Research On Information Interception Techniques Of Non-cooperative Communications Based On Convolutional Codes

Convolutional code is widely used in various kinds of communication systems because of its simplicity and good error correcting performance. In non-cooperative communication and intelligent communication systems, the receiver holds nothing or just a little information about the parameters of the sender, so the modulating and coding parameters of the sender need to be identified to gain the information. Therefore, the research of blind recognition technologies based on convolutional codes is one of the key techniques for information interception, which also has very important significance.Based on the study of blind recognition of convolutional codes and RCPC codes, the thesis studied the process of intercepting information of a communication system. The system uses convolutional code as channel coding. Besides, the research about cryptographic parameters was realized with C language.The main work and innovation of the thesis are as follows:(1) The thesis studied the blind recognition of error-free (n,k,m) systematic convolutional codes and (n,l,m) non-systematic convolutional codes with errors. For error-free systematic convolutional codes, the thesis recognized parameters by the way of simplifying matrix. Parameters mainly include the length of code, starting point, the length of information bits, constraint length, generation matrix and check matrix. A complete fault-tolerant identification method for (n,1,m) convolutional codes was proposed and verified by simulation. Code length and starting point used reduced row echelon form to identify. Then Euclidean algorithm or Walsh-Hadamard transformation was applied to identify the generation polynomials for (2,1,m). Meanwhile, the thesis presented the relevant proof of Euclidean algorithm. On the basis of these studies, a complete polynomial algorithm was proposed to expand from (2,1,m) to (n,l,m) convolutional codes.(2) A set of simulating software was established based on MFC and C language, which included blind recognition algorithms of error-free (n,k,m) systematic convolutional codes and (n,l,m) convolutional codes with errors. Each part included two modules for data generator and recognition of parameters. Euclid and W-H algorithms can be chosen to recognize parameters. The set of software was displayed by the way of user-friendly graphics which was easy to use and operate. After testing, the software can recognize (n,k,m) systematic and (n,l,m) non-systematic convolutional, whose code length ranges from 2 to 8, constraint length is less than 13 and generation matrix generated randomly.(3) Kinds of methods for blind recognition of RCPC codes were investigated and realized. Code length and starting point was analyzed by reduced row echelon form. The check polynomial was recognized by Fast Walsh-Hadamard transformation (FWHT). The thesis also modified an existing algorithm for PWHT to correct its errors so that the results were accurate. The methods which recognized generator matrix and puncturing pattern of RCPC codes, were introduced in detail on the foundation of the relation of generation and check matrix. Moreover, the thesis presented the method of establishing equations and the completing process for blind recognition of RCPC codes.(4) The thesis studied blind identification algorithm of cryptographic parameters in a communication system related to convolutional codes, which provided the necessary conditions for the information interception. After the presentation of system encryption processes, principles and implementation processes of the research over traffic channel were discussed in detail. The thesis also applied the algorithm to fast associated control channel. A method for blind identification of code parameters was proposed and simulated based on the nature of convolutional code and combination of two-way data.(5) In the fast associated control channel transmission side, the data was transferred after CRC check, convolutional coding, interleaving, scrambling and enciphering. Primary task of gaining information was blind identification of cryptographic parameters. The thesis introduced the implementation process of the identification of password parameters in detail with related parameters. The process included calculations of parity check matrix and vector, computation of the session key and initial state of the encryptor. In the process of calculating the initial state of the encryptor, the thesis set numbers for the registers of the encryptor to solve the issue of establishing the coefficient matrix. Meanwhile, a new termination condition of iterative algorithm was proposed by getting partial results and verifying the equation to improve the computing speed.