (URLLC) applications such as driverless vehicles, tele-operated driving, remote surgery, virtual reality, and industrial automation has increased substantially. Communication Service Providers (CSPs) must meet the requirements of URLLC applications. Providing a network service with high reliability and low latency is extremely challenging, and the limited physical network resources available to CSPs increase the complexity. To enable URLLC in an NFV-enabled environment, we provide a set of novel heuristic algorithms for dynamic service function chaining that simultaneously optimizes reliability and latency without using backup techniques. We employ flow prioritization and provide an adjustable priority coefficient factor ($\mu$) to reserve a specific amount of physical network resources (bandwidth, CPU, and memory) exclusively for the deployment of ultra-reliable low latency flows. Our proposed algorithms outperform existing methods in terms of end-to-end delay, reliability, bandwidth usage, and SFC acceptance rate, as demonstrated by the results.