Source-linked AI summary

Powering the Next Billion Devices with Wi-Fi

Vamsi Talla, Bryce Kellogg, Benjamin Ransford, Saman Naderiparizi, Shyamnath Gollakota, Joshua R. Smith

arXiv:1505.06815v1cs.NI

TL;DR

Wi-Fi routers could provide far-field wireless power, but a system was needed that preserved communication performance while using existing chipsets. PoWiFi co-designs router transmissions and harvesting hardware, enabling battery-free sensors, battery recharging, and real-world deployments.

  • Problem

    Wi-Fi routers had not been established as sources of far-field wireless power without significantly compromising network performance.

  • Method

    PoWiFi co-designs Wi-Fi router transmissions with a multi-channel harvester that efficiently harvests power from multiple 2.4 GHz channels.

  • Results

    PoWiFi minimizes Wi-Fi performance effects, powers battery-free temperature and camera sensors, recharges coin-cell batteries, and delivers power in six homes under real-world conditions.

  • Takeaways & Limitations

    Wi-Fi infrastructure can support battery-free sensing and wireless battery recharging without additional power infrastructure in the demonstrated settings.

  • Takeaways & Limitations

    The paper does not implement or evaluate concurrent power transmission from multiple PoWiFi routers, which remains outside its scope.

Abstract

from arXiv · show

We present the first power over Wi-Fi system that delivers power and works with existing Wi-Fi chipsets. Specifically, we show that a ubiquitous piece of wireless communication infrastructure, the Wi-Fi router, can provide far field wireless power without compromising the network's communication performance. Building on our design we prototype, for the first time, battery-free temperature and camera sensors that are powered using Wi-Fi chipsets with ranges of 20 and 17 feet respectively. We also demonstrate the ability to wirelessly recharge nickel-metal hydride and lithium-ion coin-cell batteries at distances of up to 28 feet. Finally, we deploy our system in six homes in a metropolitan area and show that our design can successfully deliver power via Wi-Fi in real-world network conditions.

1. INTRODUCTION

PoWiFi repurposes existing Wi-Fi infrastructure for far-field wireless power while preserving communication performance. The system combines router transmission changes with a multi-channel harvester and demonstrates battery-free sensing, battery recharging, and household deployment.

  • Motivation: PoWiFi delivers far-field power using existing Wi-Fi chipsets without significantly compromising network performance.The system repurposes Wi-Fi routers rather than requiring additional power infrastructure.
  • Motivation: Wi-Fi is suited to power delivery because it is ubiquitous, uses the unlicensed ISM band, and employs OFDM with a high peak-to-average ratio.Existing 2.4 GHz antennas can, in principle, support both communication and power harvesting.
  • Challenge: A 24-hour test found that a battery-free temperature sensor ten feet from an existing router could not reach the 300 mV operating threshold.Energy harvested during transmissions leaked during Wi-Fi silent periods, exposing a mismatch between power delivery requirements and the Wi-Fi protocol.
  • Design: PoWiFi co-designs router transmissions and harvesting circuits, injecting small broadcast packets across multiple 2.4 GHz channels to approximate continuous high-occupancy transmission.Its multi-channel harvester efficiently receives power across those channels.
  • Evaluation: 95.4% average cumulative occupancy across three 2.4 GHz channels was achieved while PoWiFi minimized effects on Wi-Fi performance.The prototype used Atheros chipsets and off-the-shelf analog components.
  • Demonstrations: Battery-free camera and temperature sensors were demonstrated, batteries were wirelessly recharged, and the system delivered power during 24-hour deployments in six homes.The rechargeable sensors supported energy-neutral operation at distances up to 28 feet.

2. UNDERSTANDING WI-FI POWER DELIVERY

Initial experiments show why ordinary Wi-Fi traffic is insufficient for reliable power delivery. Packet silence and hardware leakage keep the harvester below its minimum operating voltage, while low router occupancy limits available power.

  • Initial experiment: 300 mV was the lowest tested DC–DC converter threshold, yet a battery-free sensor ten feet from the router remained below it for 24 hours.The experiment measured rectifier output voltage alongside router packet transmissions.
  • Observed limitation: Power was harvested during Wi-Fi packets but leaked during silent slots, preventing the rectifier voltage from crossing the operating threshold.The silent periods are inherent to Wi-Fi’s distributed medium-access protocol.
  • Observed limitation: 10–40% router channel occupancy, mostly near the lower end, indicates that efficient power delivery requires higher router occupancy.Client devices typically transmit at only 0–2 dBm, lower than router transmit power.

3. POWIFI

PoWiFi combines a multi-channel RF harvester with selective router power traffic. The design addresses impedance matching, cold start, channel occupancy, and client-traffic interference to deliver power efficiently across Wi-Fi channels.

