Abstract: Embodiments in accordance with the present disclosure are directed to methods and apparatuses used for slow wave activity (SWA) optimization. An example method includes receiving one or more bio-signals from a user and classifying sleep stages by processing the bio-signals. The method further include determining dominant peripheral nervous system (PNS) oscillations based on the bio-signals and as a function of time and stage of sleep, and characterizing at least one property of the dominant PNS oscillations, including a phase, a phase shift, an amplitude, and/or frequency. The method further include providing an indication of an optimal window for maximizing SWA generation based on the phase, the phase shift, the amplitude, or the frequency. The indication is provided to stimulation circuitry that delivers stimulation to the user within the optimal window. Feedback is provided responsive to the stimulation based on an EEG signal of the user.