Change When The Ofdm Channel Estimation

Orthogonal Frequency Division Multiplexing (OFDM) is a multi-carrier digital modulation technology, which can solve the multi-path fading problem effectively. OFDM has a very high utilization rate of frequency band, which is applicable to wideband mobile communications. OFDM is the most promising technology in high-speed mobile communications of next-generation mobile communication systems. But the channels in OFDM systems are time-varying and frequency selective. When the NDS (Normalized Doppler Spread) is larger than 0.01, the constant channel model will not hold, so a linear or nonlinear model is required for the channel. To get the channel state information timely and realize real-time and accurate coherent demodulation, the channel must be estimated dynamicly.This paper mainly study the time-varying channel estimation in OFDM systems. Research results are summarized as follows:1. We propose a new channel estimation method based on ML(Maximum Likelihood) rule and SOPI(Second-order Polynomial Interpolation) for medium time-varying wireless channel, which uses time-domain pilot sequence. This method is implemented as follows: First, we get the CIR(Channel Impulse Response) estimation corresponding to the pilot sequence, then we use the estimated CIR in the first step to interpolate the CIR of the midpoint of an OFDM symbol by second-order polynomial interpolation (SOPI) method. We use the CIR of the midpoint of an OFDM symbol as the CIR of the OFDM symbol. Simulation results show that the new method of channel estimation has excellent performance and lower complexity, which has broad space for development.2. We study the time-varying channel estimation method in OFDM systems based on particle filtering. The main idea of particle filtering is using some discrete random sampling points to simulate the random variable probability density function of the system, replacing the sample mean by integral operator to obtain state minimum variance estimates. Actually particles are some random sampling points in state-space. Particle filtering has a good state estimation performance for non-linear system. In this article we first discuss the principle of particle filtering and perfom simulation on the state estimation performance of particle filtering, and then apply this method to time-varying channel estimation in OFDM systems. A first order AR model is adopted to model the time-varying channel first, and then we apply the particle filtering to estimate the CIR. Simulation results show that the method based on particle filtering has a good channel tracking performance, which is a worthwhile channel estimation method to continue to explore and develop.