  • Multi-Channel Harvester Design: The RF harvester receives Wi-Fi signals, rectifies them into DC power, and boosts the voltage for sensors and microcontrollers.Its sensitivity determines operational range because available power decreases with distance.
  • Multi-Channel Harvester Design: Impedance mismatch reflects incident RF energy, reducing available power and complicating efficient harvesting across the 72 MHz band spanning three Wi-Fi channels.Rectifier behavior also varies with input power, frequency, and DC–DC converter parameters.
  • Multi-Channel Harvester Design: PoWiFi co-designs the matching network, rectifier, and DC–DC converter to achieve good impedance matching across Wi-Fi channels.This co-design relaxes matching-network constraints by accounting for the rectifier’s dependence on the converter.
  • Harvester Components: The battery-free converter must both boost rectifier voltage and start from 0 V despite a nonzero minimum input threshold.The battery-recharging version can use the battery’s minimum voltage and includes charging functionality.
  • Router Transmission Design: Continuous 1 Mbps transmission would maximize occupancy but significantly deteriorate Wi-Fi performance, motivating small traffic injections across multiple channels.The harvester cannot distinguish artificial packets from client traffic or beacons, so all router transmissions contribute similarly from its perspective.
  • Router Transmission Design: PoWiFi sends high-rate UDP broadcast packets and drops them when per-channel transmit queues exceed thresholds, preserving client traffic while maintaining occupancy.The mechanism uses queue status and channel-specific packet decisions to avoid adding traffic when client packets already fill the queue.

4. EVALUATION

PoWiFi is evaluated as a multi-channel power-delivery system that preserves Wi-Fi performance while maintaining high channel occupancy. Experiments assess traffic impact, neighboring-network fairness, and harvester efficiency across Wi-Fi channels.

  • Evaluation setup: PoWiFi uses three Atheros AR9580 chipsets on channels 1, 6, and 11, with 6 dBi antennas and 30 dBm transmission power.The prototype provides Internet access on channel 1 while transmitting power traffic across multiple channels.
  • Effect on Wi-Fi traffic: PoWiFi prioritizes client traffic: UDP throughput remains roughly at the baseline, whereas NoQueue roughly halves the achievable rate.BlindUDP significantly reduces throughput because its power traffic competes directly with client traffic.
  • Channel occupancy: 97.6% mean cumulative channel occupancy was achieved during UDP experiments, while other measured experiments reached 100.9% and 87.6%.Individual channel occupancies during UDP experiments were around 5–50% across the three channels.
  • Effect on Wi-Fi traffic: PoWiFi does not have a noticeable effect on TCP throughput at the client, while BlindUDP significantly degrades it.PoWiFi can sometimes exceed the baseline because channel changes occurred during the three-hour experiment.
  • Effect on Wi-Fi traffic: 101 ms average page-load delay was measured for PoWiFi, compared with 294 ms for NoQueue over the baseline.The residual PoWiFi delay is attributed to kernel per-packet checks; users did not perceive noticeable web-performance effects in home deployments.
  • Effect on neighboring networks: PoWiFi provides higher neighboring-network throughput than EqualShare because its 54 Mbps power packets occupy the channel for less time.The experiments averaged throughput across five runs, and the fairness property is expected to extend to 802.11n and other Wi-Fi variants using the highest available bit rate.

5. SENSOR APPLICATIONS

PoWiFi powers battery-free and battery-recharging temperature sensors and cameras using Wi-Fi harvesting hardware. These prototypes operate across indoor distances and can support sensing in hard-to-reach locations, including through walls.

  • PoWiFi integrates harvesters with temperature sensors and cameras in both battery-free and battery-recharging versions.
  • Implementation: The battery-free temperature sensor uses an LMT84 sensor and MSP430FR5969 microcontroller, booting when its storage capacitor reaches 2.4 V.The microcontroller requires at least 1.9 V at 1 MHz and boots in less than 2 ms.
  • Wi-Fi powered Temperature Sensor: 20 feet is the battery-free temperature sensor’s maximum operating distance, while the battery-recharging sensor remains energy-neutral up to 28 feet.Beyond 15 feet, the battery-recharging sensor has the better update rate because it is optimized for lower input power.
  • Wi-Fi powered Camera: 17 feet is the battery-free camera’s range with one image every 35 minutes; the battery-recharging camera reaches 23 feet at one image every 34.5 minutes energy-neutrally.The battery-recharging camera also operates to 26.5 feet with one image every 2.6 hours.
  • Through-the-wall operation: PoWiFi powers a battery-free camera through walls, with frame intervals increasing as wall materials absorb more signal.The through-wall test placed the camera five feet from the router and used four wall materials.

