Research On Dynamic Resource Allocation In Broadband Powerline OFDM Communication System

Broadband powerline communication (BPLC) which uses distribution power networks as transmission media is a kind of high speed communication technologies. Its major applied areas includes:power load management, AMR, distribution automation, Internet broadband access, home networking and etc. With the construction and development of intelligent grid project in china, BPLC has become more and more attractive in recent years.Based on the detailed analysis on the research and development of BPLC, the study of this dissertation is focus on the resource allocation problem of powerline OFDM system. The main achievements of this thesis are as follows:The methods and research statues of resource allocation technology are summarized and reviewed. The main problems and the way of investigation are put forward.According to the investigation on the characteristics of the powerline channel and the theory of OFDM system, the model of the broadband powerline OFDM system is firstly established. And then, the basic principle of the water-filling is analyzed to provide the theoretical back up for the research of resource allocation problems.Through analyzing the principle of multi-user powerline OFDM system, the optimization model of resource allocation problem based on proportional fairness is established in this paper. In order to solve this optimization problem, a novel method based on quantum genetic algorithm is proposed. Comparisons between the proposed algorithm and several classical optimization algorithms are made under the typical powerline channel environment. And the simulation results show that the total throughput and proportional fairness can be improved by the proposed algorithm.In order to resolve the problem of resource allocation for mixed traffics under the power spectrum density constraints, two kinds of suboptimal allocation algorithms are proposed in this paper. One is based on the idea of cycle selection while the other is on the basis of iterative exchange. By means of simulation, the performances of two algorithms in terms of throughput and serveice fairness index (SFI) are compared under the typical powerline channel. Results show that both of the proposed algorithms are able to improve about 10% of throughput to TDMA without reduce SFI obviously. Moreover, the algorithm based on the idea of cycle selection offers a greater throughput while the other algorithm provides better SFI.Considering the situation of dynamical traffics, the idea of cross-layer design is introduced to the resource allocation problem of powerline OFDM system. And the theoretical framework including the length and structure of frame, data source model is established. On basis of the framework, a joint MAC-PHY layer resource allocation algorithm for powerline network is proposed. This algorithm aims to reduce the packet loss rate as well as fulfill the QoS of users. It also improves the throughput by means of physical layer optimization allocation. Analytical and simulation results show that the proposed algorithm is able to guarantee the QoS of users. Meanwhile, the power efficiency and system performance are significantly improved compared to the signal layer algorithms.