Towards Joint Optimization Problem for Computing and Resource Allocation in In-network Computing for Metaverse

Abstract

Metaverse poses a huge challenge regarding networking and computing requirements for task processing, especially on mobile devices. In-network computing (COIN) is a promising solution for enabling computing in the network device for faster completion of computationally intensive tasks and efficient use of resources. This paper examines three modes of metaverse task processing in a mobile network to formulate the joint optimization problem to improve task processing efficiency through task offloading and resource allocation. The mode includes mobile/local in-network computing (LIN), fog in-network computing (FIN) and edge in-network computing. The corresponding task completion and energy consumption models for each mode are examined. Subsequently, the task offload problem is transformed into a joint optimization problem of task completion time and energy consumption.