Channel Identifiability for Blind Subspace-Based Channel Estimator in Uplink MC-CDMA Systems

Subspace-based channel impulse response (CIR) estimators have been proposed for multi carrier code division multiple access (MC-CDMA) systems. Channel identifiability conditions have been formulated for blind (non-data-aided) CIR estimators. These identifiability conditions require the channel filtering matrix to satisfy certain rank conditions. In this paper, the generalized uplink MC-CDMA system model is analyzed in detail. It will be shown that each user´s signal subspace dimensionality is independent of the user´s spreading code and CIR. Furthermore, for any realization of the users´ CIR, the channel identifiability is established by only imposing a simple rank condition on the users´ spreading codes.