1: \begin{abstract}

2: %\setstretch{0.5}

3: \baselineskip 0.2in

4: %

5: The  multiple measurement vector  problem (MMV) is a generalization of the compressed sensing problem that addresses the recovery of a set of  jointly sparse signal vectors.

6: One of the important contributions of this paper is to reveal that the seemingly least related  state-of-art MMV joint sparse recovery algorithms - M-SBL (multiple sparse Bayesian learning) and subspace-based hybrid greedy algorithms - have a very important link. More specifically,  we show that  replacing the $\log\det(\cdot)$  term in M-SBL  by a  rank proxy that exploits the spark reduction property  discovered in subspace-based joint sparse recovery algorithms, provides significant improvements. In particular, if we use the  Schatten-$p$ quasi-norm as the corresponding rank proxy, the global minimiser of the proposed algorithm

7: becomes identical to the true solution as $p \rightarrow 0$.

8: Furthermore, under the same regularity conditions,  we show that the convergence to a local minimiser is guaranteed

9: using an alternating   minimization algorithm that has closed form expressions for each of the minimization steps, which are convex. Numerical simulations under a variety of scenarios   in terms of SNR,  and condition number of the signal amplitude matrix demonstrate that the proposed algorithm  consistently outperforms M-SBL and other state-of-the art algorithms.

10: \end{abstract}

11: