Development and Performance Evaluation of Optimal Low Phase Noise and Wide Tuning Range Current-Starved VCO Using Multi-objective Salp Swarm Algorithm

This paper deals with the optimal design of a current-starved voltage-controlled oscillator (CSVCO) utilizing a bio-inspired optimization technique, the multi-objective salp swarm algorithm (MSSA). The optimal design problem consists of several conflicting objective functions. Different performance parameters including phase noise, dynamic power consumption, tuning range (TR), area, and figure of merit (FOM) of the voltage-controlled oscillator are considered the objectives for optimization, while oscillation frequency is regarded as a constraint. The optimal transistor sizing of the VCO circuit is achieved using the MSSA, and the MSSA-based optimal transistors are applied to design the CSVCO circuit in Cadence software using the 1.2 V/2.5 V 1P9M 90 nm CMOS salicide process. The designed VCO oscillates at a frequency of 2.00003 GHz, and it shows phase noise of −92.09 dBc/Hz at 1 MHz offset frequency while consuming 397.86 µW power. The tuning range is 33.05%, and the physical area is 0.647 µm². The proposed design yields a 0.0015% variation in the oscillation frequency. Monte Carlo simulations are performed on the parasitic extracted netlist of the CSVCO to determine the effect of the process variations on its different performance parameters. The MSSA-based optimal design of the parasitic and process-variation-aware CSVCO circuit produces better performance parameters relative to those of previously reported studies, particularly in terms of the oscillation frequency, phase noise, and power consumption. This proposed methodology can also be applied to different very large-scale integration (VLSI) analog circuits to enhance the different performance parameters.