A novel solution based on multi-objective AI techniques for optimization of CMOS LC-VCOs

A method of optimizing components and transistors sizing for cmos cross-coupled lc voltage controlled oscillators is presented in this paper. the design constrains of power consumption,phase noise,and the figure of merit (fom) of lc-vcos are applied on multi-objective ai techniques,simultaneously. the design parameters of lc-vcos are obtained from the two strong algorithms,the multi-objective inclined planes system optimization (moipo) and the multi-objective particle swarm optimization (mopso). it was implemented in matlab,to the pareto optimal front (pof) solutions,which have an amazingly trade-off between three objective functions. the lc-vco circuits were simulated using this method in a 0.18μm-cmos process by hspice rf environment. the results show that the size of components of the pmos-only and nmos-only integrated lc-vcos and the optimal trade-off curve between minimum power and minimum phase noise.