Superfast Reduced-Redundancy Block Memoryless Linear Equalizers

This paper shows how superfast block memoryless linear equalizers should be formulated, considering reduced redundancy multicarrier and single carrier architectures. We show how to choose redundancy optimally so as to minimize the noise power at the receiver, along with an exact interpretation of the equalizer parameters as efficiently computed Kalman variables. While some recent references have considered well known DFT-based factorizations for minimum-redundancy transmissions, besides giving a broader treatment on the subject, we show that several statements and conclusions by certain authors with respect to the performance of those schemes are misleading. Simulations are included in order to corroborate our claims.