Scientists at University College London recently achieved a speed on a wireless link of 938 Gbps. That’s over 4,000 times faster than the current average speed being delivered by T-Mobile, the current fastest cellular provider in the U.S.
The team is researching techniques for multiplexing multiple radio transmissions into a coherent transmission. The scientists achieved the speeds by utilizing a huge span of spectrum between 5GHz and 150 GHz. They also had to combine multiple techniques to create and join the signals.
- The signals from 5-75 GHz were generated using traditional, but high-quality radios that used digital-to-analog converters.
- The signals from the higher frequencies, the W-band from 75-110 GHz and the D-band from 110-150 GHz were generated by mixing optically modulated signals that used frequency-locked lasers and high-speed photodiodes. By frequency-locking the lasers, the scientists were able to create a stable carrier frequency that avoided the signal noise that would have been generated by normal free-running lasers.
- The team then used orthogonal Frequency-Division Multiplexing (OFDM)and bit loading to goose the signal up to 938 Gbps over the air.
Another team of researchers recently achieved fast wireless speeds on a single channel. The team is a consortium of researchers from the Japanese firms DOCOMO, NIT Corporation, NEC Corporation, and Fujitsu. The team of companies created a wireless device that uses 100 GHz spectrum indoors or 300 GHz spectrum outdoors to create a 100 Gbps link that can transmit for about 100 meters. The companies see this first device as the prototype for developing future wireless radios that can deliver speeds only possible today on fiber.
In the past, several research teams in laboratories have created terabit speeds for a link of several feet. There is a lot of literature speculating that radios in space could reliably achieve terabit speeds between satellites without the interference created by air.
All of the research teams are pushing the cutting edge for wireless technologies with the goal of someday creating much faster wireless technology. The worldwide push to master the use of the terahertz frequencies between 100 GHz and 1 THz has been labeled as 6G, although wireless vendors have already absconded that label to describe radios that use millimeter wave spectrum. The terahertz frequencies lie between traditional radio and infrared light.
The first 6G summit met in 2022 in Levi, Lapland, Finland, sponsored by the University of Oulu, and included major wireless vendors like Nokia, Huawei, Ericsson, Samsung, and NTT, along with researchers from numerous universities as well as groups like Bell Labs. At the summit, researchers talked about creating a set of standards for terahertz frequencies by 2030. They identified the first hurdle as the development of chips that can handle faster speeds. There were also questions about whether governments would try to regulate the higher frequencies.
For now, these fast tests represent scientists pushing the edge of radio technology. These tests are not going to produce any usable technology for many years. The University of College London used a wide swath of spectrum that would never by allowed by any government. But the early success of these various tests show that faster radios will somebody be possible.
sensational piece.
It’s quite easy to get 10bits per hz on a modern radio. Given many gigaherts of band (5-150Ghz, doesn’t say how much was actually used), a wifi7 radio setup with MLO could do this. Access to the spectrum is the limit and CPU/bus speed to handle 1Tbps. We’re talking a couple Gbps on a 320Mhz WiFi7 channel today, so give it 50x more spectrum…
extrapolating 10Ghz channels on WiFi7 is about 100Gbps on ‘easy’ modulations. Technically it can get as high as 60bits per hz so I’m being very conservative here.
modulations do tend to be lesser on higher frequency radios, wifi7 has a peak of 4096QAM while most mmWave radios tend to top out at 64QAM. Lasers often in similar range to mmwave.
Throwing huge channels at this likely isn’t a realistic solution, especially in sub 10Ghz where propogation is ‘good’ ie re-use and noise is bad. Massive channels in upper bands where propogation is poor, meaning very little noise and very little interference from other radios, sounds great but it’s also increasingly short range and line of site.
back to the point, while it’s an interesting benchmark, 145Ghz of channel isn’t remotely interesting for practical use.