Standards Soup

Three companies work to develop fast, spectrally efficient standards for 3G CDMA networks.

Service providers want to capitalize on booming projections for wireless-data usage and conserve valuable spectrum at the same time. A new study by ARC Group predicts mobile-data users will total nearly 1.2 billion worldwide in 2005. These primary drivers - rising market predictions and spectrum scarcity - are accelerating the development of a new air interface for CDMA networks. Three companies - LinkAir, Lucent and Motorola (with cooperation from Nokia and Texas Instruments) - are working on proposals for what is being called Phase II of 1XEV.

But before delving into the developing technologies for Phase II, it is helpful to recap what happened with Phase I, which has not yet been deployed. Phase I is based on Qualcomm's high-data-rate (HDR) technology and also is called 1XEV-DO, the DO meaning data only. Large CDMA operators required a data-only air interface that would get them to market as fast as possible, in order to compete with the TDMA push toward EDGE, according to one wireless executive. Smaller operators - sometimes called fast followers - may not be following so quickly this time. Many of the smaller service providers, it seems, may choose to skip Phase I and move directly to Phase II in order to save one generation of upgrade costs. The additional time also would allow for in-depth observation of wireless data's popularity with the general public. The risk in this approach is obvious. If wireless data takes off fast, there will be dust to eat.

All CDMA-based service providers are encouraging the development of one standard air interface for Phase II. But with development still taking place, and harmonization of the three proposals yet to begin, high-speed packet-data transmission at average throughput rates over 1Mb/s is years away.

Jim Takach, CDMA Development Group (CDG) director of advanced programs, has been following the progress of these standards discussions for some time. The CDG, as a market representation partner member of the 3GPP2, facilitates discussion and action among corporate members of the 3GPP2.

"They created a group to evaluate different proposals ... in August," Takach said. "That group will receive the different proposals - from LinkAir, Motorola and Lucent - review, evaluate and attempt to harmonize them. We will be actively involved in that process versus establishing our own process."

The target date for publishing the baseline text of the Phase II standard is debatable. Some companies are pushing for May 2001. Others think October 2001 is more realistic.

After a technology is picked, or a synthesis of the proposals is produced, service providers jointly will begin testing the air interface with vendors.

"Our requirements were developed with the carrier community in mind," Takach said. "Obviously, high throughput and low latency (are requirements). They want a seamlessness with their current networks. They don't want it to be disruptive. We're in this age of wireless Internet and high-speed data, and it's a little unknown as to how to operate it, install it and manage it."

Despite these unknowns, efficiency remains a constant goal.

"The story behind all of this is that it's innovation within a small amount of real estate," Takach said. "3G is not about spectrum. It's all about doing more with less. Providers want to use that real estate as effectively as possible. That's the drive behind 1XEV Phase I and Phase II."

Dueling Proposals LinkAir Communications has developed Large Air-Synchronized CDMA (LAS-CDMA) as its proposal for Phase II. LinkAir recently achieved data throughput rates of more than 5Mb/s during a lab trial of LAS-CDMA in China, said Frederic Leroudier, vice president of standards and business development.

"We think that with more and more subscribers and more network activity from data and PDAs, there is a real need for 3G," Leroudier said.

He pointed out that there are nearly 65 million mobile subscribers in China, and further market growth is expected. LinkAir is aiming LAS-CDMA at current CDMA networks. However, it has plans to integrate the technology with other types of networks as well.

"The technology is not there to satisfy the demand for bandwidth," Leroudier said. "It makes sense to look for a new technology, which won't require more spectrum and significant infrastructure costs."

With LAS-CDMA, networks can expand the capacity for data and voice while using today's spectrum and equipment - primary requirements for Phase II proposals to the 3GPP2. Leroudier said LAS-CDMA uses optimized spreading codes that create a clear channel for more users. Data throughput peaks at a blistering 5.5Mb/s using LAS-CDMA. But average throughput is closer to 2.5Mb/s, though it is difficult to estimate.

LinkAir has developed a working relationship with the Chinese Ministry of Information Industry (MII), going back two years to when the MII endorsed Professor Li Daoben's invention of LAS-CDMA.

Leroudier said the technology is backward compatible with 2G networks. And LAS-CDMA can be integrated with emerging IP-based network infrastructure, which Leroudier predicts will be deployed by 2004.

LinkAir recently lured CDMA pioneer Dr. William Lee away from Vodafone Airtouch. Lee, who is now chairman of LinkAir, said in a recent interview that LAS-CDMA was the most innovative CDMA-based air interface he had seen since 1989, when he first began working with CDMA. (Lee has put his experiences into the new book, Lee's Essentials of Wireless, scheduled for publication in November by McGraw-Hill.)

LAS-CDMA capitalizes on a code-isolation scheme developed by Daoben. By isolating certain smart codes during transmission, the LAS-CDMA system optimizes spectrum and increases efficiency, Lee said.

These smart codes also allow voice to be strongly separated from data during transmission, which allows for low-power voice transmission and high-power data transmission, Lee said.

"That is the beauty of the LAS technology," he said. "Without this coding window found by the professor, it would not be possible. Without the window, it was hopeless."

Also, through the use of a time-division-duplex scheme (TDD-CDMA) rather than frequency-division duplex (FDD-CDMA), LAS-CDMA only uses one frequency for both the uplink and the downlink, rather than one for each.

Key characteristics of Lucent's proposal for Phase II of 1XEV include increased voice and data capacity, and carrier performance, said Ed Chao, Lucent senior manager for 3G CDMA data. The proposed technology also would be backward compatible with existing IS-95, 1X and future 1XEV Phase I networks.

