Vector extrapolation accelerated maximum a posteriori superresolution algorithm for real-beam scanning radar

Convergence speed is quite important for a iterative superresolution algorithm, especially in real time processing. A vector extrapolation acceleration maximum a posteriori (VEAMAP) superresoltuion algorithm is proposed in this paper. In the algorithm a predicted point is made before initial multiplicative iteration and this can make the iterations go faster along the convergence path. As the prediction is implemented by simply extrapolating the obtained vector, only little additional computation is needed. Thus, the time required to obtained the optimal solution can be greatly reduced. Simulations and real data experiments verify the effectiveness of proposed algorithm.