针对在实际宽带压缩频谱感知中难以预先获知宽带频谱稀疏度的问题，提出一种改进的稀疏度自适应匹配追踪(modified sparsity adaptive matching pursuit, MSAMP)算法，该算法在支撑集选择过程中对稀疏度进行了预估计。结合序贯压缩检测技术，给出了一种基于该算法的多认知用户合作场景下的宽带压缩频谱感知方法，理论分析和实验仿真结果表明，该方法可在频谱稀疏度先验知识缺少的情况下，有效提高宽带频谱感知性能。

Traditional spectrum sensing based on compressed sensing assumes that the sparsity is known, while actually, it is unknown. To solve the problem, we proposed a modified sparsity adaptive matching pursuit (MSAMP) algorithm. The MSAMP algorithm pre-estimates the value of sparsity in the process of choosing the support set. A cooperative wideband spectrum compressed sensing method is developed based on the MSAMP algorithm and sequential compression inspection technique. Theoretical analysis and simulation results show that the method can enhance the spectrum sensing capability without a priori knowledge of sparsity.