This is new territory for wireless network operators. Mobile data isn't the new issue; after all, from the network's point of view, voice is digital data, although delivered via a circuit-switched infrastructure. What is new is the fact that growth is no longer driven by the need to provide better voice quality. Growth is about providing new services. To better understand this, consider the brief history of mobile wireless communications.

Pre-cellular trunked radio systems date back to the 1920's and were successful in providing specialized wireless communication services such as emergency dispatch. To commercialize these services, the logical step was to pick high population areas and set up multiple towers. Different "cells" or towers were linked via proprietary X.25 packet networks to allow local roaming. Adoption of SS7 as the basis for a packet network control plane including home (HLR) and roaming (VLR) databases allowed the major carriers to achieve national footprints and establish manageable roaming agreements. Thus, "1G" was born.

As this model was still being implemented, the move was underway to replace the analog air interfaces with digital technologies based on time division (TDMA) or code division (CDMA) multiple access. This gave the end-user, who was accustomed to the digital voice quality of the fixed line network and its T-Carrier system, the same quality of voice service, providing they had decent RF signal strength. The mobile phone was now able to serve as a both a complement to, and a substitute for, the fixed line phone. This defines the second-generation or "2G" mobile wireless world we live in today: national and even global roaming capabilities with digital quality voice.

Competition and the need for differentiation are driving the industry to move us out of this relatively stable environment and into the brave new third-generation (3G) world of mobile multimedia and Internet access. 1G was a commercial venture intended to grow mobile communications business - it expanded the network. 2G was a "rescue mission" to improve the quality of the network experience - it optimized the network access. 3G is a commercial venture intended to grow mobile communication business. 4G will again be a rescue mission to improve the access to the new network.

What will the killer application or applications be that justify the purchase of a new handset and will pay for the new network infrastructure being deployed? There are a number of viable candidates, from the more mundane (email, Web access) to the more exciting (gaming, instant photos). Regardless, the wireless mobile industry has always provided one essential "Killer App": convenience.

That mobile wireless communication provides a highly valued convenience is made only more evident by the inconveniences we are willing to endure for its sake. Subscription rates continued to grow despite "can you hear me now?" becoming a daily utterance. The critical question is how much inconvenience will users tolerate in order to enjoy the convenience of mobile access to data?

To approach this answer quantitatively, we can look at the customer turnover rate, or "churn," in the mobile voice market. Of the six largest carriers in the U.S., the yearly churn ranges from 28 percent to 46 percent, according to The Yankee Group. Of those who choose to switch, approximately 18.5 percent cite service quality as the main reason according to the Federal Communication Commission. In absolute terms, this translates to nearly 40 million subscribers switching per year. And this is not a trivial switch in most cases, as it typically requires a new mobile handset and a new phone number.

So it's clear the quality of service is crucial, but will mobile data exasperate the issue? The telephone, providing reliable voice service for decades, is an example of an invisible technology: one that we do not even think about and that we expect to be working with no effort on our part. If someone picks up the receiver of a fixed phone and does not hear a dial tone, the common reaction is to put it down and pick it up again. It has been difficult for telephone users to adjust to a phone service, namely mobile service, which does not work this way. The computer, on the other hand, is a visible technology to the user. We plan daily for some type of failure in our terminal or the network. So we backup files, install virus protection, and include armies of IT support personnel on companies' payrolls. Users of mobile data therefore might not be as shocked and disappointed when they cannot immediately access the Internet or when experiencing slow downloads.

So people expect more from their voice services than data connections. But mobile voice has an arguably higher convenience quotient than does mobile data. You may really need to make that call to a colleague for a ride or to 911 for an emergency, but you may simply want to get your emails while waiting at the airport. To be commercially successful, mobile data networks will need to match the quality level of fixed data network. And as with fixed networks, the place where quality is crucial is at the access points to the network.

Expect future wireless networks to rely on multiple radio access technologies. Bandwidth, location and mobility will determine the type of access needed by the consumer, who will likely use more than one communications device. Complexity and intelligence will rise at the edges of the network. Source: Tektronix

There are two primary paths to 3G wireless which have survived - the path that begins with Global Systems for Mobile Communications (GSM) network technology and the one based on CDMA technology. In many ways, they are both quite similar. Both migrate to a topology that includes circuit switched voice service working in parallel with packet-switched data service. Both deploy spread-spectrum code division multiple access radio interfaces, although the GSM path may stop short of the transition to Wideband CDMA and still provide 3G services via Enhanced Data for GSM Evolution (EDGE).But all of these deployments share the fact that they employ the same cell-based coverage and hand-off schemes used now for voice services. There are certainly significant changes to be made to the base stations and base station controllers, but network access is still through these base stations.

With 3G voice and data services delivered via existing towers, access will become a more troublesome issue. A transceiver may be happily providing narrowband voice channels for a number of users until a "data-hog" requesting 2 Mbps downloads, roams into the cell. 3G base stations will be able to dynamically allocate bandwidth but not manufacture it. The solution, especially in urban areas, is not to add more transceivers, as interference is already a serious problem. The answer must come from complementary ways of accessing the networks.

The optimal means of network access depends on what you are doing and where you are doing it. Sitting in a hotel lobby or airport, a wireless LAN connection gives you the data connectivity you would want for catching up emails. If you're traveling down the interstate in a car while surfing the Web to find a nearby hotel, a 3G mobile wireless network will accommodate your needs. Together, bandwidth, location and mobility determine the type of access you need. No one technology is optimal. No one technology can be all things to everyone.

Intelligence and complexity will grow at the edges of the network. Services will be provisioned at the periphery of the network and independently of the transport mechanism. All the provisioning and hand-off intelligence needed to accommodate wireless access of different speeds will be moved to the periphery of the network. The center of the network flattens out with respect to functional layers, eventually approaching a pure IP transport model.

The path to 4G will be a natural evolution of networks, following the pattern of response to commercial need followed by network optimization. Operators are deploying the convenience of mobile data as an overlay and upgrade to the existing circuit-switched voice infrastructure. But this is not an optimal solution, and so to reduce operation costs and improve accessibility, we will see the intelligent edge devices managing multiple access technologies, dynamic bandwidth requirements and roaming users.