A Sequential Decoding Technique and Its Realization in the Lincoln Experimental Terminal

Abstract

Probabilistic codes when suitably matched to modulation-demodulation systems allow communications which realize the theoretical performance predicted by the coding theorem. Sequential decoding is a form of probabilistic coding which allows realization in practical equipment to achieve this end. The Lincoln Experimental Terminal (LET) uses convolutional encoding and sequential decoding matched to a modulation system employing a 16-ary orthogonal alphabet and matched filter envelope detectors, followed by an ordered list of the filters containing the 7 largest outputs. This coding system employing a constraint length of 60 bits and rates of 1 and 2 bits per orthogonal symbol achieves operation at an energy to noise ratio of 6 dB per information bit on an active satellite (Gaussian) channel. The Fano decoding algorithm is employed. After a brief description of this algorithm, the realization of the LET encoder-decoder is presented. The machine using a commercial magnetic core memory, together with about 2000 integrated circuit elements, occupies about 20 inches of 19-inch rack space.