Energy Efficient Resource Allocation using Reinforcement Learning in SDN Based Wireless Networks

摘要

A lot of factors are involved in the transmission of data in an SDN based wireless network as it holds the power of combining various optical small cell wireless networks together with the WiFi, LiFi, 5G and others in a single network domain. The main issue that needs to be addressed in such a scenario is the huge demand of energy consumption by the resources that need to run the overall network structure to maintain both the Quality of Service and speed. the energy usage is investigated in an SDN based wireless network and propose a solution to effectively reduce the energy consumption of user device without compromising on these parameters. But this task becomes quite challenging due to continuous variations in the number of resources needed and type of each needed at a particular user device. Herein, Reinforcement Learning comes to the fore which instead of concentrating on the near term goals takes into consideration the final requirement which is the overall conservation of energy in resource allocation. The results have been then plotted out to find the usage of the method in the conservation of energy during resource allocation.