Equivariant adaptive selective transmission

In this paper, we consider the problem of selective transmission-the dual o f the blind source separation task-in which a set of independent source sig nals are adaptively premixed prior to a nondispersive physical mixing proce ss so that each source can be independently monitored in the far field. Fol lowing similar procedures for information-theoretic blind source separation , we derive a stochastic gradient algorithm for iteratively estimating the premising matrix in the selective transmission problem, and through a simpl e modification, we obtain a second algorithm whose performance is equivaria nt with respect to the channel's mixing characteristics. The local stabilit y conditions for the algorithms about any selective transmission solution a re shown to be the same as those for similar source separation algorithms. Practical implementation issues are discussed, including the estimation of the combined system matrix and the reordering and scaling of the received s ignals within the algorithm. Mean square error-based selective transmission algorithms are also derived for performance comparison purposes. Simulatio ns indicate the useful behavior of the premixing algorithms for selective t ransmission.