Energy-Saving Service Scheduling for Low-End Cyber-Physical Systems

Energy consumption and timely requirements are two key factors affecting the performance of mission-critical cyber-physical systems. Little work deals with scheduling a set of time-sensitive services in a finite time interval. We consider a server serving N users in time interval [0, T]. Each user demands its service to be completed in a strict deadline. Based on convex power-speed relationship, a lazy schedule policy is designed to minimize energy expenditure of N mission-critical services without violating temporal constraints. Simulations show its significant energy efficiency comparing to a non-lazy schedule.