Qualcomm made three announcements about connectivity products today at its 2016 4G/5G Summit in Hong Kong. The first announcement, aimed at easing the transition between the 4G LTE networks we have today and the 5G networks of the future, is a partnership with Netgear, Telstra, and Ericsson to deliver the first commercial gigabit class LTE device and network. Qualcomm sees gigabit LTE as a necessary building block and foundation for 5G, providing more bandwidth over a wider area until even higher bandwidth, shorter range 5G networks come online. Telstra, an Australian telecommunications provider, is currently building and testing the first gigabit LTE network, with equipment and software from Ericsson, in a bid to bring higher bandwidth internet access to the home without the expense of laying fiber cables.

Telstra is pushing for a commercial launch in the next few months and will be providing its customers with the new Netgear Mobile Router MR1100—a gigabit LTE endpoint and Wi-Fi router that delivers up to 1Gbps download speeds through a combination of 3x carrier aggregation, MIMO technology (4x4 MIMO on two aggregated carriers plus 2x2 MIMO on the third carrier), and 256-QAM modulation. Inside is a Qualcomm Wi-Fi solution and Qualcomm’s Snapdragon X16 LTE modem paired with Qualcomm’s WTR5975 RF transceiver.

This is the Snapdragon X16’s first appearance after it was unveiled back in February of this year at Qualcomm’s annual Analyst Day. The X16 is a sixth-generation discrete LTE modem manufactured on a 14nm FinFET process that’s capable of 1Gbps (Category 16) on the downlink and 150Mbps (Category 13) on the uplink. According to Qualcomm:

The Snapdragon X16 LTE modem is designed to reach Gigabit Class LTE speeds using the same amount of spectrum as Category 9 LTE devices. By using carrier aggregation and 4x4 MIMO, the Snapdragon X16 LTE modem can receive 10 unique streams of data using only three 20 MHz carriers. Its support for 256-QAM boosts the peak throughput of each stream from ~75 Mbps to ~100 Mbps, with additional gains possible with modem data compression. Additionally, with support for Licensed Assisted Access (LAA) as well as LTE-U, this combination reduces the amount of licensed spectrum required—to 40 MHz or less—to vastly expand the number of operators that can deploy Gigabit Class LTE speeds around the globe.

The X16 LTE modem will find a home integrated inside the next, currently unnamed, Snapdragon SoC, which is the second big announcement today. This next generation SoC will be powering flagship phones in 2017, enabling faster download speeds as carriers begin rolling out gigabit capable networks over the next 12 months, according to Qualcomm.

The final big announcement is Qualcomm’s first 5G modem, the Snapdragon X50. Designed to work only in the 28GHz millimeter-wave spectrum initially (supporting the Verizon 5GTF and KT 5G-SIG specifications), it’s capable of 8x100MHz carrier aggregation, giving it access to 800MHz of spectrum versus 80MHz (4x20MHz) for Qualcomm’s X16 LTE modem. All of this extra bandwidth enables up to 5Gbps on the downlink.

The downside to using higher frequencies, however, is their limited penetration and range. Qualcomm looks to mitigate these issues by using an array of many small antennas and employing adaptive beamforming and beam tracking to enable non-line-of-sight operation, essentially using surface reflections to see around corners and objects.

The Snapdragon X50 5G modem will be paired with two of the new SDR051 millimeter-wave transceivers and the PMX50 power management IC to support fixed wireless applications or future mobile devices. Because it only supports millimeter-wave frequencies, it needs to be paired with an additional 4G LTE modem to provide data uplink and voice capabilities and to allow seamless handoff between 4G and 5G networks depending on availability.

Qualcomm expects the Snapdragon X50 to begin sampling in the second half of 2017, with the first commercial products incorporating the new components showing up in the first half of 2018.

Comments Locked

22 Comments

View All Comments

  • sherifhanna - Tuesday, October 18, 2016 - link

    Sorry that you didn't find the info graphic helpful. Here's a video of our 28GHz mmWave prototype in action: https://youtu.be/IabL8uk8A-4

    And here's an in-depth white paper on how to make mmWave mobile, including results from real world tests in urban environments: https://www.qualcomm.com/documents/heavyreading-wh...

    Hope you'all find that information much more substantial.
  • iwodo - Friday, October 21, 2016 - link

    My only problem with directional mmWave is if there are any health concern passing through a human being. What energy level are they ?
  • StevoLincolnite - Tuesday, October 18, 2016 - link

    Telstra is a horrible provider, over-priced, garbage customer service.

    But you cannot dispute their network size and quality, they tend to adopt the latest technology's before most other companies.

    Sadly though, price is an important factor, so I stick to their main competitor, Optus.
  • scottjames_12 - Tuesday, October 18, 2016 - link

    Gigabit LTE on Telstra! Great, now their customers can blow through their expensive, tiny monthly data caps in mere seconds! Well, that's assuming they're not having another network outage...
  • MrSpadge - Tuesday, October 18, 2016 - link

    > The downside to using higher wavelengths, however, is their limited penetration and range.

    You meant higher frequency and shorter wavelength.
  • Matt Humrick - Tuesday, October 18, 2016 - link

    Thanks
  • name99 - Tuesday, October 18, 2016 - link

    256-QAM. Seriously?
    How far can you get from a cell tower before ambient noise drowns up such fine amplitude detail in the signal? (I'm not being snarky, I'm genuinely curious. In a normal city, can this function usefully even a block from the tower?)

    Or is the secret that you've aggressively lengthened the window over which error correction is performed, along with aggressively iterated soft decoding, so that while individual bits are frequently incorrect, the signal when tracked over, I don't know, (16? 32?) successive bits can be extracted? (With what density of excess FEC bits?)
  • MrSpadge - Tuesday, October 18, 2016 - link

    At 28 GHz the propagation length is so short that I don't think the classical cell phone tower concept works anymore. For indoor use over a few meters the requirements won't be as harsh.
  • sherifhanna - Tuesday, October 18, 2016 - link

    Yup, 256-QAM. Turns out it performs remarkably well in real networks. Here are results from Telstra's network in Australia, which has had 256-QAM since last year on their macro network:
    http://signalsresearch.com/issue/256-qam/

    Signals Research Group is the foremost authority on cellular technology testing in live networks.
  • OEMG - Wednesday, October 19, 2016 - link

    Yeah, 256-QAM is not a joke. I like to think of it as the upper limit of the technology. Sure, you need very good conditions if you want to pack that high number of bits. But when you consider that these same "very good conditions" are also required for the highest bit rates on 3G/2G networks before it shows that even though you may never have the opportunity to use it, you're still getting benefits on more common scenarios.

Log in

Don't have an account? Sign up now