ID 原文 译文
53307 贝叶斯压缩感知理论将稀疏贝叶斯学习理论引入到压缩感知中，给需要重构向量中的每个值设置受超参数控制的后验概率密度函数， BCS method introduces the sparse bayesian learning theory into CS. It sets posterior probability density function that is controlled by hyperparameters to each value in the vector that demand reconstruction.
53308 在超参数的更新过程中，零值所对应的超参数将趋向于无穷大，与之对应的后验概率将趋向于零， In the process of updating, many of the hyperparameters tend to infinity for those values that have insignificant amplitudes, and these corresponding posterior probability tends to zero.
53309 通过这种方法剔除非重要多径，自适应地找出信道向量中的重要多径，并使用回归算法进行重构。 This method can reject trivial multipaths, find the critical paths in channel vector automatically and reconstruct them with regression algorithm.
53310 实验结果表明在信道稀疏度未知的情况下，该方法能够对原信道进行有效的重构。 The experiment results show that this method can reconstruct original channel effectively under the circumstance that channel sparsity is unknown.
53311 采用更高阶 Volterra 滤波器更好地逼近非线性系统时，Volterra 自适应滤波算法的计算复杂度呈幂级数增加。 As the orders of Volterra filter are enlarged to approach the nonlinear system better, the computational complexity of Volterra adaptive filtering algorithms increases by power series.
53312 针对此问题，本文提出了一种在 稳定分布噪声背景下基于离散余弦变换（DCT）的三阶 Volterra 滤波算法。 An adaptive algorithm of third-order Volterra filter basedon DCT in -stable noise environment is proposed.
53313 首先将Volterra 滤波器的三次项权系数矩阵分解成一组二次项权系数矩阵; First, the cubic term weighting coefficients matrixes are decomposed into a group of quadratic term weighting coefficients matrixes.
53314 然后利用正交变换，将二次项权系数矩阵变换成对角矩阵，从而大大减少了权系数个数，有效降低了算法的计算复杂度; Second, using orthogonal transformation, the quadratic term weighting coefficients matrixes are transformed into diagonal matrixes, so that not only the number of weighting coefficients is decreased greatly but also the computational complexity is effectively reduced.
53315 最后将 Volterra 自适应滤波器输出表示成线性滤波器输出形式，并由此得到权系数自适应调整算法。 Lastly, the output of Volterra adaptive filter is expressed as the output of linear filter and the adaptive adjusting algorithm of weighting coefficients is proposed.
53316 系统辨识的仿真结果表明，本算法在 稳定分布噪声背景下具有优越的性能。 Simulation results of system identification show that the new algorithm has better performance in -stable noise environment.