Back to the discussion on cellular technology after the digression to slides presented at the Telit Developers Conference. I will continue on the factors that affect the choices necessary for M2M applications. As a quick recap, Part 1 of this thread covered the differences between traditional cellular handset needs compared to M2M needs:
M2M Devices Are Different.
M2M Applications Are Different.
Network Coverage Needs Are Not The Same.
This week, I will cover the next major item for consideration when selecting the technology: Cellular Technology Longevity. This is one of the most confusing issues that people encounter when deploying M2M applications using wireless cellular for the transport of M2M data. To make sense of this issue, a simplified history of cellular technologiesas related to M2Mis necessary.
Early M2M Data Transmission Experiences
In the early days of using cellular for M2M data transmissions in North America, the option was simple: everyone used the Analog Advanced Mobile Phone System (AMPS) on the 800MHz A and B bands for their transmissions. All Carriers in North Americalarge and smallprovided the same AMPS cellular technology service. So, deploying M2M applications meant a single Device, with a single AMPS radio module, with a single transmission technology was practical. In addition, an AMPS radio could operate on both the A or the B bands in any market, since the technology was identically deployed by Carriers on both bands in the market. Thus, roaming (i.e., the ability to use any available local Carrierbasically, one of two in any given market) was relatively easy to achieve, once the business arrangements were worked out and the network and control connections for roaming (for example, as designed by pioneers like Elmer Lyle and others at GTE Mobilnet) were set up. Indeed, once the Carrier handset subsidy was paid (i.e., when the multiple year contract ended), the customer was free to choose to use the handset on any other carrier offering AMPS service.
AMPS Data Transmission Methods
The transmission protocols and methods in AMPS were generally straightforward:
People used standard modemstypically Bell 212A compatible modems at 1200 baud or Bell 103 at 300 baud. Higher speed modems were generally not used due to modulation issues created by the Rayleigh fading effect when a radio is in motion.
Or patented data transportssuch as the Aeris MicroBurst technology that allowed 25 digits for use in Home and Residential security and Vehicle Location applications, for example.
A later entry for digital data on AMPS, known as Cellular Digital Packet Data (CDPD) was, unfortunately, a bit too late and too expensive (radios and service) to make much of an impact for M2M applications.
Most importantly, though, AMPS enjoyed a long life-cycle! First deployed in the USA in the early to mid 1980s, it survived through February of 2008.
Introduction of Digital TDMA and CDMA
In the mid to late 1990s, new digital cellular technologies were introduced into service. In North America, two technologies became common: ANSI-136 Time Division Multiple Access (TDMA) and ANSI-95 Code Division Multiple Access (CDMA). Here, I use the ANSI standard numbers to identify the specific variant of the technology and avoid confusion. Describing the actual protocol and detailed transmissions methods of these technologies is not important here, but there are some important points to be noted.
Enhanced Spectrum Efficiency
The most important reason for the change was the improved spectrum efficiency. Carriers could support more phone calls in the wireless spectrum or bands that they were allocated. For TDMA, the improvement was a factor of 3 times more call capacity and for CDMA, the improvement was a factor of 6 to 20 times (depending on the tolerance for noise) more call capacity, when compared to AMPS. In the early days, however, digital cellular was not heavily used by M2M Applications since the cost of the digital radio modules was substantially higher than AMPS-only radios. This cost precluded many companies from deploying digital radios into serviceparticularly in large-scale M2M deployments. However, the trend and direction was clear.
The Start of Technology Incompatibility
The introduction of TDMA and CDMA also saw the beginning of a fundamental service incompatibility between Carriers. While AMPS could be used with any Carrier, TDMA mode could only be used on TDMA Carriers, and CDMA mode could only be used on CDMA Carriers. Along with mergers and consolidations, this meant that TDMA handsets could typically only be used in digital mode on the 850MHz A band Carriers: such as AT&T and others (in the USA), Rogers and others (in Canada) and Telcel (in Mexico). Similarly, CDMA handsets could typically only be used on the 850MHz B band Carriers: such as Verizon, Alltel and others (in the USA), Bell Mobilite, Telus and others (in Canada) and IUSACEL (in Mexico).
Roaming Was Dependent on AMPS
Since the digital roll-outs in all the markets took a number of years, and large mergers and acquisitions were not very common yet, the roaming agreements and ability to use a given cellular handset anywhere, became completely dependent on the common availability of AMPS service. Thus, cellular handsets used Dual-Mode radiosi.e., the radios were either CDMA/AMPS dual-mode or TDMA/AMPS dual-mode. This allowed these digital radios and handsets to operate on the other digital network in AMPS mode and somewhat fore-stalled any potential concern of the creation of cellular monopolies unfortunately.
