How Much Timing Information Is Enough? Awake Hardware Evidence and Wake/Listen Energy Models for PToB
How little new timing information could sustain a useful clock between listens?
Overview
Six randomized awake T096 runs used 125 ms training windows and 5, 10 or 20 second HOLD intervals. Only 42 of 120 windows refreshed the fit; 78 were empty. The study combines that information-budget pilot with explicit guard, drift, wake and reacquisition models.
What this paper covers
- Distinguish a training aperture from physical radio duty and measured energy.
- Keep inherited fit history, missing pairs and post-turnover evidence visible.
- Define retained-counter sleep and metered-power experiments that could test the proposed savings.
Research scope
Radio reception and the precision counter remained active. Scheduled training fractions are not measured airtime or energy savings. Only one run replaced all acquisition pairs, leaving 10.17 seconds of post-turnover evidence. Physical sleep, wake, power and cold-start qualification remain future work.
Read it in context
Read this earlier awake pilot separately from the 30 October W12 holdover-duration study. They use different targets, estimands and observation populations.
Cite this edition
Jonathan Yantis. How Much Timing Information Is Enough? Awake Hardware Evidence and Wake/Listen Energy Models for PToB. 19 October 2026. Research draft. Original PDF.
Summary permalink: https://ptob.dev/research/awake-timing-power-scheduling/
Jonathan Yantis leads protocol design, hardware development and research. GPT 6.1 contributes implementation, analysis and documentation. This page summarizes the linked edition; retain its original evidence dates and measurement references.