Optimal Design of a Biomedical Amplifier for Minimum Offset Using a Modified ABC Algorithm

The optimal design of an Improved High Swing Self-bias Folded Cascode (HSSFC) amplifier with current scaling considering minimum random offset voltage as the design objective under various design constraints for application in biomedical systems has been presented. The modified ABC algorithm with selection strategy (MABC-SS) has been employed for this purpose. MABC-SS uses a novel rate of change approach to modify the initial population and bee position based on average cost. It has been reported to demonstrate significant improvement in optimal results for benchmark mathematical functions. This work demonstrates the novel scope of the MABC-SS for designing an optimal biomedical amplifier, which has not been reported previously. Biomedical Amplifiers have a stringent requirement of achieving minimum specifications for various design parameters such as gain and noise for optimal performance. Offset voltage has a direct impact on the precision and accuracy of these parameters and as such needs to be minimized. Biomedical signals are furthermore characterized by low amplitude and frequency, so a high offset leading to error in the results is undesirable. This chapter focuses on the random offset modeling of the biomedical application-based HSSFC amplifier using Pelgrom's Design Model, which yields a cost function involving the offset voltage standard deviation. MABC-SS is then used to resolve the subsequent cost function while meeting the least possible design constraints like noise and gain, among others. Comparative analysis with state-of-the-art techniques is performed in terms of the various statistical parameters and convergence speed, which further demonstrates the efficiency of the proposed approach. Eventually, the algorithmic results are verified using a redesign of the circuit using the best design parameters in the Cadence circuit 180 nm technology design tool. Simulations and Monte Carlo analysis are then carried out, and the results illustrate an excellent match with the computational results.