Abstract: LS algorithm has been widely used in the OFDM channel estimation because of its simple operation. In actual systems, however, the algorithm is influenced by the noise of Inter-Carrier Interference（ICI） due to the presence of non-ideal factors. Meanwhile channel response will be interfered with outside noise in the burst transmission, which makes the estimated channel response amplitude fluctuate a certain range and have spikes and glitches. In order to solve this problem, an OFDM channel estimation based on the optimization of the FIR filter（Chebyshev filter） algorithms is proposed. Chebyshev filter attenuates quickly in the transition zone, and the error between it and the ideal filer frequency response curve is the smallest. So this algorithm can guarantee the performances of the local frequency are optimal compared to the traditional filer algorithm by simulation. The unexpected noise is introduced in the white Gaussian channel environment and a simulation of the proposed scheme is given. The simulation results show that this method can effectively eliminate the spikes and glitches in the passband caused by burst transmission.

Key words: Orthogonal Frequency Division Multiplexing（OFDM）, channel estimation, Least Square（LS） algorithm, Finite Impulse Response（FIR） filter, Mean Squared Error（MSE）

摘要： OFDM信道估计中，LS算法因其运算简单，得到了广泛应用。但是在实际的系统中，由于有非理想因素的存在，使得该算法受到载波间干扰（ICI）噪声的影响；并且突发传输时，信道响应会受到外界噪声的干扰，使估计的信道响应幅值在一定范围内上下波动，并带有尖峰和毛刺。为了解决此问题，提出了一种基于LS算法的最优化FIR滤波器（切比雪夫滤波器）信道估计算法。切比雪夫滤波器在过渡带衰减很快，和理想滤波器的频率响应曲线之间的误差最小，因此该算法较之传统的加窗算法，能保证局部频率点的性能也是最优的，进而有效地减小均方误差（MSE）。在高斯白信道环境下引入突发噪声，对所提方案进行了仿真，其结果验证了该方法能有效消除通带内因突发传输引起的毛峰和尖刺。

关键词: 正交频分复用（OFDM）, 信道估计, 最小平方（LS）算法, 有限长单位冲激响应（FIR）滤波器, 均方误差（MSE）