"We have a number of enhancements that we've had in our hip pockets ... that by the time this (standard) comes out we're going to be ready to implement from an economic standpoint," Chao said, "but in a way that will make sense to allow footprint and base stations to be deployed economically."

Chao said that some specific goals for Lucent's Phase II proposal have not been defined.

"We took a look at it and found that 1XEV Phase I is as good as it's going to get for the next couple of years in terms of data, specifically, because it is optimized specifically for data," he said. "Use that as a benchmark, and if you can get to that, great. But if you can get to 75% or 80% of the way there, that's probably OK as well. We are going to shoot for as much as we can get. Phase I gets you up to 2.4Mb/s peak rate. We want to get to somewhere around that capability."

Chao predicted that the peak data rate for Phase II would be about 2.4Mb/s, or equal to that of Phase I. However, he acknowledged that other proposals claim a peak rate of 5Mb/s.

"I know that the 1Xtreme camp has advertised rates of like 5Mb/s, but that requires a whole new base station, a whole new set of antennas, new amplifiers, new everything," Chao said. "So from a feasibility standpoint, we know we can do 1XEV Phase I, so let's shoot for that as the bar."

Lucent had recommended a baseline text target date of the end of next year, Chao said. But others have advocated moving that date up to May or June 2001.

"We'll work with them to do that," he said.

Chao said the three proposals aim for the same goal but differ on mainly theoretical levels.

"We've had discussions with the 1Xtreme and the LAS-CDMA camps," he said. "Actually, when they first saw our proposal, which we've had in our hip pocket for awhile, they said there were many great ideas in it. I think they're trying to achieve some similar objectives that the CDG had defined, and they want to work with us."

Motorola's Phase II proposal, 1Xtreme, will provide increased efficiencies in voice and data capacity, according to Tim Berghuis, manager of strategic product management for CDMA systems.

"From a straight 1X perspective, we believe that the (1Xtreme) voice capacity is about equal to 1X," Berghuis said.

But by using non-air interface techniques such as transmit diversity, Motorola proposes to improve 1Xtreme's voice capacity, he said.

"From a data perspective, we believe we can double the capacity of that carrier relative to Phase I," Berghuis said. "That's more than a 4X increase in data-throughput capacity from 1X. We are looking at (average throughput) capacity of around 250kb/s to 300kb/s for 1X, HDR being around 600kb/s to 650kb/s, and all of our simulations have shown that we can reach about 1.2Mb/s on 1Xtreme - average throughput."

Berghuis noted that peak rates would be much higher; however, average rates are more realistic.

The evolution to 1Xtreme would differ depending on a service provider's current network and subscriber base, Berghuis said.

"What I would envision is that an operator would add a new 1Xtreme carrier, as an example," he said. "If these services are as great as everyone is telling us they're going to be, there is going to be a demand. To do that, your existing subscribers are going to have to start to migrate, which is going to mean a handset trade-in. And when they start to do that, the existing spectrum that is being used is going to become underutilized. Eventually, it would get to the point where it would make sense for an operator to remove that carrier. It's like with analog today. There are not nearly as many analog channels as there were five years ago. I would envision that same type of an evolution."

Although acknowledging the differences in all three proposals, Berghuis said Motorola is willing to work with LinkAir and Lucent to reach consensus for the Phase II standard.

"We actually think there might be some potential there for collaboration," he said. "Those discussions have yet to take place. We are hoping that potentially the operator community could drive a unified solution. That will be an ongoing process for us."

* May, the CDMA Development Group published its "Standards Requirements for the Evolution of cdma2000 1X, Version 2.0." Those requirements must be met by all proposals for Phase I and Phase II of 1XEV.

The document defines Phase I as a system optimized for non-real time, high-speed packet-data services. Phase I is being called 1XEV or 1XEV-DO, and employs Qualcomm's HDR technology.

Phase II proposals should concurrently support high-speed packet data and real-time service. In an integrated approach, the objective for Phase II is to integrate Phase I capabilities on the same carriers, while retaining the ability to maintain packet-data services on a separate carrier.

Phase I has been approved and is scheduled for trials in January 2001; Phase II is scheduled for a baseline text some time in mid- to-late 2001.

Phase I and II proposals must incorporate include the following:

- The Phase I standard will enable high-speed, non-real-time packet-data capability; interoperability - including hand-off with cdma2000 1X channels for packet services; improved signaling and packet-data throughput; deployment in all existing and planned frequency bands where cellular and PCS systems are or will be deployed; asymmetric data rates to serve the majority of Internet applications; fast connection times compared to existing IS-95 standards; always-on capability for data services; use of existing IS-95A/B and cdma200 antennas and active, smart and directional antennas for specific applications; simple user-device activation; a graceful evolution from IS-95 A/B and cdma2000 that minimizes impact to terminals and infrastructure; a graceful evolution to Phase II that also minimizes impact to terminals and infrastructure.

- Phase II will build on the aforementioned Phase I capabilities. In addition, the Phase II standard will enable three modes of traffic delivery (real-time, non-real-time and mixed real-time/non-real-time); interoperability with cdma2000 1X channels for packet and voice services; mixed real-time and non-real-time traffic on the same carrier; network support for multiple, concurrent packet sessions; voice and packet service options and capability for future globally harmonized audio and video codecs; consistently excellent voice quality; use of existing cdma2000 and IS-95 A/B antennas; double the voice service spectral efficiency of cdma2000 1X; improved packet information data rates; and quality of service parameters including priority, minimum acceptable data rate, maximum permitted data loss rate and latency or delay.

- Also, the Phase I and II standards should achieve total call-set-up time to home country destinations of less than two seconds; system acquisition, registration, authentication and network log-on time of less than two seconds (international calls may take longer); packet-access failure rate of less than 0.5% of access attempts; and undesired packet-session terminations of less than 1% of completed terminations.