论文部分内容阅读
为了有效解决稀疏度未知的压缩信号快速重构问题,提出了一种适应范围较广、效率突出的信号重构方案.在该方案中,压缩信号的上下界按照等距性质获得,并将最接近其中值的整数作为信号稀疏度的估计值;通过减少迭代时观测向量在支撑集上的投影次数,降低该方案的运算复杂度;给出了能够反映整个信号重构概率的评估体系,并以此体系验证分析该方案的有效性.通过实验表明,该方案有效实现了稀疏度未知信号地快速重构,同时其重构的成功概率也高于现有的回溯类重构方案.“,”The paper presents a signal reconstruction program, which has wider adaptation scope and highlighting efficiency in order to effectively solve the problem of sparsity unknown compressed signal rapid reconfiguration. In this scheme, bounds of compressed signals are obtained in accordance with the equidistant nature. It uses the nearest integer value as the estimated value of signal sparsity. Through reducing the projection number of the observation vector on the foothold set when iterating, the operational complexity of the program is reduced; Gives an assessment system that could reflect the entire signal reconstruction probability, and uses this system to verify the effectiveness of the program analysis. Experiments show that the program effectively achieves quickly sparsity unknown signals reconstruction, at the same time, the success probability of its reconstruction is higher than the existing backtracking class-remodeling program.