Blind adaptation of a matched filter using the constant modulus algorithm coupled with an optional correction method

Blind correlation-based DFE receiver for the equalization of single input multi output communication channels

“…(*) is the convolution operator. We can conclude from the derivation of the above that the effective channel over the DFE block of the architecture in Blind Calculation of the Matched Filters: As being mentioned earlier in the Introduction, the unknown matched filter can be estimated blindly by the use ofthe CMA, if the matched filter is the sole unknown [7]. The said condition is satisfied, once the DFE weights are calculated with sample averaging.…”

“…The wellknown blind channel estimation methods in the literature involves implementation inefficient and expensive mathematics, such as Singular Value Decomposition (SVD) [4] Eigen Value Decomposition (EVD) [5] or the need for the use of Higher Order Statistics (HOS) of the received data [6]. In the work [7], Ozcelik et. al.…”

“…Constant Modulus Algorithm (CMA), if the matched filter is the sole unknown in the receiver. To build the multi-dimensional matched filter in our blind SIMa receiver, we have benefited from the findings of [7].…”

Multi-dimensional matched filter identification technique for channel equalization deployed in spatial diversity receivers

“…Therefore, the adaptation of the receiver in [4] for SIMO channels would both improve the efficiency and performance in addition to its computational cost.…”

All-Adaptive Blind Matched Filtering for the Equalization and Identification of Multipath Channels—A Practical Approach

“…The Blind Matched Filter (BMF) receiver [1], [2] aims to implement Forney's receiver [3], which is the optimum receiver in the presence of ISI and Additive White Gaussian Noise (AWGN), blindly without the need for the explicit knowledge of the CIR. For this purpose the BMF approach makes use of an adaptive filter, updated via a computationally simple 1) The communication doesn't depend on the transmission of training signals, as the use of training symbols, which are carrying no valuable information, making them inherently inefficient and wasting the bandwidth of the communication channel as well as resources at both ends, and 2) Neither matrix decomposition methods, which are implementation inefficient and expensive (such as [4], [5] using Singular Value Decomposition (SVD) and [6], [7], [8] using Eigen Value Decomposition (EVD)) nor higher order cumulant functions of the received data, e.g.…”