A linear interference canceller with a blind algorithm for CDMA mobile communication systems

This paper proposes a new blind algorithm for a linear interference canceller that can increase the capacity of CDMA cellular systems. An orthogonalizing matched filter (OMF) is one of the effective linear interference cancellers. OMF employs the constrained minimum mean square (CMMS) criteria for controlling the combining coefficients, and operates as the decorrelator, or the orthogonalizing filter, under the interference dominant condition and as the matched filter in the noise dominant condition. The new blind algorithm evaluates a steering vector which is essential for achieving good cancelling performance. Singular value decomposition (SVD) of the received signal correlation matrix constitutes a time-varying code vector space, and the steering vector embedded in that space is obtained and utilized. The effectiveness of the blind method is verified by computer simulations under non-fading and fading environments.