中文摘要：

针对测量值的一比特量化，提出了一种新型贪婪迭代算法：符号子空间追踪算法。该算法融合了一致性恢复和贪婪迭代的原理，将量化误差对重构的影响降到最小。仿真结果表明，在高比特率的情况下，该算法的重构误差比只考虑稀疏性和只考虑一致性恢复的算法分别低13dB和21dB。

英文摘要：

A novel greedy iterative algorithm is proposed for the measurements quantized to one bit： sign sub- space pursuit algorithm. This algorithm combines the principle of consistent reconstruction and greedy iteration, and minimizes the influence of the quantization error on the reconstruction. In the case of high bit rate, simulated result shows that the reconstruction error of the algorithm is about 13 dB and 21 dB lower than traditional recovery algo- rithms that use only sparsity or only consistency.