Typical Interference Environment, Ofdm System Signal Detection Study

Because of robustness against multipath channels, high spectral efficiency and elimating inter-symbol interference by adding Cyclic Prefix, orthogonal frequency-division multiplexing (OFDM) is becoming an indispensable key technology for the future of mobile communications. However, the use of OFDM in an interfered senario poses several technical challenges, including interference suppression and channel estimation.Time-domain elimination method, adaptive notch filter and interference cancellation method based on threshold are three tipical technologies for interference suppression. Although these methods can eliminate interference to some extent, there are some problems with them. In time-domain elimination method, it is difficult to find out a well-performed parameter estimation algorithm. When the estimated paremeters are not accurate, it does not only fail to eliminate the interference but also introduce another new kind of interference, which may on the contrary affect the effectiveness of the method. In adaptive notch filter, the interference and transmitted signal are filtered simultaneously. Furthermore, this method works in an iterative fashion and the convergent process is usually quite time-consuming. The combination with encoding and interleaving are required for the third method to suppress interference, otherwise the effectiveness of the method is not obvious.For the above reasons, an interference suppression algorithm based on energy-spreading transform (EST) is proposed. In multi-tone interference senario, EST which spreads each symbol energy to different times and frequencies is operated for the modulated signal in transmitter. At receiving end, the Inverse Energy-Spreading Transform (IEST) which can increase the correlation between sub-carriers is performed for the signal after the time-domain elimination method. Finally demodulation is performed. The increase of the correlation between sub-carriers means that the residual interference after time-domain elimation method is decreased. To further improve the accuracy of signal detection, we send the reconstructed signal after modulation and EST to the time-domain elimation module. Through substracting the reconstructed signal from the received signal, the accuracy of parameter estimation can be improved, and the interference can be eliminated more effectively. Our dissertation analyzes the feasibility of the iteration fashion theoretically and verifies this conclusion by simulation further. In narrowband jamming (NBJ) senario, a frequency-domain detection algorithm based on moving average is proposed. We divide an OFDM symbol into several sets, each one contains several sub-carriers. The threshold factor is developed using the theory of constant false-alarm rate (CFAR) and order statistics (OS) after getting the mean square of the elements contained in each set. Compared with the mean square of single sub-carrier, the mean square of the elements contained in each set can more accurately describe the power of transmitted signal. Therefor, it can reduce the probility of false alarm. Meanwhile, the minimum value selected in our algorithm makes the threshold factor be more reliable because it is extremely insensitive to the NBJ. The plagued carriers are clipped using the proposed algorithm. Then the IEST and demodulation are performed for the processed signal successively. Analysis and simulation results show that our algorithm can effectively reduce the effect of NBJ on the systems.In order to maximize the performance of an OFDM system, channel estimation is usually required to obtain the information of wireless channel in communication receiver. Currently, both the pilot-based interpolation algorithms and the nonlinear estimation algorithms based on maximum likelihood (ML) or minimum mean square error (MMSE) criterion can perform well in low mobility applications. However, the performance of these algorithms deteriorates in interference environment.For this reason, the channel estimation in interference environment is discussed in our dissertation. Firstly, the typical estimation algorithms devised for conventional OFDM system in the present of interference are introduced, including joint detection and channel estimation algorithm and Expectation Maximization (EM) algorithm. Secondly, an estimation algorithm of path delays is proposed to overcome an unpractical assumption that the path delays are known in receiver in EM algorithm. The estimation algorithm includes NBJ detection and suppression and estimation of path delays in the present of interference. For the NBJ detection, according to the statistical property of channel frequency reponse on pilot carriers, the plagued pilot carriers are detected and suppressed by dual threshold Consecutive mean excision (CME). For path delays estimation, the mathematical model of channel frequency reponse on pilot carriers is analyzed. Using Alternating Notch Periodogram (ANP) algorithm, we can get path delays of channel. Simulation results show that normalized mean square error (NMSE) of channel by the improved EM algorithm are very close to the performance of EM algorithm which assumes that path delays are known in receiver, so is the corresponding BER performance of systems.