from Hacker News

5G researchers manage record connection speed

by ericthegoodking on 2/25/15, 4:11 PM with 32 comments

  • by vmarsy on 2/25/15, 10:31 PM

    "We need to bring end-to-end latency down to below one millisecond"

    The incredible bandwidth is surely great but they don't mention how to improve latency here, they forget to tell what was the latency of this 100m transfer from an antenna to a receiver.

    One of the big issues with today apps is latency, this is easily frustrating. It also prevents apps such as real time multiplayer games to be viable.

  • by simonebrunozzi on 2/25/15, 11:46 PM

    Latency, latency, latency. Everything else is pretty much irrelevant at this point.
  • by femto on 2/26/15, 3:25 AM

    It's a simulation [1]. The claim is that the technique is applicable to center frequencies below 6GHz and the 1Tbit/s used a bandwidth of 100MHz (for a spectral efficiency of 10^7 bit/s/Hz???). I'm guessing that whilst the technique might be applicable below 6Ghz, the 1TB/s rate isn't.

    There are fundamental limits on the information capacity of an antenna using the EM spectrum [2], based on the surface area of the volume of space it occupies, in units of wavelength. (Related to the Holographic Principle?) I haven't done the calculation, to see if the claimed rate is within this limit, but a spectral efficiency of 10^7 bit/s/Hz is about 5 orders of magnitude beyond what others have reported (less than 100bit/s/Hz [3]). It will be interesting to see the details!

    [1] http://www.ispreview.co.uk/index.php/2015/02/university-of-s...

    "UPDATE 25th Feb 2015

    We’ve been finding the 1Tbps claim a little difficult to digest and so have been prodding Professor Rahim Tafazolli for further details, specifically a greater clarification of how the performance was achieved.

    According to Tafazolli, the new class of Detector (a completely new approach) was tested through computer simulations (these simulated a real mobile/wireless environment) and were found to achieve the 1Tbps rate claimed. In our view that’s quite a bit different from conducting a practical test.

    Next year Tafazolli said that his team would work to implement this in a proper hardware/software platform and test it in a real environment in the 5GIC outdoor testbed. Hopefully they will be able to announce the performance in 2017."

    [2] http://arxiv.org/pdf/cs/0701055.pdf

    [3] http://ieeexplore.ieee.org/xpl/articleDetails.jsp?reload=tru...

  • by jblow on 2/25/15, 11:15 PM

    I don't even get 4G speeds almost ever. Bandwidth is always massively oversold. Often I can't even successfully load a random web page when I have 4+ bars of "4G".
  • by nikanj on 2/25/15, 11:09 PM

    I wonder if we'll still be stuck with our puny data caps when this arrives. Blowing through your monthly quota in a minute might be astounding from a technical point of view, but it only takes one broken page with an ajax infinite loop to get you.