The Research On Test Methods Of RFID Signal Modulation

RFID is an advanced automatic identification and data capture technology which is considered as one of the top ten key technologies in the 21st century, and it has broad application prospects in production, sales distribution and security areas. As key supporting technologies, RFID tests exert a significance effect on RFID's further application. RFID tests include mainly performance testing and conformance testing together. In this thesis some tentative research has been taken on recognition of RFID signal modulation belonging to conformance testing.As a kind of whole new radio mindset, software radio is based on hardware platform and software hierarchical structure, and it achieves communication through dispatches software and hardware resources. Thus,concepts and corresponding technologies of software radio give a new solution to RFID tests faced with multi-standards and complex testing environments. Then, RFID tests system based on software radio build a new path for modulation identification of RFID signals.Wavelet transformation is characterized by its advantages both in time domain and frequency domain. Artificial neural networks are non-linear statistical data modeling tools, with their remarkable ability to derive meaning from complicated or imprecise data, can be used to extract patterns and detect trends that are too complex to be noticed. In this thesis, a new test methodology for RFID signal modulation is proposed, which is based on wavelet analysis combined with artificial neural networks. First, the thesis introduces the basic theories and the three architectures of software radio. According to the present designing condition, the wideband and band-pass sample architecture is proved to be the most practical one. Based on the architecture, the general model of multi-mode modulation is proposed. Then, the thesis discusses the advantages of wavelet analysis applying on modulation recognition, and introduce a way of feature imformation capture by wavelet analysis. After that, two classifiers respectively based on BP and RBF algorithm nets are proposed. At last, the procedures and results of simulation are demonstrated in the thesis. The results reveal that the methodology proposed in this thesis could make a sound identification among 5 RFID modulation types, including 2ASK,4ASK,2FSK,4FSK and BPSK, even with a low SNR.