Optimum Design of Balanced Surface Acoustic Wave Filters using Evolutionary Computation

Surface Acoustic Wave (SAW) filters are small, rugged and cost-competitive mechanical band-pass filters with amazing frequency response characteristics. Therefore, SAW filters have played an important role as a key device in various mobile and wireless communication systems such as personal digital assistants (PDAs) and cellular phones (Campbell, 1998). Recently, the balanced SAW filter becomes widely used in the modern Radio Frequency (RF) circuits of cellular phones. That is because the balanced SAW filter can provide not only the band-pass filtering function but also some external functions, namely, the unbalance-balance signal conversion, the impedance conversion and so on. In other words, if we use balanced SAW filters in a modern RF circuit, we can reduce the number of the components of the modern RF circuit, as well as their mounted area (Meier et al., 2001). As a result, we can miniaturize the modern RF circuit of a cellular phone. The frequency response characteristics of SAW filters including balanced one are governed primarily by their geometrical structures, namely, the configurations of Inter-Digital Transducers (IDTs) and Shorted Metal Strip Arrays (SMSAs) reflectors fabricated on piezoelectric substrates (Hashimoto, 2000). Therefore, in order to realize desirable frequency response characteristics of SAW filters, we have to decide strictly their geometrical structures, which are specified by using many design parameters such as the numbers of the fingers of respective IDTs, the length of the electrodes, and so on. Several optimum design methods which combine the optimization algorithm with the computer simulation have been reported to decide suitable structures of SAW filters (Franz et al., 1997; Tagawa et al., 2002; Goto & Kawakatsu, 2004; Tagawa et al., 2007). In these optimum design methods of SAW filters, Evolutionary Algorithms (EAs) such as Genetic Algorithm (GA) have been also used as the optimization algorithm (Prabhu et al., 2002; Meltaus et al., 2004; Tagawa et al., 2003). However, conventional optimum design methods of SAW filters have been contrived for the unbalanced SAW filter. Therefore, conventional optimum design methods can't be applied directly to the balanced SAW filter. The primary purpose of this chapter is to demonstrate the usage of the latest Evolutionary Computation (EC) technique in a real-world application, namely, the optimum design of balanced SAW filters. In the proposed optimum design method of balanced SAW filters, Differential Evolution (DE) is employed as the optimization algorithm. DE is one of the most O pe n A cc es s D at ab as e w w w .in te ch w eb .o rg