2G GSM & GPRS — The Shutdown Begins
I don’t know if you noticed it.
A few days ago, AT&T announced that it has started refarming (i.e, shutting down) its 2G GSM cellular services at 1900 MHz spectrum in the New York City area, for 3G and 4G services.
M2M Applications using 2G GSM and GPRS—such as Business and Residential Alarm units in large metropolitan markets—are going to be the first M2M services to be hit by this soon.
This event should not surprise anyone.
Regardless, as mentioned in my last post, this is also the start of an activity that has enormous cost implications for the M2M industry in the next year or two.
The GPRS Shutdown Was Inevitable
More than six years ago, when Aeris’ customers were reviewing options for new cellular technologies before the AMPS Sunset in 2008, I believed that a GSM (and, hence, GPRS & EDGE) shutdown was inevitable.
In presentations to our Customers, I stated the only uncertainty was the actual date when GSM and GPRS services would be unavailable in any given major market.
The underlying encoding protocol of GSM is Time Division Multiple Access (“TDMA”). The spectrum efficiency (or how much bandwidth is used to transmit data) of TDMA simply is not as high as Code Division Multiple Access (“CDMA”).
Thus, GSM Carriers (like AT&T and T-Mobile) had to transition their 2G customers to more spectrum efficient technologies at some point. This would allow them to compete more effectively with CDMA Carriers (like Verizon, Sprint, US Cellular and Alltel).
To help our customers, I recommended that our AMPS Customers switch to CDMA to avoid another sunset too quickly—i.e., the GPRS Sunset that could occur within the product life-cycle of many M2M Applications. Needless to say, those who switched to CDMA are very happy to have made that decision.
Lower GSM Radio Costs
However, at the time, the cost of GSM radios was too low to be ignored by a large portion of the overall M2M market—particularly by cost-conscious M2M Applications that needed low-cost radios.
This lower cost of GSM radios was due, in part, to:
- Less complex cellular chipsets used in GSM than in CDMA (smaller chips cost less).
- Large numbers of GSM handsets sold throughout the world (higher volume provides scaling cost reductions).
For example, the following information from GSM World in early 2009 shows that the deployed number of GSM units was greater than the other technologies combined:
|CDMA2000 1xEV-DO Rev. A||
While it is not clear how many of those GSM units were GPRS-capable, choosing to deploy 2G GPRS data units was an easy choice for some M2M Customers. This substantially reduced the radio costs for their M2M Applications.
The T-Mobile Acquisition Failure
Last year, it became clear that AT&T did not have enough spectrum in many markets for their 3G and 4G network deployments. Over the previous few years, the introduction of smart-phones using 3G data had skyrocketed. As a result, their 3G networks have not kept up with increasing demand and load. AT&T attempted to acquire T-Mobile to obtain additional spectrum, but was unsuccessful.
During the discussion on the acquisition effort, AT&T executives made very clear statements regarding the impacts if the acquisition were denied. In her blog, Joan March, AT&T’s VP of Federal Regulatory stated what would happen:
First, AT&T would promptly shut down its 2G GSM network—a network that currently supports tens of millions of devices, including handsets for our pre-paid products that are particularly important to fixed and low income customers. As a result, all those handsets would go dark and that customer base would be required to go purchase new mobile broadband (UMTS) handsets, which are generally more expensive.
In this news report, Kris Rinne, AT&T’s SVP of architecture and planning, underscored this same need at a GSM Association event. As the report put it:
Rinne’s logic was not completely flawed. As AT&T fills the 18 MHz it has set aside for 4G, it could fall back to its AWS spectrum. And then, if that band were to get full, it can leverage its PCS spectrum. But right now, AT&T is using those bands for its 2G and 3G networks. It will have to transition those customers to 4G before those airwaves could be reused, which can be a painful process. Rinne said, “We will have the opportunity to re-utilize this spectrum in the future.”
About two months ago, it was reported that AT&T began notifying its 2G GSM customers in New York City that their phones would need to be changed to devices operating on their more advanced networks.
This was a clear indicator of what was about to occur shortly.
The GPRS Sunset
And now, the moment is here. The GPRS Sunset has begun.
Without fanfare, and most importantly, without any FCC mandate guaranteeing 2G GSM service continuity through a date that is certain (or one that is far enough in the future to make it palatable). There will not be a five year 2G GSM sunset, similar to what was available for AMPS about a decade ago.
As the spectrum crisis worsens for AT&T, with the continued sale and deployment of high-usage data devices, such as smartphones and tablets, they must continue to refarm their 1900 MHz 2G GSM networks across the country.
Eventually, they will also need to refarm their 850 MHz spectrum. When this happens, this action will effectively shut down all their 2G services.
And, M2M Application Devices operating on AT&T in that market will go dark.
Can Customers Move Service to T-Mobile?
T-Mobile has stated that they have no current plans to shut 2G GSM.
Thus, moving the 2G GPRS units to operate on T-Mobile seems like an option. However, this requires replacing the SIM in the Device, since it is unlikely that AT&T would allow expensive roaming operation for low-ARPU M2M Devices to operate on a competing Carrier using an AT&T SIM.
Sending field technicians to “touch” the units to replace the SIM is an expensive action, if End-user customers cannot be expected to do the work.
Physical SIM or Device Swaps
In some cases, the SIM may be sealed inside Devices, or the Devices may be physically installed in relatively inaccessible places. This would require physical Device swaps instead of simpler SIM swaps.
T-Mobile and Other Coverage
Since T-Mobile does not have the coverage footprint of AT&T, changing service to T-mobile is not a solution for those units that are in markets not covered by T-Mobile.
Selecting another local GSM Carrier may not be an option, since the number of local contracts needed to maintain a nationwide M2M deployment would be impractical (and it would still require many SIM swaps!).
In other cases, the local GSM Carriers may not be able to provide GPRS data services at a price that would work for the Carrier or the M2M Customer.
In such markets, the 2G M2M Devices would necessarily cease to operate.
T-Mobile Futures for 2G Service
Furthermore, moving M2M service to T-Mobile is temporary at best.
T-Mobile has been focused on deploying HSPA+ in their available spectrum, and is planning to deploy LTE in the additional spectrum received from AT&T as a result of the failed acquisition attempt.
But, after this work is complete, they must also begin converting their network from 2G to 4G to remain competitive with the other Carriers.
Thus, within the next few years, T-Mobile will also need to start the process of stopping 2G services to re-use their spectrum for 4G services.
CDMA Remains the Answer
As I have often stated in the past five or six years, and our customers are more than happy to have done so, M2M Applications must use CDMA for service longevity. Until 4G LTE radio prices drop low enough to make it a viable option, this is the best approach.
We are now at a critical juncture for Customers on 2G GSM GPRS service—there is simply no time left for delaying any further.
Although this GPRS shutdown over the next year or two will be gradual for the Carriers (and it will not happen everywhere simultaneously), it will still occur at a speed that may be far too rapid for M2M Customers.
Please see my recent posts on spectrum (particularly from last week), to understand why it is important to begin deploying CDMA units as soon as possible … to avoid the Death of a Thousand GPRS Cuts.
What are your Comments?
Agreements? Disagreements? Please post your comments. Thanks!