Study On Algorithms Of Signal Modulation Recognition Of Software Radio Receiver

Traditional communication radio is designed for single system with given modulation and bandwidth. It's application field is finite and it can not adapt to the system of multi-modulation, multi-service at present. For the using of multi-modulation, the precondition of receiving and demodulating signal is to decide the type of the modulation. So automatic recognition of modulation signal is a required function of software radio receiver. This paper is based on the present algorithms of modulation recognition. In the part of parameter extraction, author combines the key features of Nandi's and V.Ramakonar's and puts forward new parameters based on the study of the instantaneous information. In the part of recognition, the decision tree with simple structure, less calculation, good performance of real time is adopted. So the analogue modulation, digital modulation and the mixture of analogue and digital modulation can be recognized separately. The main works in this paper can be summarized as follows: 1. Six types of analogue modulations using the key features which put forward by Nandi etc. are recognized. It is shown that the average recognition success rate is over 98% at the signal to noise ration beyond 5dB. 2. By analyzing the instantaneous information, the new key features are extracted and the method of recognition of high-level digital modulation signals is given. The simulation proves that the performance of this algorithm exceeds that of the preceding algorithms. It not only has higher recognition success rate at the low signal to noise ration but also uses fewer key features. 3. Fourteen types of the mixture of analogue and digital modulations are recognized by the key features which put forward by Nandi etc. and in this paper. It is shown that the average recognition success rate is 100% at the signal to noise ration beyond 5dB. 4. Three algorithms are simulated separately in stationary interference and multipath environment. It is shown that the affection in Gaussian noise environment is between stationary interference and multipath environment at the same signal to noise ration. The affection in multipath environment is most severe.