from the bright-antennae-bristle-with-the-energy dept.
NIST Finds Wireless Performance Consistent Across 5G Millimeter-Wave Bands:
Wireless systems are moving to the mmWave spectrum at 10-100 gigahertz (GHz), above crowded cellular frequencies as well as early 5G systems around 3 GHz. System operators tend to prefer lower bands of the new mmWave spectrum. [...]
NIST (National Institute of Standards and Technology) researchers developed a new method to measure frequency effects, using the 26.5-40 GHz band as a target example. After extensive study in the laboratory and two real-world environments, NIST results confirmed that the main signal path — over a clear "line of sight" between transmitter and receiver — does not vary by frequency, a generally accepted thesis for traditional wireless systems but until now not proven for the mmWave spectrum.
The team also found that signal losses in secondary paths — where transmissions are reflected, bent or diffused into clusters of reflections — can vary somewhat by frequency, depending on the type of path. Reflective paths, which are the second strongest and critical for maintaining connectivity, lost only a little signal strength at higher frequencies. The weaker bent and diffuse paths lost a bit more. Until now, the effects of frequency on this so-called multipath were unknown.
"This work may serve to demyth many misconceptions about propagation about higher frequencies in 5G and 6G," NIST electrical engineer Camillo Gentile said. "In short, while performance will be worse at higher frequencies, the drop in performance is incremental. So we do expect the deployment at 5G and eventually at 6G to be successful."
The Friis transmission equation says that for fixed effective antenna area, direct line-of-sight signal detection is independent of the frequency. However, non-line-of-sight (nLOS) signals reflect off of materials and the amount they reflect decreases with frequency, so there have been concerns about pushing 5G and beyond to very high frequencies. This work confirmed the Friis equation at these frequencies and showed that nLOS signal loss isn't that big of a deal.
Damla Guven, et al., Methodology for Measuring the Frequency Dependence of Multipath Channels Across the Millimeter-Wave Spectrum [open], IEEE Open Journal of Antennas and Propagation, 3, 2022
(Score: 0) by Anonymous Coward on Thursday May 12 2022, @09:06AM (2 children)
No big surprise if you look at the Attenuation versus frequency plots for the mmWave band(s). The peaks are at 22, 60, and 180 GHz, which this 26.5-40 GHz result avoids.
Of course, the hardware needed to send and receive will have frequency-dependent factors that'll be a hhard problem to span the octaves.
(Score: 0) by Anonymous Coward on Thursday May 12 2022, @08:39PM
The issue apparently was not the direct line of sight stuff. There was a lot of concern about the multipath signals, and the concern was the higher frequency signals attenuated too much when they reflected off of the normal stuff they'd encounter in the wild. And if you read the paper intro, apparently there were multiple attempts at measuring this that showed inconsistent results, so there was an expectation that the lower frequencies would work better and everyone wanted to be down in the 3GHz region. These guys were able to separate out the direct from multipath contributions and show that it isn't a big deal and companies shouldn't worry too much about staking out higher frequencies in the 5G/6G spectrum auctions.
(Score: 0) by Anonymous Coward on Thursday May 12 2022, @11:49PM
Fear not, this is where the 5G nanobots they put in DannyB's COVID vaccine will come into play.