Research On The Technology Of Convolution Code Blind Identification

Channel coding technique is used to ensure the quality of information transmission and improve the reliability and validity during the transmission. The encoded information has the ability of error detection and correction automatically by increasing the number of redundant information artificially. In the non-cooperative communication system, channel coding type and parameters are unknown, which requires the channel coding identification method to identify relevant parameters to obtain reliable and valid information to make timely and correct response. Channel coding identification has a very important significance in communication reconnaissance.In this paper, we studied the blind identification algorithm for common convolution codes,and solved the identification of packet starting point, code length, rate, register length and parity check matrix. Existing algorithms assumed that packet starting point and other information is known, and they will not be able to play an effective role with the unknown parameters in these circumstances, which cannot achieve the purposes of blind identification. In addition, existing algorithms have some other disadvantage, including the poor fault tolerance and requiring a great amount of data.This paper proposed a blind identification algorithm for convolution code algorithms based on previous studies, which does not require any prior knowledge. This article first proposed a fast blind frame synchronization code recognition algorithms to overcome the disadvantages in traditional methods. In this paper, the improved method based on data matrix model for code length information bit length and register length identification achieved good fault-tolerant performance and reduced the calculation. And then replace the traditional Walsh-Hadamard transform with the fast Walsh-Hadamard transform to identify the check matrix, which greatly reduced the amount of calculation and achieved good recognition results. Computer simulation and practical tests show that the algorithm proposed can achieve a good identification performance and meet the demands in practice.