Blind Estimation Parameters Of Complex Linear Block Code

The blind identification of channel codin is a heated research topic in the field of non-cooperative communication,which is widely used in cryptography,information interception and electronic confrontation.In the field of non-cooperative electronic reconnaissance area,the first step from the signal layer into the information layer is identify the type of channel coding correctly and its coding parameters from the intercepted information stream to achieve the correct decoding,which is of great significance.Linear block code is commonly used in channel coding.In this thesis,we will focus on the blind estimation method of linear block codes.The mainly work is listed as follows:1.The fundemantal knowledge and important properties of abstract algebra needed in channel coding based on the discussion and implementation of the blind estimation parameters for the linear block code.2.A blind estimation parameters of BCH-product code is proposed.Using the correlation between the code elements,the estimation of the code deviation and the code length is done by the normalization of the data matrix.Based on the estimation method of decoding matching,the estimation of the primitive polynomial part in the generator polynomial is completed by the statistical code word correct rate,and the nonproductive polynomial of the least poly nomial is deduced.The product of the two is the final generate polynomial.3.A blind estimation parameters of RS code based on primitive element is proposed.By dividing the number of code words containing the factors in the RS code,the code length and the primitive polynomial are completed.Using the RS code to generate polynomials is composed of continuous roots with primitive elements.After filtering the error code words,the estimation of the generate polynomials can be done by verifying the continuous roots.This algorithm has the advantages of low complexity and only requires a small number of correct codewords in the symbol sequence to complete the estimation.Therefore,even in the case of longer code length,the algorithm also has strong error resistance.4.A blind estimation parameters of RS code based on partial codewords is proposed.The core idea of the algorithm is to use the nature of the RS code equivalent quasi-cyclic code,because there is a correlation between the sub-code,and the estimation of RS code length is obtained by extracted the dimension of each sub-code space.The advantage of this algorithm is that it only need to use part of the codeword to complete the parameter estimation,which saves the spatial complexity of the algorithm operation and runs faster than using the complete codeword in the Galois-extended domain more than doubled.5.A blind estimation algorithm for deleting RS code is proposed.By improvement the traditional Euclid algorithm,the use of two different to deleting RS code-words to find the maximum common factor is still equal to the generate polynomial.Assuming that the maximum common factor between the two different symbol codewords under different parameters satisfies the condition.This algorithm is more resistant to error,the code length is longer and the bit error rate is lower than 0.001 case,still can reach more than 90% recognition probability.