A novel method for identifying recursive systematic convolutional encoders based on the cuckoo search algorithm

Abstract

The existing methods for identifying recursive systematic convolutional encoders with high robustness require to test all the candidate generator matrixes in the search space exhaustively. With the increase of the codeword length and constraint length, the search space expands exponentially, and thus it limits the application of these methods in practice. To overcome the limitation, a novel identification method, which gets rid of exhaustive test, is proposed based on the cuckoo search algorithm by using soft-decision data. Firstly, by using soft-decision data, the probability that a parity check equation holds is derived. Thus, solving the parity check equations is converted to maximize the joint probability that parity check equations hold. Secondly, based on the standard cuckoo search algorithm, the established cost function is optimized. According to the final solution of the optimization problem, the generator matrix of recursive systematic convolutional code is estimated. Compared with the existing methods, our proposed method does not need to search for the generator matrix exhaustively and has high robustness. Additionally, it does not require the prior knowledge of the constraint length and is applicable in any modulation type.