The Research On Test Technology Of RFID Signal Modulation At Low SNR

RFID has many advantages of transfering and disposing information when no direct contact,recognizing multiple targets synchronously,and be applied in many harsh environments.At present,it is becoming the hot technology of IT.RFID test is the technology presupposition of its application,and the research on RFID test technology and building of open platform has become Urgent problems.RFID test includes technology test and application test,the former tests weather the performance of RFID product and reliability Conform to Standard,and the latter tests the application in the Special environment.RFID modulation recognition belongs to technology test.With the development of RFID, the coexistence of multiple standards is made up.Because of modulation's diffirence in various standards,it is difficult to communicate each other.Accordly,correct identification of RFID modulation method is precondition for relizing communication each other,RFID test and so on.With the Increasing sophistication of communication environment,it sets higher requirements for SNR(signal-noise ratio) of modulation identification,and at low SNR research on RFID modulation recognition has becom hot research hot in RFID test at home and abroad. This paper proposes a wavelet filter,which pre-processes recognition signal by using self-study threshold wavelet denoise, at low SNR on common digital modulation modes in RFID standard. And it adopts a feature parameter Ra, which is the inverse of R and aviods false positives by slight R error. In the end, it designs a recognition classifier of GA-BP (genetic algorithm-back propagation) nerual network,which gives full play to global optimization of GA and self-learning of neural network.It can improve the classification effectiveness by Combining with the former.