Researchers from Pusan National University in South Korea recently published research into a backscatter technique they argued could prove key to enabling IoT offerings spanning smart cities, industry, and public and private services.

The University team researched a technique involving the use of multiple antennas for simultaneous transmission and reception of radio signals. This so-called polarisation diversity was used alongside an accurate modelling of in-phase/quadrature (I/Q) load modulators, resulting in a system with a spectral efficiency of 2.0 bps/Hz which the Pusan team said is 40 per cent more energy efficient than traditional set-ups.

To fully appreciate the potential, it of course helps to have some understanding of what backscatter is and why it is regarded as something of a holy grail for the IoT sector.

The university alluded to the growing importance of wireless communication between devices when announcing its study.

It noted device-to-device communication has become key to services spanning “smart homes, wearable technology and industrial automation”. The team argued its technique is essential to “widespread adoption of the IoT”.

In 2020, the ITU published Backscatter Communications With Passive Receivers: From Fundamentals to Applications, in its journal on future and evolving technologies. The paper was authored by researchers at Stony Brook University, the University of Texas and Cornell University, who explained backscattering “is a form of wireless transmission based on modulated reflection of external RF signals”.

The authors argued backscattering is essential to fully realise the potential of the IoT by enabling devices to “communicate directly with each other”.

A key focus in this is enabling devices with no batteries to be deployed: the researchers focused on backscatter systems for small-sized RF tags based on passive radio receivers, noting such devices could be used to detect the frequencies used in communication links.

Passive transceivers can be used because the signal is external, “allowing devices to function in an extremely low power regime” of less than 10 microwatts. Any power needed for the transmitter “can be harvested from the external RF signal itself and, thus it is possible for such devices to be battery-less”.

Backscatter itself is not new: the Institute of Electrical and Electronics Engineers (IEEE) published research in 2019 which explained “various kinds” of communication system using the technique had been developed in the preceding 70 years, “which will enable the low-power communications as required in the IoT and green communications”.

Several other papers and journals explain backscatter is, at heart, a system capable of being powered by the wireless signal itself, with those received modulated and reflected to the source.

Most draw parallels with RFID systems.

Proposed IoT technique
The Pusan National University team focused their research on “discrepancies between simulations and real-world measurements” when modelling reflection coefficients used to choose modulation processes like Quadrature Amplitude Modulation (QAM), which the team explained are used to deliver “low bit error” and “high data” rates.

In a statement, lead researcher Professor Sangkil Kim, explained the team employed transfer learning to “accurately model” I/Q load modulators. The approach “involves applying knowledge gained from one task to enhance performance on a related” undertaking.

An artificial neural network (ANN) was trained using “simulated input bias voltages” for the in-phase and quadrature elements, to teach it how modulators typically perform across different power ranges.

The network was then fed experimental data “to predict reflection coefficients” based on the voltages.

“This transfer of knowledge enabled the ANN to improve its predictions, achieving a minimal deviation of only 0.81 per cent between modelled and measured reflection coefficients”. The “accurate models” were then used to select “optimal” 4-QAM and 16-QAM set-ups by aligning the predicted reflection coefficients with set points.

Pusan National University’s team noted the optimisation is the key to the energy efficiency of their approach, “with total consumption below 0.6mW, much lower than conventional wireless systems”.

The researchers also explored a multiple MIMO transceiver configuration involving two transmit and two receive antennas with different polarisations. The team wrote the set-up “enhances signal reception, throughput and efficiency” of backscatter, with a gain of more than 11.5dBi and cross-polarisation suppression of 18dB achieved in tests using a Vivaldi Antenna.

Emphasising the industrial potential of the approach, the researchers tested their system in the 5.725GHz to 5.875GHz frequency band, one of the ranges the ITU has set aside for industrial, scientific and medical uses.

Along with the 2.0 bps/Hz spectral efficiency, the Pusan team “attained an error vector magnitude of 9.35 per cent, indicating high reliability and efficiency in data transmission”.

Kim said the mix of “accurate circuit modelling, advanced modulation techniques and polarisation diversity, all tested in over-the-air environments, presents a holistic approach to tackling the challenges in ISC and IoT”.

Eliminating backscatter
A few weeks later, Japanese giant NTT revealed the results of a project undertaken with Okayama University into the transmission of ultrasonic radio signals which took backscatter out of the equation.

This is not a like-for-like study with that conducted by Pusan National University, but offers an interesting counterpoint regarding backscatter technology.

In the Japanese teams’ case, the focus was on filters used to identify and tune into the correct ultrasonic radio signal, those in which “matter vibrates at frequencies between kHz and GHz”.

NTT stated filters are increasingly crucial in a world in which wireless communication proliferates in uses spanning IoT, home appliances and vehicles.

A problem is the number of ultrasonic filters needed. NTT placed the figure at “nearly 100” for a current smartphone and predicted an ever-expanding requirement as the IoT advances.

The conglomerate added the filters will also need to get smaller, setting the stage for its research into how best to direct signals while minimising backscatter.

NTT described a problem in directing ultrasound waves within narrow confines without creating backscatter when the signal is bent sharply.

The company worked with Okayama University to develop a “topological ultrasonic circuit” capable of propagating GHz ultrasonic waves with “reduced backscattering” by protecting the signals with specifically shaped periodic holes. The approach effectively takes the shape of the signal path (waveguide) out of the equation, while enabling production of an ultrasonic filter measuring in the hundreds of square micrometres rather than tens of thousands.

A simplified explanation of the approach of NTT and Okayama University is the waveguide comprises periodic holes tilted 5-degrees clockwise or anti-clockwise: “When an external ultrasonic wave is applied to the edge of this structure, valley pseudospins are generated in both regions, which rotate in opposite directions”, resulting in ultrasonic waves propagating “along the edge in one direction”.

Valley pseudospin-dependent transport “results in robust and stable travelling waves protected by topological order” NTT explained. The characteristic “solved the problem of backscattering in the folded small waveguide structure”.

The miniaturisation element of NTT’s work is a key area, potentially opening the door for IoT devices to handle a broad range of RF without impacting the size and battery-less operation the researchers at Pusan National University focused on.

And, while the focus here is on IoT, the Japanese research could prove important for future smartphones and wearables, boosting the amount of internal space available without compromising on the range of frequencies the devices can handle.

The South Korean research, meanwhile, offers a possible foundation for other consumer electronics, along with monitoring devices for the healthcare sector, smart infrastructure for urban settings and environmental sensing services, among others.