RF Antenna Adaptive Impedance Matching Based On Genetic Algorithms Method

RFID (Radio Frequency Identification, RFID) is a developing automaticidentification technology stemming from90s in the20th century. Non-contact andtwo-way communications are adopted to obtain efficiency object identification anddata exchange. As the RFID reader’s moving, the inductance of antenna andimpedance variability will lead to transmission power loss, and reduce therecognition ability. Therefore, the RF antenna adaptive impedance matching hasbecome a serious problem to be solved.An adaptive impedance matching method based on genetic algorithm has beenproposed in the paper.Compared with many traditional matching networks, this paper uses the π typenetwork to make the RF network matching. Unlike traditional L type network, whichhave matching restricted areas, the π type network not only can get very accuracymatching results, but also because of the simple topology, the matching speed of theπ type network is very fast.For the RF narrowband impedance matching, the problem can be considered asthe impedance matching at the center frequency of the bandwidth range because ofthe high-Q property of the circuit. On the basis, an adaptive impedance matchingmethod is proposed by adopting floating point genetic algorithm (FGA). Comparisonwith simple genetic algorithm (SGA), FGA ignores the tedious process of encodingand decoding which may avoid "Hamming cliff" problems. Moreover, the FGAoperators are more flexible than SGA operators. Therefore, the search ability of FGAis also more efficient than SGA.For the case of RF broadband impedance matching, the problem is simplified asa numerical computation in N points in the frequency bandwidth, which draws thestanding wave ratio method and the real frequency method. To solve the abovematching problem, this paper proposes an adaptive impedance matching methodbased on multilevel genetic algorithm (MLGA). In the evolution program of FGA,because of the greedy rule, FGA always appears “premature convergence”(PC).Compared with FGA, MLGA can overcome PC, because MLGA uses the principle ofasynchronous parallel. For the high frequency, the RF inductor appears non-ideal factors. In this paper,two types of the actual inductor model at2GHz are provided. The effect of thenetwork parameters such as parasitic resistance is analyzed in the paper.