Efficient clustering techniques for channel equalization in hostile environments

On the least-squares performance of a novel efficient center estimation method for clustering-based MLSE equalization

“…Using (6) in (7) 3 Clearly, the above procedure does not apply when L 2 A different approach must be taken in this case [5]. In the rest of the paper it will be assumed that L 2…”

“…It belongs to the class of the so-called Clustering-Based Sequence Equalizers (CBSE) (e.g., [3]), since it is based on the idea that the set of all possible (noiseless) channel output values, needed at the Viterbi stage, are simply the centers of the clusters formed by the received observations at the receiver front end and can thus be estimated from the noisy observations via a clustering approach. In contrast to earlier CBSE methods, however, which appeal to clustering in a high-dimensional space defined by successive observations, 1 the novel algorithm operates in a one-dimensional space [6].…”

Fractionally Spaced Clustering Based Equalizer for Optical Channels

“…In this paper, a Fractionally spaced Clustering Based Sequence Equalizer (Fs-CBSE) is proposed which treats equalization as a classification task [4,5]. The proposed method focuses on the clusters, which the received data form.…”

“…Matrix Σ i can be similarly estimated. The number of clusters formed (q) is 2 L+2 , for the two dimensional space with binary data and for a channel length equal to L. Details about the CBSE are given in [4,5].…”

“…As a possible alternative, equalization methods based on recursive Volterra nonlinear filters have been proposed, [3]. In this paper, a new equalization algorithm based on data clustering techniques [4,5] is proposed for high capacity optical communication channels. The major advantage of the proposed method is that no explicit channel modeling is required.…”