An adaptive channel SVD tracking strategy in time-varying TDD system

An adaptive channel singular value decomposition (SVD) tracking scheme is discussed for time division duplexing (TDD) systems. The proposed strategy is based on the signal subspace decomposition. The advantage of the scheme is that only the unitary matrix U and singular matrix /spl Sigma/ need to be tracked. The proposed algorithm reduces stringent processing requirements significantly and easy to implement in real time applications. The algorithm is simulated by using i.i.d. channels and the indoor channels measured at Victoria University of Technology. The complexity of performing SVD reduces from 26a/sup 3/ to 16a/sup 2/, compared with using R-SVD algorithm.