Study On Dynamic Resource Management In Time Division Duplex Heterogeneous Networks

With the rapid development of mobile Internet and the popularization of smart terminals, mobile services experience explosive growth. To satisfy the ever-increasing demand of people for high-data-rate services, mobile communication systems are transforming from the conventional cellular networks into the heterogeneous networks which include various types of densely deployed small cell base stations and support the device-to-device(D2D) technology. On the one hand, by densely deploying small cell base stations, the coverage in blind areas and hotspots can be effectively improved. On the other hand, as an alternative option of the 5th generation mobile communication systems(5G), the D2 D technology enables mobile users in proximity to establish a direct link by reusing cellular resources and bypass the base stations, thereby offloading the network infrastructure and providing increased spectral efficiency. The emergence of various new types of mobile applications also makes the traffic in different cells have bursty and time-varying characteristics, leading to traffic asymmetry in networks. Time division duplex(TDD) can accommodate uplink/downlink(UL/DL) traffic asymmetry by flexibly varying the UL/DL subframe configurations, so as to improve the utilization of network resources. To better accommodate the dynamic characteristics of mobile services, the dynamic TDD allows the base stations to dynamically adjust cell-specific UL/DL subframe configurations according to the instantaneous traffic conditions of the serving cell. Therefore, the TDD heterogeneous networks are regarded as one promising approach for future wireless networks.Many types of interference exist in the TDD heterogeneous networks, including the inter-tier interference between macro cells and small cells, intra-tier interference among macro cells or small cells, intra-tier interference between D2 D terminals and macro cells or between D2 D terminals and small cells, and intra-tier interference among D2 D terminals reusing the same cellular resources. Meanwhile, the introduction of dynamic TDD to the heterogeneous networks causes cross subframe interference, including the interference caused by the DL base station to the UL base station and that caused by the UL terminal to the DL terminal, which aggravates the network interference. The complicated types of interference have become the major factor limiting the improvement of network performance. Therefore, it is extremely important to design effective dynamic resource management and control strategies for TDD heterogeneous networks to mitigate the interference and further enhance the system performance. The main achievements and results of this dissertation are summarized as follows:1. We have proposed an optimal resource allocation strategy based on dynamic TDD to accommodate the network traffic distribution in a large-scale TDD small cell network. The strategy can minimize the cell service time and maximize the network energy efficiency. Chapter 2 of this dissertation first classified small cells according to the service requirements in the cells and studied the impact of different traffic loads on the success transmission probability of a random cellular link. Then, given the distributions of service requirements of small cells and mobile users, with the objective of minimizing the service time of each type of small cells, we have derived the UL/DL subframe configuration factor that can optimally accommodate the UL/DL traffic conditions in each type of cells. Finally, by using the dynamic TDD technology, we have configured the ratio of UL/DL subframes and realized the optimal match between network services and network resources. Compared with the conventional semi-static TDD scheme, we find that the dynamic TDD scheme can significantly reduce the cell service time when there is a great difference in the service requirements among different types of cells. Furthermore, we have investigated the relationship between resource allocation and network energy efficiency. The results show that the proposed UL/DL subframe configuration scheme not only minimizes the cell service time but also achieves optimal network energy efficiency.2. We have proposed an interference modeling scheme which is applicable to the hybrid networks including both of dynamic TDD heterogeneous cellular networks and D2 D networks, and designed a channel access scheme for D2 D terminals. The channel access scheme can effectively mitigate the mutual interference between D2 D terminals and cellular terminals, and improve the success transmission probability of the link and network throughput. With the tools from stochastic geometry, Chapter 3 of this dissertation first characterized the network interference and derived the success transmission probabilities of a random cellular link and a D2 D link, as well as the average throughput of the network. Then, we studied the impact of important network performance parameters, such as UL/DL subframe configuration factor, base station density and bias factor, on the UL and DL success transmission probabilities of a random cellular link, and the impact of access control scheme for D2 D terminals on the network throughput. Meanwhile, we investigated the tradeoff between the success transmission probability of the link and the activity of D2 D terminals, and derived D2 D terminal channel access parameters that can maximize the network throughput.3. We have designed a distributed D2 D cooperative transmission scheme in the TDD cellular network to enhance the network throughput and energy efficiency. Chapter 4 of this dissertation considered the scenario that, in a single-cell TDD cellular network, the D2 D terminals are allowed to reuse the uplink cellular channel resources. D2 D terminals access channel resources in their concurrent transmission with the cellular user, and help the cellular user retransmit cellular packets to compensate for the loss of cellular throughput during the concurrent transmission. In the proposed D2 D cooperative transmission scheme, we have fully exploited the gain achieved by the cooperation between D2 D terminal and cellular user, and among D2 D terminals. As such, we have significantly enhanced the spectral efficiency and energy efficiency of the hybrid network on premise that the cellular user’s Qo S requirement is satisfied. With the Rayleigh fading channel model, we have derived the closed-form expressions of normalized throughput for cellular users and D2 D terminals, and analyzed the network energy efficiency. Simulation results have demonstrated the superiority of the proposed D2 D cooperative scheme in improving the network throughput and energy efficiency.