Anti-Interference Wireless Multi-User Communication System Based On Autocorrelation Function Coding

Due to the increment of wireless applications, wireless communication channels are crowed by many users. For multiple user communication system, inter-user interference and co-channel interference becomes one of the main factor that impact communication quality. Traditional anti-interference methods rely on the training code or pilot signals. They need to transmit extra signals in the communication process and therefore reduces time for the transmission of user signals. Most anti-interference methods base on the reliable synchronization of time and/or frequency, which is not applicable to the environment where interference power is too high to obtain precise synchronization. As such it is required to find new interference blocking and multiple user isolation method which is able to isolate multiple users and blocking high power interference. The new method is also required to be insensitive to both the time and the frequency synchronization. In this thesis, an autocorrelation function Coding based anti-interference multi-user system is studied.In this system, autocorrelation function of each user is design to be linear independent and their linear space disjoins to that of the autocorrelation function of interferences. By signal processing on the covariance matrices of the received signal, the spatial filter that block interferences and isolate uses can be obtained. In the thesis the user isolation and interference blocking algorithm for flat-fading channel are studies first. The analysis of the covariance matrices of received signals shows that they consist of components from both the interferences and tha user signals. We have proposed a linear operation algorithm on the covariance matrices to extract the components of the interferences and at the same time canceled the components related to user signals. From the extracted components we have shown that the spatial filter that block interference and isolated users can be obtained by matrix decomposition. Theoretical analysis and simulation shows that the proposed method is insensitive to the power of interferences and Gaussian white noises. When the solution of the required spatial filter is not unique, an optimal solution is proposed which can maximize the SNR of the output. Furthermore, we have proposed the Autocorrelation Function Zero-forcing algorithm, which can simplify the procedure of extracting interference information form the covariance matrices of the received signal. One important feature of the autocorrelation function zero forcing algorithm is that it is insensitive to the carrier frequency error. This property is useful for systems that have strict requirement on frequency synchronization, and it can be applied to communication systems on high speed vehicles.In the second part of the paper, the user isolation and interference blocking method on frequency fading channels are studied. We have proposed a modified Autocorrelation Function Zero-forcing algorithm for the frequency fading system. We have also studied the application of the algorithm to the multi-user OFDM system. According to the specific properties of OFDM systems, the design criterions for the autocorrelation of multi-user OFDM system are proposed. Since OFDM is one of the widely used modulation method and it is sensitive to interference and synchronization, the proposed method is important for the implementation of multi-user OFDM technique under interference environment.In the end of the thesis, we have discussed several extensions of the proposed methods. Such as the reduction of receiving antennas and the incorporation of autocorrelation function Coding to the MMSE receiver to maximization output signal to interference-plus-noise ratio and reduces bit error rate.