Research On M2M Communications Random Access Technologies Based On Cellular Systems

With the development of communications and internet technology, a large numbers of machine type devices will have the capability to access the network. The spontaneous data communication between machine device and machine device without the intervention of human is defined as Machine-to-Machine (M2M) communication, which is one of the most important parts of the Internet of Things (IoT) technology. M2M communications which studied in 3GPP is also called Machine Type Communications (MTC). The machine devices of MTC mainly use cellular network to transfer data. Since the existing cellular network infrastructure which is mainly designed for Human to Human (H2H) communication is not suit for M2M communications, the re-evolution of current cellular network is imperative needed. In this thesis, M2M communications random access technology based on cellular systems is investigated.Firstly, the urgent problem which the cellular M2M communications faced that the severe degradation of success access probability of M2M users and utilization of the resources of cellular systems when a great number of M2M users send random access request to the base station simultaneously is discussed. The random access network and core network of the cellular network, such as LTE-A, will be congested at the same time. The overload control schemes which listed in 3GPP such as Slotted Aloha, Dynamic Resource Allocation ,Specific Backoff and Slotted access and their advantages and disadvantages are discussed. A mathematical model is proposed to evaluate the performance of each scheme. Simulation results show that the Slotted access scheme is currently the best solution compared with the others.Subsequently, we proposed an M2M users random access scheme based on spatial reuse and discuss it respectively from grouping of users, allocating of resources to splitting of random access procedure. To divide the cell into local and global partitions, the optimal circle coverage algorithm which grouping according its geography location in the cell is proposed. We heuristically proposed the out-band and the in-band local resource allocation algorithm, which make the M2M devices maximized use the limit resources of the cellular systems. We split the procedure of random access into local completion sub-procedure and global access sub-procedure and creatively design the protocol of each sub-procedure. So the M2M user can launch the local completion sub-procedure that sending the preamble which is randomly selected from local resources with a low power, the winning user can continue to send the preamble which is selected from global resources and then complete the procedure of random access. The simulation results show that the proposed scheme outperforms the exiting random access schemes under the same parameters, especially in the performance of access delay and utilization of random access opportunitys(RAOs), the average access delay is reduced by approximately 60% and the utilization of RAOs is increased by 25% compared with the slotted access scheme.Finally, the energy efficiency of M2M users is further studied in this paper. On the one hand, K-means algorithm in machine learning is used to group users to reduce the radius of each group and the target of increasing the energy efficiency is achieved in the end. Simulation results reveal that this algorithm reduce the radius of each group effectively; On the other hand, the power optimal algorithm is proposed to reduce the user power of local completion and global access. The simulation results show that the energy efficiency of M2M users is increased by 100% after optimization algorithm is used.