An Autonomous Resource Management Model towards Cloud Morphing

This paper proposes an autonomous resource management model for future edge clouds with abundant and diverse computing resources. The model utilizes a pseudo cost function to allocate resources based on load, and a convex cost function to enable load balancing and idle-resource pooling. Stakeholders can manipulate the cost function or resource weights for micro-jobs to control resource utilization. We explore various cost tuning methods and present their performance through simulations.