Algorithm 1: SOMP |
---|
(1): Input: training sample matrix , testing sample matrix |
(2): Output: Row-wise sparse solution |
(3): Initialization: residual , , non-zero rows , i = 0 |
(4): While stopping criterion false do |
1). Find a new atom from matrix A to best approximate the current residual based on q-norm: |
2). Update the non-zero row support . |
3). Update the signal estimation , where denotes the sub-matrix of A consisting of the atoms from matrix A, and the residual: . |
4).i = i + 1. |
(5): End while |
(6): Return: |