The main theme of this dissertation is the design of adaptive medium access protocol to alleviate the traffic overload and to provide resource efficiency for the massive number of devices. This dissertation focuses on the development of mechanisms to cope with the traffic overload due to the time-variant arrival of devices, and the mismatch between the estimated number of devices and the actual number of devices in cellular networks. This dissertation also focuses on efficient random access (RA) procedure for Internet of things (IoT) devices.
The dynamic allocation of RACH resources (DARR) is one of solutions to alleviate the traffic overload from massive devices when the resource for RA can be increased. This dissertation discusses the challenge of a gap between the theoretical maximum throughput and the actual throughput in DARR. The gap occurs when the BS cannot change the number of preambles for a RACH until multiple numbers of RACHs are completed. In addition, a preamble partition (PP) approach is proposed that uses two groups of preambles to reduce this gap. The simulation results show that the proposed approach can achieve the throughput which is closer to theoretical maximum throughput than other approaches.
The resources for the machine-to-machine (M2M) communication devices can be limited and shared with other type of devices. In this case, both the DARR and the access class barring (ACB) are required to provide resource efficiency and to alleviate the traffic overload. This dissertation discusses the gap between the theoretical maximum throughput and the actual throughput since the BS cannot change the number of preambles for a RACH until multiple number of RACHs are completed. Based on the discussion, a preamble partition and a stochastic gradient descent approach are proposed to reduce the gap when both of DARR and ACB are used for mobile network system. The simulation results show that the proposed protocol shows the throughput which is close to the throughput with ideal selection.
The data transmission using a RA procedure for human-to-human (H2H) communications requires the overhead for signaling messages. The overhead for the signaling is very high compared with the size of data in small data transmission (SDT). Thus, the recent release of LTE-A standard includes two SDT procedures that reduce the number of exchanges. This dissertation evaluates the performance of conventional SDT procedures in the viewpoint of resource usage and resource throughput. In addition, this dissertation also proposes an SDT procedure to reduce the resource usage and to increase the resource efficiency. The numerical evaluation shows that the proposed approach decreases the resource usage and increases the resource efficiency.