T-Mobile, NSN Team for Faster 3G

T-Mobile US Inc. and Nokia Networks are teaming up to back Long Term HSPA Evolution, a technology that is a new evolution in the HSPA standard. Both companies say this super-fast mobile network technology will be commercially deployed in 2013. (See T-Mob, NSN Speed Up HSPA.)

When the network update is complete, T-Mob will be delivering peak data rates of more than 650 Mbit/s, nearly reaching the 1Gbit/s peak promised by Long Term Evolution (LTE)-Advanced, and upping HSPA+ speeds from today's 42 Mbit/s.

Both companies say it is important to evolve HSPA networks while deploying LTE networks to accommodate mobile data growth. NSN, which counts 200 operators globally on its HSPA customer roster, demonstrated data rates over 100 Mbit/s at this year's Mobile World Congress. It says its Single Radio Access Network (RAN) platform is already equipped to handle the evolution path, which will be backwards compatible with earlier HSPA and relatively easy to install, not to mention more cost effective than deploying LTE.

The initial key proposal for the standard was approved during the 3rd Generation Partnership Project (3GPP) RAN plenary meeting held this week.

Why this matters
T-Mobile hasn't formally announced plans to deploy an LTE network, instead rolling out its FauxG HSPA+ network with speeds comparable to LTE, but its plans to wrest the most speed possible out if its HSPA network make sense.

Heavy Reading senior analyst Gabriel Brown says most operators will use both technologies -- LTE where there is new spectrum or in some existing GSM spectrum bands -- and, where there is HSPA, operators will develop it. (See 3G vs LTE: No Contest.)

What's not clear, he says, is what specific features will actually get implemented on evolved HSPA networks. An NSN spokeswoman says the company is targeting several features with the final details to be worked out in the standardization process. The features, outlined in a whitepaper available here, include an HSPDA multi-carrier evolution path, multiband frequency band aggregation, and HSPA and LTE carrier aggregation. Longer-term, it will also include enhanced MIMO (multiple input multiple output) functionality and SON features.

For more
For more on NSN's role in the evolution of HSPA and migration to LTE, please check out the following stories:

— Sarah Reedy, Senior Reporter, Light Reading Mobile

sarahthomas1011 12/5/2012 | 4:15:56 PM
re: T-Mobile, NSN Team for Faster 3G

Evolving HSPA has to be a lot cheaper for T-Mobile than deploying a new LTE networks. Seems like a good strategy.

Gabriel Brown 12/5/2012 | 4:15:54 PM
re: T-Mobile, NSN Team for Faster 3G

Good comment. Dual-carrier (w/ 64-QAM) is the preferred HSPA+ feature over the next couple of years. For now, operators tend to prefer 6-Sector to MIMO for HSPA capacity upgrades. 

Interested to hear if you think differently?

Also interesting to note some of the recent simulations for LTE performance. In loaded networks it's coming out at around 1 bit/s/Hz vs the 2 bit/s/Hz observed in commercial, but lightly loaded, networks today. Obviously, there are still technology innovations to come in LTE as well. 

wirelesse2e 12/5/2012 | 4:15:54 PM
re: T-Mobile, NSN Team for Faster 3G

Primary benefit of long term evolution is not the increase in peak throughput but growing the total cell capacity due to significant improvement at the cell edge. For example moving from category 14 (21 Mbps) to category 20 (42 Mbps) results in an improvement of over 25% in throughput for 95% of cell area. On the other hand, for the same change the improvement is limited to about 6% when it is measured in the center of the cell (about 5% of the cell area).

Likelihood of having a 650 Mbps evolved HSPA+ network in 2013 with a handheld terminal form factor is nil. Since this requires 40+40 MHz spectrum along with 4*4 MIMO capability, multiple terminal design constraints have to be tackled at once. Considering after more than three years, we still don't have any MIMO HSPA+ devices (category 21-24) working within a 28 Mbps HSPA+ network, I consider the likelihood of overcoming similar challenges at 4*4 in the next three years very low.

Nevertheless, it is time to recognize the long term potential of HSPA+ evolution and underline the room for growth in overall data capacity. I hope you'd publish more detailed articles on this topic to separate hype from reality. 

wirelesse2e 12/5/2012 | 4:15:50 PM
re: T-Mobile, NSN Team for Faster 3G

I agree that seems to be the trend. I recently looked at GSMA HSPA site for a list of 28 Mbps networks. There are three networks mentioned: Singapore M1, Italy TIM and Germany O2 (planned). In none of these operators websites there is a mention of 28 Mbps service or a device with HSDPA category 15 or 16 capability. Fastest device on M1 and TIM websites seems to be Huawei E1820 (21 Mbps, class 14 device).

I also looked at the same GSMA website for equipment that are category 16. List includes Qualcomm modules MDM8200A, MSM8260, QSD8672 as well as a number of OEMS: Samsung Broom, ZTE MF678, Novatel Ovation MC998D and a number of devices from Fly Card. Out of this entire list the only device with a reasonable spec publicly available is Ovation MC998D. Unfortunately Samsung, ZTE or Fly Card details are either non-xistent or sketchy.

Out of the three Qualcomm modules listed, only MDM8200A has solid references as the chipset used in commercial devices. Sierra Wireless modules MC8700, MC8704 and 8705 all use Qualcomm MDM8200A. Similarly Sierra Wireless USB sticks 306, 308 and 310A all use the same module. However, Sierra Wireless does not use the device as a cat. 15 or 16 (with MIMO); instead it restricts it to cat. 14 (21 Mbps).

Novatel Ovation MC998D is also another USB stick that uses Qualcomm MDM8200A. Unlike Sierra Wireless they mention the ability to have MIMO and cat. 16 but refer to some future firmware upgrade to enable it. Considering it is not being sold in any of those three operators, it is hard to believe they have much motivation to enable that capability.

Net/net, as you said, operators go the easier way of channel-bonding to reach cat 24 (42 Mbps). However, that is the end of the road since cat. 25 (46 Mbps) to cat. 28 (84 Mbps) al require MIMO. Before we start expecting 672 Mbps in 2013, we should first see when we will start having commercial grade HSPA+ MIMO devices and networks that support them.

Murat Bilgic


Sign In