Study On Multi-RAT Concurrent Transmission In Wireless Heterogeneous Networks

With the explosive development in communication, more and more heterogeneous networks with different characteristics and complementary advantages will emerge and coexist for ages. However, after decades of sustained growth, the transmission rate of single Radio Access Technology (RAT) is gradually approaching the Shannon capacity, while the urgent demands for Data-hungry Applications are still augmenting progressively and rapidly. In this light, the network cooperation technology has been devised to satisfy this trend. However, these traditionally loose coupling mechanisms are failed to guarantee the QoS as well as to handle the high power consumption and the severe interference problems. The concurrent transmission, regarded as a tight-coupling technology for heterogeneous networks, can transmit the traffic via multiple RATs simultaneously and has become one of the key techniques to handle this issue. This dissertation mainly focuses on how to address the above problems via utilizing concurrent transmission technology in heterogeneous networks scenario. The delivery of Data-hungry Applications is firstly achieved by the proposed network structure and data splitting strategies. Then, the potential of concurrent transmission to promote network energy efficiency (EE) is investigated. Finally, the mentioned strategies are extended to address the interference problems in heterogeneous networks. The main achievements and results are listed as follows:1. This dissertation classifies concurrent transmission network structure, establishes the optimal model of concurrent transmission and derives the theoretical delay bound. The fundamental problem of concurrent transmission is how to couple various kinds of existing network systems together. To address this problem, based on the comprehensive researches on characteristics of different networks, this dissertation categorizes the concurrent transmission network structure into partially concurrent transmission network structure and entirely concurrent transmission network structure. Furthermore, we analyze the application scenarios, traffic transmission models and the effect on data splitting strategies under different network structures. On this basis, the equivalent queuing model is adopted to analyze the end-to-end delay performance of perfect splitting strategy and the theoretical delay bound is obtained, which provides reliable theoretical basis for our subsequent researches.2. We propose different data splitting strategies for the proposed network structures, analyze their performance, and validate the proposed strategies with DEMO. The quality of data splitting strategy determines the performance of concurrent transmission directly. In this dissertation, we focus our attention on how to obtain the optimal delay performance of concurrent transmission in heterogeneous networks. Furthermore, based on the different network structures, three data splitting strategies to jointly schedule the multi-RAT resources, namely, Minimum Queuing Delay based splitting strategy (MQD), Minimum Queuing Length based splitting strategy (MQL) and Unified Queuing Management Based Splitting Strategy (UQM), are presented. Two-dimensional discrete-state continuous-time Markov process is exploited to analyze these strategies. DEMO results, implemented by WLAN and WiMAX, demonstrate that our proposed data splitting strategies utilize the system resources efficiently, outperform the existing strategies, and approach to the theoretical delay bound especially in the heavy load region. More importantly, MQD reduces the delay jitter remarkably.3. This dissertation establishes the EE concurrent transmission optimization model, proposes the optimal EE splitting theorem and devises the EE maximization concurrent transmission strategy. Nowadays, the extremely high power consumption caused by the data hungry application has been become one of the main shackles on wireless network development. Aiming at this problem, the benefits of concurrent transmission from the perspective of green communications are studied in this dissertation. The basic principle is elaborated by exploring the theoretical relationship between the transmission rate and EE. Furthermore, the EE concurrent transmission optimization model is established and the traditional EE-Bandwidth tradeoff is modified. In addition, the nature of multi-RAT unified scheduling diversity gain is discovered to improve the system EE. Besides, the traffic splitting threshold and the optimal rate allocation of each flow in the downlink transmission are derived to maximize the system EE according to the relationship of available channel state information of different networks, energy consumption and data rate. And then, the optimal RATs selection and transmission rates for establishing concurrent transmission among multiple RATs in the uplink transmission are derived under some given static circuit power conditions. Furthermore, we formulate this problem as a mixed binary integer programming by considering the influence of static circuit power. Based on the relaxation of integer constraints, a low complexity heuristic algorithm for general static circuit power conditions is presented, which can achieve the near-optimal solution. Simulation results validate the effectiveness of the proposed strategy and demonstrate that the EE performance of the multi-mode user equipment can be significantly improved.4. We propose an interference migration strategy via concurrent transmission which can utilize the interference distribution diversity gain efficiently. And we analyze the optimal threshold and quantity of this strategy. As the home base stations are widely employed to enhance indoor coverage and network capacity, the interference becomes one of the main challenges to deteriorate the system performance. This dissertation first explores the non-uniform distribution property of interference, which is caused by the random deployment of femtocells and the unbalanced load distribution. Then, an interference migration strategy with concurrent transmission, which connects the originally separated regions together, is proposed to transfer the interference among different interference regions. Meanwhile, an interference intensity index is defined to depict the non-uniform interference distribution. Then we derive the threshold for executing interference migration and the optimal transmission splitting probabilities for EE maximization. The results demonstrate that the proposed strategy significantly promotes the EE.