Research On Channel Estimation Algorithms Of The OFDM System

As a high-speed data transmission technique, orthogonal frequency division multiplexing (OFDM) has been widely used in wireless communication systems due to its capability to combat multi-path fading. In the wireless communication systems, it's necessary for a dynamical estimation of channel since the properties of time-varing channel.In the past, When the researchers studied on the channel estimation,they often used a fixed channel model(multi-path fading and doppler frequency shift)to estimate the channel impulse response with some classic algorithms. To some extent, this method could not only improved the estimated performance on the particular circumstances, and the scope of the application of algorithms were greatly restricted,so lead to the algorithm used in wireless communications transmission system was greatly restricted, then the paper does a more study in-depth for this issue,and based on this, to study the wireless multi-path variable channel statistical characteristics. At the same time,it is establishing a new channel mathematical model of uncorrelated scatterring hypothesis to make it more suitable for a wireless communication system.Based on the above mathematical model,we investigated the estimation algorithms from the point of views of frequency domain and time domain in order to improve the scopes and system performance by improved algorithms. In most OFDM systems, it usually does these assumptions that the length of cyclic prefix is longer than the length of the channel impulse response. However, if the length of channel impulse response is longer than cyclic prefix, the traditional OFDM frequency domain algorithms will not effectively estimate channel. Based on it,in this dissertation, we propose one channel estimation method which are based on time domain pilots. The practical reasoning and simulation results indicated that the proposed algorithms can obtain the approximate estiamtion performance as in the fixed-channel own, and it not only reduce the number of estimation dimention,but it also simplify the complexity of system and effectively eliminated the excessive interference that produced as a result of much longer length of channel impulse response. In time domain,pilots are transformed by DFT at first,then combining frequency domain MMSE with time domain LS to estimate channel impulse response and does the simulation with the above same condition. The simulation results showed that under the time-varying circumstances, when the number of pilots is bigger than the channel length scaled by system,time domain estimation is better than frequency domain estimation, when the number of pilots is smaller than the channel length scaled, MMSE estimation which has the second statistic properties of channel is better than time domain LS estimation.Based on the estimation algorithms of channel mathematical model of uncorrelated scatterring hypothesis which the paper has built, the practical simulation results showed that it has greatly enhanced the system's ability against to bad channel environments and improved the ability to influence of combating the multi-path fading,etc.in a word,the algorithms has made great improvement to the whole system performance.