Time-invariant spatially coupled low-density parity-check codes with small constraint length

We consider a special family of spatially coupled low-density parity-check (SC-LDPC) codes, that is, time-invariant low-density parity-check convolutional (LDPCC) codes, which are known in the literature for a long time. Codes of this kind are usually designed by starting from quasi-cyclic (QC) block codes, and applying suitable unwrapping procedures. We show that, by directly designing the LDPCC code syndrome former matrix without the constraints of the underlying QC block code, it is possible to achieve smaller constraint lengths with respect to the best solutions available in the literature. We also find theoretical lower bounds on the syndrome former constraint length for codes with a specified minimum length of the local cycles in their Tanner graphs. For this purpose, we exploit a new approach based on a numerical representation of the syndrome former matrix, which generalizes over a technique we already used to study a special subclass of the codes here considered.