The multiscale distributed hierarchical fusion algorithm for multisensor inverse problem

We develop a distributed hierarchical fusion inversion algorithm to multi-sensor inverse problem, which was motivated by the advantage of characterizing signal at different resolution levels of the wavelet transform. When we use the criterion based on relative error covariance matrix (RECM) to evaluate the performance of the fusion inversion algorithm, we show that the distributed hierarchical fusion technique resulted in an estimate comparable to that of [Eric Miller et al., 1995], and that it efficiently alleviates the computational burden of the central processor.