Utility-Optimized Resource Allocation in Computing-Aware Networks

Abstract

The rapid development of 5G and Artificial Intelligence technologies has brought various applications with specific requirements in both computing performance and network performance. At the same time, ubiquitous computing devices are interconnected to the Internet Service Provider’s network. Computing-aware networking, as the integration of networking and computing, has become a hot topic in both academia and industries. Motivated by this new network architecture, we present a Computing-aware Bandwidth Allocation method based on utility optimization, called CABA. The proposed method enables the network to schedule eligible computing nodes to serve the requested service, and coordinate network resources (e.g., network cost, bandwidth) so that the traffic can be routed along the desirable path. Experimental results show that our proposed method can achieve high utility in relation to bandwidth allocation and provide service scheduling scheme leveraging both computing capability and network cost.