The Key Technologies Of The Communication Signal Recognition And Realization

Intelligent communication system needs automatic modulation types recognition. Communication signal differentiation system plays an important role either in military or in civilian regions, and the key points, i.e. modulation recognizer and estimator, are fast developing fields in signal analysis. There are lots of articles appeared aimed at the classification algorithm of modulation signals, mostly as three methods, i.e. statistical pattern recognition framework, decision-theoretic method and artificial neural networks.This thesis fully discusses the design, academic analysis and realization of the communication signal differentiation system's four parts, i.e. pretreatment module, modulation recognizer, estimator and demodulation output module. This paper is also concerned with proposing new and less computationally demanding algorithms based on decision-theoretic approach for automatic modulation recognition of communication signals. For parameter estimator and demodulation output module, this thesis detailedly discourses upon the algorithms of all the required communication signals and compares lots of traditional methods and novel methods. Based on the technical requirement of the task, a total implement frame is presented and the system is carried out in Verilog HDL and C language on FPGA and DSP chips.Different analog and digital modulation types can be classified by the proposed flow chart, such as single-tone modulated and audio modulated AM signals, LSB, USB, CW, FM/PM, 2ASK and 2FSK. Computer simulations show that these presented algorithms are able to distinguish between the different modulation types of interest at a SNR of 10dB with average success rate 99.95% (the lowest 99.5%) for audio modulated signals, the average success rate 100.0% at a SNR of 16 dB (the lowest 100.0%) for both single-tone modulated and audio modulated signals. Therefore the performance reaches the task's requirement. The whole communication signal differentiation system based on FPGA and DSP chips has gone throug a preliminary test and good results of modulation recognition function in real environment are gained.