A NOVEL NONCIRCULAR FOURTH-ORDER DOA ESTIMATION ALGORITHM EXPLOITING THE DIFFERENCE AND SUM CO-ARRAY CONCEPT

In this paper, a novel co-array based noncircular (NC) fourth-order MUSIC algorithm is presented for detecting NC sources. The proposed method, whose array model can be arbitrary linear array, provides a strategy for efficiently combining NC property, cumulant operation and co-array concept. As a result, the difference and sum co-array (DSCA) can be constructed. Comprising both the difference set and the sum set, the DSCA has much larger aperture than the difference co-array (DCA) and the traditional NC virtual array. Particularly, since the cumulant operation instead of the Khatri-Rao (KR) product operation is adopted for constructing the co-array, the coherent issue usually generated by the KR product operation will not occur. Therefore, the available degree of freedom (DOF) will not be reduced by half. Furthermore, the proposed method can be considered as the basic theory for further constructing a high-order DSCA with higher DOF. Simulations evaluate the performance of the proposed method.