Learning to Communicate Effectively Between Battery-free Devices
Talkat 19th USENIX Symposium on Networked Systems Design and Implementation (NSDI '22) · 2022-04-04..04-06·
summary
Bonito builds pairwise connections between intermittently powered devices by having each node learn its charging-time distribution with online SGD and exchange models to pick a common wake-up interval via the inverse joint CDF. A battery-free prototype on an nRF52 with a boost converter and a 47 µF storage capacitor sustains connections across hundreds of wake-ups, incurs about 25% energy for learning but saves far more by avoiding re‑synchronization, and yields 10–80× throughput improvements in experiments.
AI-written summary · about
chapters (16)
- 0:00Intro
- 0:15Going Battery-free
- 0:42Intermittent Operation
- 1:14Battery-free Networks
- 2:01Key Problem: Interleaving
- 2:42Real-world Data
- 3:55Bonito Protocol
- 4:30Connection Interval
- 5:11Bonito Overview
- 5:49Predicting Charging Times
- 6:53Learning Distribution Paramet
- 8:03Computing the Connection Int Model node 1
- 8:56Implementation
- 9:25Evaluation
- 9:54Example
- 10:44Key results
related talks
10mBootstrapping Battery-free Wireless Networks: Efficient Neighbor Discovery and
06mIntermittent Learning: On-Device Machine Learning on Intermittently Powered System
01h 19mEnergy and Efficiency
16mAlterWear: Battery-Free Wearable Displays for Opportunistic Interactions
06mBFree: Enabling Battery-free Sensor Prototyping with Python
45mWorkshop: Building a device that can run for years on small batteries
53mBecome a Wireless Connectivity Ninja 101
46mDevice Power Management: Journey to 5uA