Research On Low Power Consumption Coverage Enhancement And Resource Scheduling Of M2M Communication Of 5G System

Machine-to-Machine(M2M)communications is a special type of communication between devices without any human intervention.As the key technology of Internet of things(IoT),M2M communication can be applied into many areas,such as in transportation,finance,smart home,environmental monitoring and smart grids.With characteristics including high speed transmission,large coverage,high reliability,easy to deploy,mobile cellular networks would be ideal carriers of IoT services.However,existing cellular networks are optimized and designed based on Human to Human(H2H)communication,the unique characteristics of M2M communication will challenge the cellular network.Such as low power wide coverage(LPWA)service,need to provide improved coverage for massive connections of low-throughput low-cost IoT machine type communication(MTC)devices with low device power consumption in delay-tolerant applications.However,frequent connections required by massive numbers of MTC devices cannot be satisfied because of the limited capacity of cellular networks.Therefore,the solution of massive MTC devices to access problem is one of the key problems of 5G(the 5th Generation mobile communication technology)systems.In this paper,a scheme on coverage enhancement of narrow band M2M communication based unlicensed spectrum is proposed to solve the access problem of IoT LWPA services in 5G.Meanwhile,reinforcement learning based M2M distributed scheduling algorithm is proposed to solve the problem of M2M resource scheduling in cellular network.The main contents and innovations are as follows:Firstly a coverage enhancement of narrow-band M2M communication system is proposed to satisfy the characteristics of LPWA and the scarcity of licensed spectrums.In this paper,the physical layer design scheme of the system is introduced in detail.Meanwhile,in the study of M2M communication coverage enhancement,the scheme of repetitions and low-level modulation and coding in transmitter is proposed,and corresponding receiving mechanisms are proposed to achieve low power consumption and wide coverage.Simulation results prove that a 10?21 dB gain of the coverage performance can be achieved compared with the current LTE systems.Secondly,to enable efficient resource scheduling for massive MTCD in 5G systems,a distributed resource management scheme based on reinforcement learning is proposed.M2M devices are firstly classified according to the characteristics of data traffic and delay for the problem of resource scheduling in 5G cellular networks.According to the M2M service type,the terminal equipment of the same type is divided into a cluster,and further is divided into multiple access groups based on the location information.Then,the group head device applies for scheduling resources on behalf of all members of the group.On the basis of grouping,a distributed M2M scheduling algorithm based on reinforcement learning(RL)is proposed.The original scheduling problem is modeled as a multi-agent RL system.Based on collecting environment information,the leader device capable of the RL ability tries to find the best wireless scheduling resource by trials and errors to complete the data transmission.Simulation results demonstrate the fesibility,fairness and superiority of the RL based algorithm comparing with the other current mechaisms.