An analog reconfigurable Gaussian-shaped membership function generator using current-mode techniques

This paper proposes a procedure to implement a Gaussian-shaped membership function for fuzzy systems. It is characterized by independency of any references to generate the membership function and simple adjustability of the Gaussian function parameters. Power consumption is decreased through the use of the current- mode design techniques. Higher Power Supply Rejection Ratio (PSRR) is also achieved by the use of differential structure. To verify the circuit performance, simulations are done in 0.35 μm standard CMOS technology. The power consumption is equal to 105 μW and the chip occupies an area of 400 μm². Simulation results are in excellent agreement with theoretical results and good performance of circuit is also demonstrated.