Complex Multisnapshot Sparse Bayesian Learning for Offgrid DOA Estimation

Table 1

Complex sparse Bayesian learning algorithm for offgrid DOA estimation.

---

Input quantities: the received signals , and the sampling grid in the DOA range

Output quantities: offgrid

Initialization: giving the initial values of parameters , and setting the initial values of the convergence threshold and the maximum number of iterations max_iter, computing the array steering matrices and .

(1) Matrix construction: Computing from , , and .

(2) Parameters update:

1)Updating matrices and based on equation (19) from the current values of .

2)Acquiring the estimated values (the closest grid to the true DOA , ) from the position of nonzero rows in .

3)Updating quantities from the current values of and with equations (25) and (26).

4)Updating the offgrid parameter from the current values , , , and according to equation (32).

(3) Iteration termination: Calculating the residual value . If the residual value is less than or the number of iterations is more than max_iter, terminating the iteration process, otherwise, jumping to (1) to start the iteration, and updating the matrices , , and based on equation (33).

(4) DOA detection: DOAs will be estimated in equation (33).

---