- High frequency signals collapse when walls or people block their path
- The neural networks have learned the bending of the beam by simulating countless basketball strokes
- Metasurfas integrated into signals in the form of transmitters with extreme precision
For years, researchers have had difficulties with certain vulnerabilities in ultra-high frequency communications.
Ultra-important frequencies are so fragile that signals that promise a huge bandwidth can collapse when faced with even modest obstacles, because walls, libraries or simply the displacement of people can stop advanced transmissions.
However, a new approach to Princeton engineers suggests that these obstacles may not be permanent roadblocks, although the leap from experience to the deployment of the real world remains uncertain.
From physics to adaptive transmissions
The idea of folding signals to avoid obstacles is not new. Engineers have been working for a long time with “airy beams”, which can bend in controlled way, but applying them to wireless data has been hampered by practical limits.
Haoze Chen, one of the researchers, says that most of the previous work focused on the demonstration of the beams could exist, and not to make them usable in unpredictable environments.
The problem is that each curve depends on countless variables, leaving no simple way to scan or calculate the ideal path.
To make the beams useful, the researchers borrowed an analogy from sports. Instead of calculating each shot, basketball players learn through a repeated practice which works in different contexts.
Chen explained that Princeton’s team was aimed at a similar process, replacing test and error athletes with a network of neurons designed to adapt its answers.
Rather than physically transmitting bundles for each possible obstacle, Atsutse Kludze doctoral students built a simulator which allowed the system to train virtually.
This approach has considerably reduced training time while anchoring models in air beam physics.
Once formed, the system was able to adapt extremely quickly, using a metasurface specially designed to shape transmissions.
Unlike reflectors, which depend on external structures, the metasurface can be integrated directly into the transmitter, which allowed the beams to bend around sudden obstructions, maintaining connectivity without requiring a clear line of aim.
The team has shown that the neural network could select the most effective beam path in the congested and changing scenarios, which conventional methods cannot achieve.
He also claims that this is a step towards exploitation of the Sub-Terahertz band, part of the spectrum that could support up to ten times more data than today’s systems.
The main investigator Yasaman Ghasempour argued that the fight against obstacles is essential before such a bandwidth can be used to require applications such as immersive virtual reality or fully autonomous transport.
“This work tackles a long -standing problem that has prevented the adoption of these high frequencies in dynamic wireless communications to date,” said Ghasempour.
However, challenges remain. Translation of laboratory demonstrations into commercial devices requires equipment scaling, refining training methods and proving that adaptive beams can manage the complexity of the real world at high speed.
The promise of wireless links approaching the flow of the Terabit class can be visible, but the path around obstacles, both physical and technological, is always wound.
Via techxplore
