Research On Channel Estimation Algorithms For OFDM System

Along with development and improvement of society, many researchers pay much attention to developing system of wireless broadband digital communication. In wireless broadband digital communication, Inter-Symbol Interference (ISI) caused by time disperse of radio channel is one of main factors that limit bandwidth of communicated information and system performance of communication. In receiver, Orthogonal Frequency Division Multiplexing (OFDM) system can easily cancel ISI and equalize frequency-selective failings that are both caused by time disperse of radio channel. For this, it has become a prime candidate for B3G communication. The requirements for B3G systems such as high data rate, high quality and high mobility make channel estimation more strict.Channel estimation has attracted widespread attention, and many related algorithms are presented these years, among which pilot symbol aided modulation pilot symbols are placed in separate subcarriers .Which result in a significant bandwidth loss. Hence we use pilot symbols which can be added to data symbols. This paper presents a new iterative channel estimation based superimposed pilot. A joint time domain maximum-likelihood (ML) and correcting of channel impulse response channel estimation scheme was developed for more precise channel estimates. Theoretical analyses and simulation results show that the proposed method has better performance than iterative maximum-likelihood (ML) algorithm based no adding compute complexity.Semi-blind channel estimation save bandwidth by avoiding the use of training symbol and make it more strict. This paper proposes a new subspace-based adaptive semi-blind channel estimation for zero padding OFDM systems. Employing the pilot information of ZP-OFDM systems, this algorithm gives a accurate initial channel parameters. Then channel estimation is computed adaptively using iterative recursion and power method. Not only reduces this algorithm computation complexity, but improves the blind channel estimation in convergence rate. Simulation results also show that the algorithm has better performance.