Abstract
This publication describes systems and techniques for implementing an adaptive, self-tuning hysteresis timer that determines when to transition a device to a low-power state to balance power savings and performance based on individual user behavior. Hysteresis timers are commonly used to transition a device to a low-power state after an idle period without user activity, then transition the device back to an operating state once the user reengages the device. Hysteresis timers typically are preprogrammed with one or more static hysteresis thresholds that specify an interval between a user’s last engagement with the device and when the device transitions to a low-power state. If the static hysteresis threshold is too long, power may be wasted; if the static hysteresis threshold is too short, user experience suffers because the device transitions to a low-power state too quickly between interactions. The disclosed systems and techniques include an adaptive, self-tuning hysteresis timer that adjusts the hysteresis threshold based on the user’s pattern of activity. Utilizing an integral derivative feedback loop, the duration of idle periods between consecutive wakeup events is measured. The hysteresis threshold is iteratively updated to converge toward a duration that represents the average idle period plus a specified bias. The systems and techniques allow the hysteresis threshold to adapt to changing usage patterns within a few clock cycles. Consequently, the need for manual tuning is reduced, and power management is optimized for individual user behavior, rather than the behavior of an average, hypothetical user.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Majumder, Anirban; Gupta, Aditya; Sardana, Tushar; and Sangaiyah, Pandithurai, "Adaptive, Self-Tuning Hysteresis Timer with Integral Derivative Feedback Loop for Power Management", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/9991