Application Research On MUD In CDMA Communication System

Code Division Multiple Access(CDMA) communication system is an interference-limited system. The main factors of affecting the capacity and performance are multi access interference(MAI) and near-far problems. With the increasing number of users, MAI is the main interference among others. In order to solve this problem, the traditional receiver is used in CDMA system. It consists of a group of matched filters and correlated decisions. The principle is to separate the desired information and other user’s information. The desired information is regarded as useful signals and other user’s information as interference. This method still exists MAI problem. Compared with the traditional receiver, an effective technique is proposed called multiuser detection(MUD) techniques and it has been widely used in CDMA system. The principle is to use all signals as useful information. This technique can effectively overcome MAI by analyzing several parameters such as user’s spreading code, useful signal amplitude, timing and the delay characteristics. It also increases communication capacity and improves performance in CDMA system.The study on this thesis is to research several typical methods based on multiuser detection(MUD) techniques. The theory of MUD has undergone a process of continuous development and improvement. This thesis mainly focuses on the following points:(1) It is outline some basic theory of CDMA communication system such as basic principle, key technologies, and its problems, which thus leads to the concept of MUD. Besides, it is naturally elaborated on the basic principle, basic structure and performance measurement.(2) Based on the study of the principle of MUD and more classical MUD techniques, the principle and performance of conventional Detector(CD), decorrelating detector(DD), MMSE detector, successive interference cancellation(SIC) detector and parallel interference cancellation(PIC) detector are explained and analysis in detail as well as the advantages and disadvantages of different methods. In order to compare different methods, it is simulated a synchronous CDMA communication system with additive white Gaussian noise(AWGN). Simulation results show that MMSE detector has a good anti-proximity effect against near-far problem and better performance of signal to noise ratio(SNR), while SIC and PIC has low implementation complexity.(3) The principle of two blind multiuser detector of least mean square(LMS) and recursive least squares(RLS) are descripted firstly. Under the simulation communication system with AWGN channel, the performance measurements of SNR, mean error and anti-interference are compared in simulation system. The experimental result indicates that these two algorithms can achieve the optimal performance state of the communication system. LMS and RLS both can have a recovery state automatically in the case of the mutation status.