Abstract: Based on satisfying the requirements of end-to-end delay and reliable service among nodes,to solve the problem of high energy consumption of existing multi-path routing protocols,a quality of multi-service(QoS) constrained route algorithm for edge computing and node energy optimization(MQEN) was proposed. The QoS constraints of end-to-end delay,reliability,and energy expenditure were considered. The related technologies of edge computing and machine learning were utilized to create a sensor network model of multi-constrained majorization path,introducing a wake-up strategy of energy-aware node as well as reward and punishment mechanism based on learning automaton. This algorithm combined edge computing to preprocess the original data of the node,speeding up effective data transmission and treatment. The automata-environment interaction approach was adopted to accelerate algorithm convergence. The technique of sleep activation of control node was employed to optimize network power consumption and extend the network life cycle. Experiments indicate that the MQEN algorithm reduces network power expenditure and corresponds to the demands of multiple QoS constraints for the end-to-end delay and credible services.

Key words: wireless sensor, edge computing, learning automata, machine learning, energy-aware