Stanford engineers have demonstrated a way to transmit data using thermal noise, achieving rates in the kilobits-per-second range without generating a dedicated signal carrier at the transmitter. The laboratory work could inform wireless sensors that send small readings while using little power.
The research was published online in Proceedings of the National Academy of Sciences on October 20. Its authors are Geneva Ecola, Jida Zhang, Joseph M. Kahn and Zerina Kapetanovic.
The system is not a replacement for a commercial mobile network or a claim that communications need no electricity. It addresses a narrower problem: how a transmitter can convey information without spending energy producing its own radio carrier.
Making background noise carry information
Thermal noise comes from the random motion of electrical charges. Engineers usually try to stop it obscuring a signal. In this work, controlled differences in the noise received at an antenna carry the data.
The paper’s abstract describes a thermal-noise backscatter mechanism. Backscatter means changing the way energy is reflected rather than generating a new signal carrier.
The researchers developed a channel model, a mathematical description of how information travels from sender to receiver. It explains how switching the transmitter’s electrical impedance, its opposition to an alternating electrical signal, changes the thermal noise measured at the receiver.
They used the model to devise a transmission scheme and reported data rates more than two orders of magnitude above earlier demonstrations. Two orders of magnitude means a factor of 100.
That comparison is with earlier thermal-noise experiments. It is not a claim that the system is 100 times faster than Wi-Fi, cellular service or every competing low-power radio.
The receiver remains part of the energy bill
Stanford’s explanation describes experiments in an insulated chamber designed to keep other signals out. The transmitter switched between open and short circuits to represent the zeros and ones of digital information.
At the receiving end, an amplifier helps produce the contrast needed to distinguish those states, even though the equipment and its surroundings are at room temperature.
The absence of a dedicated transmitter carrier does not eliminate the receiver, switching circuitry or processing. A business assessing any future product would need the energy consumption and costs of the complete installation, not just the transmitting element.
A small reading sent occasionally also has a different job from a video stream. A sensor reporting a measurement may be useful with a modest data rate, provided it works reliably at the required distance and the information reaches the system that needs it.
As we explained in our coverage of the Internet of Things in business, connecting physical equipment allows its measurements and status information to be used elsewhere. The cost of maintaining that connection is part of the service.
Laboratory performance still needs field testing
Stanford says the team is now examining networks with multiple transmitters and operation in noisier surroundings outside the insulated chamber. Environmental sensing and settings sensitive to electromagnetic interference are among the applications the researchers discuss.
These remain prospective uses. The announcement does not establish a commercial launch date, a product price or a measured reduction in maintenance costs for an operating business.
For a sensor deployment, replacing batteries can involve labor and access costs as well as the price of the battery. Reducing the transmitter’s power needs could be attractive if the resulting installation remains dependable. A more demanding receiver or a shorter operating range could change that calculation.
The published model gives engineers a firmer basis for designing those experiments. A field system still has to deliver readings among competing signals, accommodate more than one device and operate at a cost that users can justify.