6. HOME DEPLOYMENT STUDY

A six-home deployment evaluates PoWiFi under unpredictable neighboring and client traffic. The system maintained high cumulative channel occupancy, delivered power in real-world conditions, and generally did not reduce users’ perceived experience.

  • Deployment: PoWiFi was deployed in six metropolitan-area homes for a week while occupants used replacement PoWiFi routers.The first two deployments occurred over a weekend and the remaining deployments on weekdays.
  • Channel occupancy: 78–127% was the range of mean cumulative channel occupancies across the six home deployments.Per-channel occupancy varied because carrier sense reduced transmissions when neighboring networks were heavily loaded.
  • User experience: Users in homes 1–4 reported no noticeable difference, while one user reported improved page-load and streaming-site experiences.
  • Channel occupancy: PoWiFi’s current deployment does not reduce power traffic to keep cumulative occupancy below 100%, although the authors state this can be done using occupancy measurements.
  • Sensor performance: Battery-free temperature sensors placed ten feet from the routers successfully received Wi-Fi power under real-world network conditions.The result was obtained from computed update-rate distributions across the homes.

7. RELATED WORK

Prior work established RF harvesting from RFID, ambient broadcasts, and 2.4 GHz signals, but did not demonstrate powering devices with existing Wi-Fi chipsets. PoWiFi addresses this gap through high cumulative occupancy over multiple Wi-Fi channels.

  • RFID: RFID systems have powered microcontrollers, LEDs, sensors, and cameras from dedicated 900 MHz readers, making PoWiFi complementary to RFID.
  • Ambient RF harvesting: Ambient TV and cellular harvesting is stronger outdoors but attenuates indoors, whereas Wi-Fi can augment these approaches for indoor power harvesting.
  • 2.4 GHz harvesting: Earlier 2.4 GHz harvesting studies demonstrated power harvesting from continuous-wave transmissions, but none powered devices with existing Wi-Fi chipsets.
  • Contrast with prior work: PoWiFi masks Wi-Fi traffic burstiness by creating high cumulative channel occupancy at the router rather than relying on continuous-wave transmissions.
  • Other wireless-power systems: A back-of-the-envelope estimate for competing 1 W, 15-foot systems requires 83.3 dBm EIRP, or 213 kW, which would exceed FCC point-to-multipoint limits by 50,000 times.The passage contrasts this with PoWiFi’s operation within FCC limits and minimal impact on Wi-Fi traffic.

8. DISCUSSIONS AND FUTURE DIRECTIONS

The paper discusses extensions of PoWiFi to USB charging, MIMO, multiple routers, and future ISM-band designs, while identifying security implications. These directions preserve the goal of delivering power with minimal impact on Wi-Fi traffic.

  • Wi-Fi router as a charging hotspot: PoWiFi can act as a wireless charging hotspot for devices such as FitBit and Jawbone activity trackers through a general-purpose USB charger.
  • Wi-Fi router as a charging hotspot: A 2 dBi antenna attached to a custom harvester charged a Jawbone UP24 from empty to 41% in 2.5 hours.
  • PoWiFi with MIMO: The authors state that MIMO and PoWiFi have no fundamental tradeoff because concurrent transmissions can use multiple antennas and adjacent channels.
  • Multiple PoWiFi routers: Multiple PoWiFi routers could transmit power packets concurrently because collisions are acceptable for UDP broadcasts that need not be decoded by a specific client.Implementing and evaluating this multi-router solution is outside the paper’s scope.
  • Security implications: A rogue device could cause power starvation by generating signals that trigger carrier-sense events at a PoWiFi router.The paper identifies efficient security mechanisms against this power denial-of-service threat as an open research opportunity.
  • Future clean-slate designs: Future designs are intended to generalize PoWiFi’s multi-channel approach while avoiding solutions that jam specific frequencies or deteriorate Wi-Fi performance.

9. CONCLUSION

The paper addresses the challenge of powering smaller, more numerous IoT devices by introducing far-field power delivery through existing Wi-Fi chipsets and demonstrating it in real homes.

  • The system targets the difficulty of powering increasingly small and numerous IoT devices without inconvenient, large-scale plugging-in.
  • The authors introduce far-field power delivery using existing Wi-Fi chipsets and prototype battery-free temperature and camera sensors.
  • The system also demonstrates wireless recharging of nickel–metal hydride and lithium-ion coin-cell batteries.
  • Deployment in multiple metropolitan-area homes demonstrates successful power delivery via Wi-Fi under real-world network conditions.
Loading 1505.06815v1…