Automatic Modulation Recognition Of Digital Communication Signals

Purpose of communication modulation signals recognition is to capture a section communication signals, under the premise of no modulation information contents, bases on less prior information, judges out the modulation method of communication signals. Along with the flying development of communication technique, the style of communication signals becomes more complex and more various. Signal environment is gradually intensive, and this condition also makes communication signals modulation recognition more difficult.Recent several decades, domestic and international scholar proceeds large quantities explore at the aspect of communication signal recognition, and acquires a lot of new recognition methods. But usually, characteristic parameter of these methods has more requests for prior information of the signals, particularly for the prior estimate of signal noise ratio. How to reduce the influence of noise? How to reduce the influence of signal noise ratio parameter in characteristic parameter? All of these problems are the topic way of the paper research.According to this premise, aim of the paper is research of digital communication modulation recognition algorithms which can resist noise influence. Put great emphasis on researching characteristic parameter extraction of digital phase modulation signals, analyzing and simulating the performance of these classifications characteristic parameter. The paper brings up new recognition arithmetic of digital communication modulation signals and its subgroup.Main research result of the paper slightly describe as follows:1. Aim at MPSK signals recognition problem based on likelihood, according to signals estimation theory, firstly the paper establish two assumption examination, secondly expand to multi assumption examination, finally strictly deduce out classification adjudge formula of maximum average likelihood, and bring up a MPSK modulation recognition algorithms (DAML_MPSK). It realizes the classification of all MPSK the signals, and give out the detailed performance analyze and simulation result.2. Via analysis of DAML_MPSK algorithms main parameter, the paper realized to simplification and predigest of this algorithms, get another MPSK modulation recognition algorithms (QDAML_MPSK). It optimized the calculation counter and the dynamic scope of signals, obvious decrease calculate complex and its performance lose not much.3. The paper complete the analysis of communication signals spectral correlation in Gardner's literature, detailed treatise and give out the spectral correlation function expression of communication signals. For the MPSK signals, particularly to 8PSK/16PSK signals, bring upped a new expression, import sequence belonged to the scope of {±1,0}, simplify for the complex that MPSK signal spectral correlation characteristic deduce, and deduce out expression of 8PSK/16PSK signals spectral correlation.4. Deeply analyzed spectral correlation function of quarter equation of MPSK signals, the paper get three unitary characteristic parameters used to modulation recognition. The three parameters have little influence from modulation signals, little influence from sampling parameters, and suffer the steady noise small etc.5. Bring up the MPSK signals modulation recognition algorithms. Primarily include below three contents: Make sure classification values of three classification parameters; Make sure the realizing process of algorithms; Give out the calculated engineering expression of algorithms.6. The paper simulate performance of MPSK signals modulation recognition algorithms under different original transmit modulation type. The simulation signal environment include added Gauss white noise channel (AWGN) and flat falling channel. In the AWGN, the paper analysis and simulate several cases, include the different signal noise ratio, and carry estimation deviation, code speed estimation deviation, molding filter, code the interference influence etc.7. The paper discuss a complete arithmetic of digital modulation recognition, firstly identify modulation vast sorts, secondly under already known modulation vast sorts, identify modulation ranks each from own modulation recognition algorithms, finally get the classification of detailed type of modulation signals, and get simulation result.