Design And Implementation Of FPGA-based Linear Interference Alignment Communication System

Interference widely exists in wireless communication system. As the number of users increasing, the multi-user interference severely limits the capacity of the system. Interference has become one of the major problems of future wireless communication. The traditional methods manage interference, the spectrum efficiency of a single user drops rapidly faced with current dense multi-user system. The traditional interference management methods applied to current dense multi-user system bring about a low single-user spectral efficiency. Interference Alignment(IA) is new technology, which is from the Angle of the overlapping Interference to deal with the multi-user Interference. By precoding technology, IA makes the interference overlapped in the null space of desired signal at the receiving end, eliminating interference on the influence of the desired signal, so that system capacity was doubled. Since the interference alignment technique requires to know the global channel information, so we must find a method to reduce the amount of channel feedback information in engineering implementation.Relying on the National Natural Science Fund Project “Interference management and capacity analysis in wireless networks”, this thesis from the perspective of project implementation, proposes interference alignment link control protocol and a method of reducing the amount of channel feedback information and explores how to design and implement of linear interference alignment OFDM communication system under the background of the TD-LTE based on Field Programmable Gate Array(FPGA).Firstly, by balancing the channel estimation accuracy and system transmission performance, the thesis proposes interference alignment link control protocol, and in which a method of reducing the amount of channel feedback information is used. According to this protocol, a mobile station estimates channel at the beginning of each wireless frames on a frequency domain, and then transform frequency-domain channel estimation to time-domain to reduce the amount of feedback channel information. Base station broadcasts the received global channel estimation information, while using the channel estimation information obtains interference alignment precoding matrix, and the subsequent valid data of the current frame is transmitted by pre-coding processing. The mobile station utilizes the global channel estimation information from base station broadcasting to calculate decoding matrix. The valid data portion of current frame is decoded with the decoding matrix by the mobile station. As a result, Interference alignment is implemented.Secondly, after verifying the theory, this thesis implements the transceiver module design and coding work of OFDM baseband system, which contains linear interference alignment technology, based on FPGA technology. The communication system platform uses the top-down design method. Firstly, design the architecture map of whole system, and then use the Xilinx company development suite ISE 14.6 and simulation software Model Sim to complete the design and coding of each module through the Verilog hardware description language. This thesis mainly describes the design thinking and implementation method of scrambler and descrambler module, mapping and demapping module, the reference sequence distribution module, channel estimation and feedback module, FFT/IFFT module and peripheral interface module. Finally this thesis evaluates the performance of the platform and discusses the different performance between using and not using IA technology.Simulation and operation results show that the OFDM baseband communication system design includes IA technology meet basically the requirements of the codec module. Simulation and hardware test shows that the modules work properly and three pair users can communicate at the same time,which verifies the proposed IA link interaction protocol and the method of reducing the amount of channel feedback information. The results of analysis and evaluation of the platform performance show that IA technology platform throughput of approximately 2 times the traditional platform using multiplexing.