"UMTS 3G barely got off the ground before HSDPA was launched, and HSUPA is on the horizon," one of the managers laments. "And are mobile WiMax and FLASH-OFDM 4G, or is 4G something else entirely?"
Maybe that's not real-world barstool chit-chat, but it does reflect the maze of terms and acronyms -- and confusion -- that is accompanying new and future generations of mobile broadband service.
"There is a great confusion about what all these acronyms mean and how the different services are packaged," said Anthony Ephremides, a professor of electrical and computer engineering at the University of Maryland.
This confusion reverberates through the hallways of businesses as IT and telecom managers struggle to understand newly available mobile technologies, those that are coming soon and how to pick the best one for their enterprises. The first step is understanding what the technology currently is -- and will be.
Track 1: 3G And 3.9G
Think of mobile broadband as a train station. On Track 1 is third-generation, or 3G, cellular data service, which operators have been deploying for the past couple of years. 3G service now covers most large and medium-size cities in the U.S., Europe and large swaths of the rest of the world, particularly in Asia.
Ephremides noted that the third-generation designation comes after 1G, which was analog voice service, and 2G, which was digital voice service. Early cellular data systems such as 1xRTT (1x Radio Transmission Technology) and EDGE (Enhanced Data Rates for GSM Evolution) often are called 2.5G.
Confusing matters is the fact that carriers with different technologies have deployed different flavors of 3G. For instance, carriers that have CDMA (Code Division Multiple Access) cellular networks, such as Verizon Wireless and Sprint in the U.S., have deployed the first generation of EV-DO (Evolution-Data Optimized) 3G service. UMTS (Universal Mobile Telecommunications System) and, more recently, HSDPA (High-Speed Downlink Packet Access), have been deployed by Cingular and other carriers worldwide that have GSM (Global System for Mobile Communications) networks. All these current forms of 3G provide typical download speeds of about 500 Kbit/sec. and typical upload speeds of less than 200 Kbit/sec.
In the short term -- say in the next year to 18 months -- many cellular operators will roll out enhanced versions of 3G that will increase speed, coverage and capacity, according to David Debrecht, Nokia Corp.'s director of technology for broadband wireless access. For instance, EV-DO Revision A and HSUPA (High-Speed Uplink Packet Access ) will increase typical download speeds to more than 1 Mbit/sec.
"Rev A is coming sooner than some people think," Peter Cannistra, a director in Sprint's Broadband Strategy Group, said in an interview. "We'll see typical speeds of 1 Mbit/sec., but upload speeds will be the biggest improvement over [first-generation] EV-DO -- it'll be two or three times as fast."
In the years after that next generation of 3G is deployed, cellular operators will roll out significantly improved technologies. The first of these will be what some people call 3.9G, or Long-Term Evolution (LTE), according to Debrecht. It will offer speeds several times that of current 3G services such as EV-DO, he said.
Current 3G is circuit-based technology, much like what cellular operators have been using for years. That is to say, it's a point-to-point technology in which each connection requires a dedicated circuit. However, LTE will be IP-based, Debrecht said, with the more flexible packet-based routing of such systems.
"It's not quite 4G, which is the system everybody's trying to get to in the long term," Debrecht said. But LTE is OFDMA-based, at least for the downlink, he said. OFDM (Orthogonal Frequency-Division Multiplexing) and its follow-on technology, OFDMA (Orthogonal Frequency Division Multiple Access), are the basis for newer types of wireless broadband, which are discussed in the next section.
LTE is in the middle of the standardization process now, Debrecht said. Optimistically, LTE could be standardized, and infrastructure equipment could be available in as soon as two to three years, he added.
Track 2: Wireless broadband
On Track 2 are so-called wireless mobile broadband technologies. The best known of these technologies is mobile WiMax (Worldwide Interoperability for Microwave Access), which could start appearing in a handful of markets in pre-certified form by the end of this year. A fixed version of WiMax is already available in many markets.
Mobile WiMax is best known because its most visible advocate, Intel Corp., has been waging a relentless marketing campaign for the technology over the past couple of years, just as it did for Wi-Fi. The technology scored a major victory recently when Sprint announced that it would use it for a new nationwide wireless network. A fixed, point-to-point version of WiMax already is available in many markets.
However, mobile WiMax is not the only type of mobile wireless broadband. Qualcomm is the champion of FLASH-OFDM, which it acquired when it bought Flarion Technologies last year. Qualcomm also owns an extensive portfolio of patents related to both OFDM and OFDMA. Another player is IPWireless Inc., which developed a type of wireless broadband called UMTS TDD. Both FLASH-OFDM and UMTS TDD already are available and have met with some success in markets around the world.
At this early stage, the modulation technology used for fixed WiMax is OFDM, but OFDMA is the improvement that everybody expects to start becoming commonplace soon. In fact, mobile WiMax is OFDMA-based, according to Debrecht.
"OFDMA systems provide more carrier signals and support more users in a more flexible way compared to OFDM, Debrecht said.
From a slightly more technical point of view, Ephremides noted that OFDM-based systems split the signal into different frequencies and simultaneously transmit the signal over a set of small bands spread over a wider band.
