HSPA Downlink From 1 to 80 MBit/s and Beyond

With the addition of ever more features into HSPA by 3GPP, peak data rates keep rising and one can find many different peak data rate numbers in the media. To bring some order into this, I've decided to put together a table with shows which features bring which rough peak data rates:

  •   3.6 MBit/s : Baseline HSPA with 16QAM modulation
  •   7.2 MBit/s : 16 QAM, more simultaneous channels)
  • 14.4 MBit/s : 16 QAM, even more simultaneous channels
  • 21 MBit/s    : 64 QAM modulation
  • 28 MBit/s    : MIMO (Multiple Input Multiple Output = 2 antennas) + 16 QAM, (3GPP Rel 7)
  • 42 MBit/s    : MIMO + 64QAM (3GPP Release 8)
  • 42 MBit/s    : 64QAM + Dual Carrier (3GPP Rel 8)
  • 82 MBit/s    : MIMO + 64QAM + Dual Carrier (i.e. 2 x 5MHz) (3GPP Release 9)
  • + more in case 3GPP decides to increase the number of carriers that can be bundled in Rel 9 or beyond.

In the near future, operators that can upgrade their base station to 64QAM modulation without a hardware replacement are probably tempted to do so. If they were smart with the backhaul in previous upgrades, they might already have the capacity at the base station to support the added traffic. All further steps require new hardware at the base station so operators will probably think a bit about it before actually deploying it.

Important: These are peak rates, i.e. they are only achieved under very favourable coverage circumstances and microcell environments. In most macro radio environments, speeds are much lower due to interference and less then optimal signal strength.

5 thoughts on “HSPA Downlink From 1 to 80 MBit/s and Beyond”

  1. Nice table. They can be seen as “target data rates”. There will be still a lot of work to do to achieve these rates… In my daily experiences i still see many bottlenecks on the backhaul side more than on the radio side…

  2. The “new hardware” that you mention for the further steps is actually not that much.

    For MIMO, on virtually every outdoor site, you only need an extra amplifier. This is because the existing antenna systems already have two branches for RX diversity, that can both be used for MIMO transmission without changes.

    Multi-carrier functionality is just SW that runs on a normal 2-carrier base station.

    So, I would say that the road to 84 Mbps is an easy-ride for most operators….
    … when they have the transmission in place (as you wrote) of course 😉

  3. Hi Jan,

    Thanks for the comment!

    Yes, you are right, from an antenna point of view the extra amplifier probably does the trick by extending the usage for the double antennas for RX diversity feature used today.

    I guess most network operators don’t have a lot of pain to upgrade to the 7.2 MBit/s or 14.4 MBit/s version of HSPA as that doesn’t require anything extra but software. Beyond that, some of the UMTS base stations deployed today might not have enough processing power to keep up with the load. And as you go to MIMO that’s certainly not getting any better. Also, some of the deployed RF modules might not be capable of doing 64QAM (speculation on my part).

    And on top, the many UMTS base stations in the field today can only go up to 8 E-1’s and don’t yet have an Ethernet interface. So for the extra speeds, a new backhaul interface is needed, 16 MBit/s for a 3 sector cell with 64QAM and 2 carriers per sector won’t do the trick. Even for the 7.2 MBit/s variant the limit the backhaul and not the radio access 🙂


  4. thank you for this good information, but is there any paper or article that show the simulation of data rates increased from 3.6 Mbps to 82 Mbps (16QAM only, then 16QAM + Dual Carrier, then 16QAM + Dual Carrier + MIMO,…)
    please if there is any, can you send it to me

Comments are closed.