Sparse Detection With Integer Constraint Using Multipath Matching Pursuit

Byonghyo Shim, Suhyuk Kwon, Byungkwen Song

Abstract: In this paper, we consider a detection problem of the underdetermined system when the input vector is sparse and its elements are chosen from a set of finite alphabets. This scenario is popular and embraces many of current and future wireless communication systems. We show that a simple modification of multipath matching pursuit (MMP), recently proposed parallel greedy search algorithm, is effective in recovering the discrete and sparse input signals. We also show that the addition of cross validation (CV) to the MMP algorithm is effective in identifying the sparsity level of input vector.