对于服从广义高斯分布(Generalized Gaussian distribution,GGD)的稀疏信号进行了参数分析.首先给出了广义高斯分布信号的一些性质,通过对信号等高线的分析,导出了计算稀疏性参数的公式,通过该公式的计算可以得到,对Laplace信号稀疏性参数为1,对Gauss信号为2.参照Laplace信号和Gauss信号,对于给定的服从广义高斯分布的信号,通过稀疏度量的计算可以直观地知道它究竟多么稀疏.实例表明只有当信号充分稀疏时才能通过稀疏表示方法实现欠定盲源分离.
The parametric analysis is presented for the sparse signal followed generalized Gaussian distribution (GGD). At first, the properties of GGD signal are discussed. A mathematical formula is established to compute the parameter of sparseness. It is shown that the parameter of Laplacian signal is 1, and that of Gaussian signal is 2. For a given GGD signal, comparing with Laplacian signal and Gaussian signal, we can intuitively know how sparse it is by calculating the sparse parameter. Two examples are given to illustrate the fact that only when the source signals are sufficient sparse, we can achieve underdetermined blind source separation (BSS) by sparse representation.