Enhancing sparsity in linear prediction of speech by iteratively reweighted 1-norm minimization | IEEE Conference Publication | IEEE Xplore