Digital Service Expanded and Changed
Introduction of CDMA-Only Handsets
In the early part of this century, as the CDMA Carriers completed their full digital footprints, they introducedprior to the demise of AMPSdigital-only CDMA handsets that could only operate in CDMA mode. This meant that handsets designed for CDMA-only mode could not be operated in TDMA markets at all. The incompatibility problem increased for most handset customers, but the M2M applications that began to use early CDMA radio modules were still able to operate in AMPS mode in some markets.
Demise of ANSI-136 TDMA
(I emphasize the use of the phrase ANSI-136 TDMA herealthough GSM is also a Time Domain Multiple Access technology, it was not compatible with ANSI-136 TDMA deployments.)Around the same time, AT&T chose to abandon ANSI-136 TDMA for a more commonly used, world-wide standard: the Global System for Mobile Communications (GSM) cellular technology. This was an understandable step, since the large number of GSM deployments world-wide meant that cost reductions due to scale could be achieved more easily than with ANSI-136 TDMA. Of course, since GSM was not compatible with AMPS in any way, GSM radios could not (and cannot) operate in CDMA markets, and vice-versa. Dual-mode GSM-AMPS radios and handsets were never developed or deployed. Thus, this demise of ANSI-136 TDMA became the next step in the direction of technology incompatibility. Customers had to choose between GSM or CDMA service and handsets, and roaming between these two technologies was impossible.
Introduction of PCS Bands
In the USA, Carriers began deploying service at 1900MHz, called the Personal Communications Service (PCS) bands. In PCS, there were no particular restrictions on what cellular technologies to deploy. Thus, some Carriers deployed the new ANSI-2000 CDMA technology, and others deployed GSM. Since there were 6 available bands (A, B, C, D, E and F) in the 1900MHz spectrum, the incompatibility concerns became a bit confusing. For example, many people assumed that since both Sprint and Verizon were deploying ANSI-2000 CDMA, any ANSI-2000 CDMA handset would be usable on either Carrier relatively easily. Unfortunately, business practices made that somewhat difficult: often, the very same handset sold by a manufacturer for Verizon could not be used on Sprint, and vice-versa, due to settings inside the handsets. For example, the Samsung SCH-i730 had two essentially identical versions: one for Verizon and one for Sprint.
ANSI-2000 CDMA Variation Increased
When Sprint deployed their service in the 1900 PCS bands, they did not need to deploy AMPS service, of course. Their digital-only deployments were ANSI-2000 CDMA, but there were other service difference that evolved. For example, remote handset provisioning and programming in CDMA was (and still is) accomplished differently: Verizon uses the Over The Air Service Provisioning (OTASP) and Over The Air Parameter Administration (OTAPA) method defined by the 3rd Generation Partnership Project 2 (3GPP2), and Sprint uses the Open Mobile Alliance Device Management (OMA-DM) method from the Open Mobile Alliance (OMA). This difference also leads to additional incompatibility: M2M module manufacturers do not want to incorporate the cost (albeit small) of OMA-DM support for units intended for Verizon service for example, or the cost (also small) of OTASP/OTAPA support for units intended for Sprint service for example. In M2M applications, this meant that migrating units from one Carrier service to another became more difficult, albeit not impossible. Most Customers also use tethered-programming at the factory, where the radio is programmed using a wired serial interface.
The AMPS Sunset Requirement
In late 2002, it became evident that digital cellular was going to be available everywhereeither CDMA or GSM. Thus, the Carriers asked the FCC to allow them to remove the comparatively inefficient AMPS services they offered. Particularly, since the CDMA and GSM technologies were fundamentally incompatible, and AMPS roaming was no longer possible, since GSM handsets did not have AMPS to begin with, and CDMA handsets had had AMPS mode removed pre-emptively. However, since a large number of modules in many longer-lived M2M (and other) deployments were AMPS-only, the FCC set a five year AMPS sunset requirement for the Carriers. All CDMA and TDMA operators in the 800MHz A and B bands were required to keep AMP service operational through February of 2008, and Customers were required to migrate their units to digital service within that period. Although there were some attempts by Customers to extend the sunset, particularly as the date neared, these did not succeed. CDMA and TDMA Carriers maintained their AMPS networkwith some relatively minor degradation in service qualitythrough February 2008. At that time, AMPS devices went off the air ... including many Aeris M2M Customer units that had been operating without any problems since 1997 and 1998 on! As mentioned in the last parts of this thread, expectations of M2M device operation longevity is one of the key reasons that people are concerned about future technology changes.
To Be Continued ...
Next week, I will cover more of the cellular technology history that impacts the technology selection process for M2M applications.