Synthesis of Passive RFID from Backscatter Using Soft-Computing Techniques

By analyzing backscatter signal from antenna, using stacked rectangular metal patches, one could determine the characteristics (poles and zeros) of the antenna. In an ideal situation, when the incident chirp contains the resonant frequencies of the patch, the scattered signal would clearly display them. The characteristic parameters could then be used as a signature of the patch, i.e., information stored in its physical structure. This could open the possibility of a new technology for cheap chipless RFID tag, if the scattered signal could be successfully analyzed, even in a noisy real world working environment. The analysis of the scattered signal poses two problems : (1) unwanted scattering from metallic objects and other noise sources in any realistic environment, (2) an additional phase shift, different for various frequency components of the chirp, due to positioning of the reception antenna at a distance. In addition, because the transfer function is non-linear, we need a fast optimization algorithm to get least-square-error parameter values in real-time. In this preliminary work, we explain how we solved these problems using soft-computing techniques and compare results with true parameter values.