A Rank-Iteration Based Blind Identification Method for Convolutional Codes

Abstract

A blind identification technology of channel coding has received increasing attentions in recent years. In this letter, we focus on blind identification of convolutional codes without a candidate set. To improve the blind identification accuracy, a novel rank-iteration based blind identification algorithm for convolutional codes is proposed in this paper. Specifically, a weight-dependent threshold for the Binary Symmetric Channel (BSC) channel is firstly proposed to calculate the rank of a received data matrix more accurately. Then, we propose a new iterative operation that repeats the Gauss-Jordan Elimination Through Pivoting (GJETP) algorithm for several times, and select the smallest rank as the rank of the data matrix. Employing this strategy, the rank of a rank-deficient matrix in noisy environments can be close to its noiseless rank. Further, a rank-deficient threshold is introduced to amend the rank of the full-rank matrix after the end of the whole iterative process, which can reduce the rank deviation of the full-rank matrix. Finally, the effectiveness of the proposed method is verified by simulations and comparisons.