Channel Simulation And The Channel State Information Is Estimated That Feedback Technical Studies

The radio wave which can propagate without any media is the carrier of signals for wireless communications. Because of the outside environments and the moving of mobile stations, the properties of the radio wave can be changed during the propagations. The whole environments which change the properties of the radio wave are abstracted and named the wireless channels in the wireless communication researches. Wireless communications are affected by the wireless channels greatly, we must have a deep understanding of the wireless channels in order to design a good wireless communication system.This dissertation focuses on the topics about the wireless channels. As the computer-aided designs are widely used today, the wireless channels for simulation are required in order to have compute-aided verifications in the design phase. In chapter 3, after elaborating the channel models and implementation methods, a design for the MIMO channel based on the COST207 standard on the Synopsys'COSSAP? platform is put forward. The bigaussian power spectrum is implemented in the constructed channel, and so does a method which produces the specific correlation coefficients matrix from the physical parameters. The constructed channel well simulates the real environments in a series of tests.Because of the importance of the wireless channels for the wireless communications, the estimation of channel state information is needed in real systems. In chapter 4, a channel fading coefficients estimation algorithm is put forward for SISO-OFDM and MIMO-OFDM systems by using the EM algorithm framework. This algorithm approaches to the ML estimation by iterating. From the simulation results, the algorithm has a lower BER and system overhead than traditional interpolation method, while it is close to the performance boundary.Besides the researches on the estimation of channel fading coefficients, in chapter 5, the dissertation also proposes an algorithm for estimating the channel noise variance in MIMO-OFDM systems and a feedback scheme for channel fading coefficients. The algorithm for estimating the channel noise variance needs only two OFDM symbols transmitted from each transmit antenna, while without any channel coefficients information. The feedback scheme for channel fading coefficients uses the discrete cosine transform to compress the information. From the simulation results, the noise variance estimation algorithm has good estimation accuracy, while keeping the system overhead low. The proposed feedback scheme reduces the feedback data rate while keeping the integrity of the information.