"You gain some hedging against impairments such as fading and jamming, which gives it some robustness over [3G systems], but there also are disadvantages," Ephremides said. For example, 3G systems do better in situations with uncontrolled numbers of users.
"If it happens that 200, 300, 400 users show up on the system, [OFDMA systems] will suffer more," he said. "[3G] is more resilient when traffic gets heavy."
Besides the fact that wireless broadband is IP-based, the biggest difference between it and 3G may be non-technical. That difference is that wireless broadband service will be offered by some cellular operators, but also by wireless ISPs.
For example, Sprint recently committed to building a nationwide mobile WiMax network. But so has a much smaller company, Clearwire, which is run by cellular-industry pioneer Craig McCaw. So far, Clearwire provides fixed WiMax service only in a handful of relatively small markets. However, Intel and Motorola recently invested almost a billion dollars to help Clearwire roll out a nationwide mobile WiMax network in the U.S.
What this new competition means in terms of pricing and service remains to be seen, of course. Some believe that voice over IP using wireless broadband can hurt the mobile carriers, while others don't think such applications will have much impact. At the very least, more competitors providing more types of service can only be good for users, the experts agree.
4G: The tracks start to converge
Ultimately, as Debrecht suggested, the cellular carriers will adopt 4G. The problem is that nobody is quite sure what 4G means, since there are no standards. That isn't deterring Sprint, which, in announcing its nationwide mobile WiMax network, called that service 4G. For that network, it will use spectrum in the 2.5-GHz range that it amassed in large measure through its merger with Nextel.
"We call 4G the technology we choose for the 2.5-GHZ spectrum," said Sprint's Cannistra. "That has certain characteristics that are distinct from our 3G network." Among those characteristics are what Cannistra called "a global ecosystem that will allow for an inexpensive chip set that will be part of laptops. It's a very Wi-Fi-like model."
By that he's referring to how Wi-Fi developed into a worldwide success. A wide variety of vendors, ranging from consumer gadget manufacturers to enterprise networking company Cisco Systems Inc., sell products that support that wireless local-area networking standard. Intel, in particular, was very aggressive -- and successful -- in marketing Wi-Fi as a brand and says it will apply the same effort to mobile WiMax.
Others are quick to point out, though, that mobile WiMax isn't 4G at all.
"You have to be careful when you talk to an operator when they call it 3G and 4G," Nokia's Debrecht said. "There will be a 4G in terms of what the standardization bodies determine, but that won't be the same as what the carriers may be calling 4G."
In other words, the two tracks -- 4G and wireless broadband -- likely will converge in the future in the form of OFDMA-based technology of some sort. However, beyond that, nobody can talk particulars. One big reason for that is that the 4G standard-setting process hasn't started in earnest.
"If people start throwing ideas around now, we could have a standardization body in 2009 or 2010, so it could be 10 years before we see true 4G technology -- 2011 or 2012 and probably even longer than that," Debrecht said.
What does the future hold?
New mobile technologies will provide fast, ubiquitous mobile access, and new competitors to the traditional cellular and fixed telecom operators will emerge. There is little disagreement about those two assertions.
"We feel that, in two years, anything you can do today, you'll be able to do on a mobile basis," Sprint's Cannistra said.
Beyond that, however, the long-term future of mobile technology is unclear. One area of murkiness is how the two technologies -- 3G and wireless broadband -- will co-exist. Many, including Cannistra, believe that the cellular data service now called 3G and its successors will be available primarily for cell phones, smart phones and other mobile devices. That service will be used for applications such as downloading media, video conferencing and other applications that don't focus on heavy use of enterprise data. Wireless broadband such as mobile WiMax will be used by laptops for more data-centric applications. .
"I think we'll see stratification by device," Cannistra said. "It also will be application-dependent."
Sprint's gamble is that enterprises and consumers will prefer to buy both types of mobile broadband service from a single operator. He acknowledges, though, that all manner of new service providers will emerge that have the potential to compete directly with cellular carriers such as Sprint.
"It's a megatrend that we are aware of here," Cannistra said.
Farther down the road, past 4G, Ephremides sees what he calls full-fledged multi-hop ad hoc networks, something the U.S. military is currently investigating.
"Currently, when your BlackBerry goes to a base station, everything goes in one hop," he said. "This [multi-hop ad hoc networking] would involve multiple hops to the base station -- you'd go through other users, so some users would be like passive repeaters bouncing other people's signals. Ad hoc networking means no infrastructure."
Current Wi-Fi mesh networks, such as those being installed by some municipalities to cover an entire metropolitan area, are crude versions of such networking, Ephremides said.
"But they still require very densely placed access points," he said. "[Ad hoc networking] would free you of that, so in a couple of decades, that older type of architecture will be passe." He acknowledged, however, that while the military has been pushing development of such networks, early versions don't work very well.
With all these changes occurring so quickly, it seems likely that there will be interim confusion not only for users, but also for providers.
"We'll make the decisions with the best information we have, but it won't be perfect information," Sprint's Cannistra said. The same, alas, will have to be said about both individual users and enterprises.