Self-Adjustment Channel Rank Based on Reordered Cholesky Factorization
Résumé
This paper presents a self-adjustment of the length of the channel
impulse response using a classical Cholesky Factorization. This is
of special importance in the context of algorithms containing a
matrix inversion and the rank of this matrix is strongly related to
the length of the channel impulse, i.e. interferer cancellation
algorithm.
When the length of the channel impulse response is overestimated,
the matrix inversion is carried out with a singular
matrix. A self-adjustment is then required to avoid an increase in
the noise level.