CDMA Technology
Members Sign-In
Set Menus

Delays to the launch of GPRS handsets are not bad news for everyone. It could be an opportunity for the CDMA2000 community to steal a march on its GSM-based rivals. Are CDMA handset, chipset and service developers ready to make the most of this opportunity?
By Ray Hegarty

The anticipated launch of 2.5G networks based on GPRS has stumbled in recent months. A delay with the delivery of GPRS handsets has hampered operator's plans. European network operators--desperate to launch faster 'always on' data applications based on packet-based networks--are blaming the top three GSM handset suppliers. At the moment Nokia has yet to introduce commercial GPRS handsets, Ericsson was recently forced to recall faulty GPRS handsets and only Motorola has managed to release handsets, albeit in limited quantities.

In part, the delay has been blamed on the tendency for the GPRS handsets to overheat when more than one timeslot is used to enhance bandwidth. In trials GSM operators are said to be getting a maximum10kbit/s per timeslot, which can be expanded to 60-70kbit/s by using an entire radio carrier--however, the use of an entire carrier is deemed to be economically inefficient.

Supporters of the CDMA2000 path to 3G claim that 1XRTT handsets will be more technologically efficient than GPRS. What's more, CDMA advocates claim that 1XRTT can be deployed in existing spectrum, requiring only 1.25MHz of spectrum in each direction for each deployment. For operators that do not have the luxury of additional 3G spectrum, 1XRTT, its next stage enhancement 1XEV-DO (Data Only) and, ultimately, 1XEV-DV provide a compelling alternative technology path to that of GSM.

At the CTIA Wireless show in Las Vegas in March, Motorola showed how CDMA2000 1XEV-DV could be used to simultaneously run a web browser at a high data rate as well as streaming video. Significantly the demonstration ran over wireless carrier Sprint PCS's existing spectrum. The Motorola demonstration was designed to eliminate skepticism that EV-DV will not be able to achieve data rates the CDMA industry envisions for 3G technology. 1XEV-DV was designed to offer data throughputs of up to nearly 5Mbit/s, with an average user throughput of about 700kbit/s. When the technology is commercialized, an operator can install 1XEV-DV by a simple channel card replacement and software upgrade to an existing network, all within 1.25MHz of spectrum, claims Motorola's Scott Wyman.

CDMA handset vendors, including Kyocera, Samsung, Motorola and Ericsson, claim that future CDMA handsets--1XRTT, 1XDO (data-only technology based on Qualcomm's former high data rate product) and 1XEV-DV--can utilize existing spectrum more efficiently and benefit from cheaper network upgrade costs. And they reiterate the point that CDMA 1X and EV require only 1.25MHz spectrum in each direction for deployment; in comparison WCDMA requires a full 5MHz of clear spectrum.

In addition, Qualcomm chairman and CEO, Dr Irwin Jacobs, claims that all CDMA 1XRTT handsets will be backwards and forwards compatible with earlier versions of CDMA, while GPRS and WCDMA handsets will require dual mode chipsets to work on pre-existing GSM networks.

But what makes CDMA handset and chipset suppliers think they can deliver on these promises? One of the core reasons for confidence is CDMA's reliance on IP. The GPRS network is not based on standard IP network elements, resulting in a more complicated integration than the CDMA2000 packet data path. CDMA is designed with standard IP in the handsets and network.

IP protocols
By incorporating standard IP protocols and network IP routing CDMA2000 offers an internet experience at least equivalent to that offered by GPRS and EDGE but with lower overall investment in equipment and human resources--all without decreasing voice capacity. Network operators are also more likely to find the skills to integrate equipment and develop services because the same IP elements are used on the landline internet, according to CDMA Development Group executive director, Perry LaForge.

Next generation handsets, with increased data capabilities will have to try and conserve battery life as much as possible. CDMA handsets include variable rate voice coding enabling speech bits to be transmitted at only the rates necessary for high quality voice, conserving the battery power of the subscriber handset. In other words, the handset requires lower radio frequency power requirements. This allows the phone to transmit at significantly reduced power levels, explains Qualcomm senior vice president, Lou Lupin. Multipath signal processing techniques combine power for increased signal integrity. Additional benefits to the subscriber include increased talk times for handsets, more secure transmissions, and special service options such as data, integrated voice and data, fax and tiered data.

Building on the existing foundations, Qualcomm has started to employ several additional device-side technologies designed to add more sophisticated data services to the handset.

At the start of the year the company introduced BREW (Binary Runtime Environment for Wireless). BREW is a thin application environment that is supposed to provide developers with an open, standard platform for CDMA2000 devices on which to develop products. The idea is to enable end users to download applications over the air through their carrier's network and configure their wireless devices to suit their personal requirements.

BREW is an alternative technology to Sun Microsystems' J2ME-Java 2 Micro Edition (although, confusingly, several companies including Hewlett-Packard plan to add their own Java virtual machine on top of BREW).

Both BREW and Java enable programmers to create small applications for mobile phones thereby enabling end users to download small applications off the operator's network. Hewlett-Packard's MicroChai JVM will give BREW handsets access to existing web-based Java applets. HP has integrated its MicroChaiVM onto a BREW-enabled prototype Kyocera handset and prototype color-display handsets manufactured by Samsung and Denso Corp according to William Woo, general manager of HP's Embedded Software Operation. In March, Qualcomm put MicroChaiVM and an application written for the Java platform on its test download server and began testing over-the-air downloads across US carrier Verizon's network.

The key difference between BREW and J2ME, according to Irwin Jacobs, is that BREW enables developers to create applications that use cdmaOne and CDMA2000 without having to customize code for each different handset model. Such an advantage is destined to be short-lived once versions of the J2ME Connected Limited Device Configuration and Java K Virtual Machine have been stabilized, effectively negating the 'write once, run anywhere' advantage BREW might claim.

Nonetheless, Tadashi Onodera, executive vice president of Japanese wireless carrier KDDI, claims the technology has other benefits including better performance and better security. BREW, by compiling the run-time code, provides for better handset application performance and less code compared to the interpreted code approach that is used by Java. That translates as faster and less taxing on the handset processor than Java.

To ensure secure m-commerce, BREW uses a digital signature or digital key system--a Qualcomm-tested software approach that handsets use for authentication--allowing the application to be downloaded and used on the handset. The advantage of such an approach is that applications can run without being limited to a prescribed area, giving greater access to core handset functionality. In comparison, Java uses a protected environment or 'sandbox' approach to security, where applications are able to run in a delimited and protected area of the device. BREW also benefits from better load and runtime because it is binary, which translates as being less memory- and processor-intensive.

In order to encourage the development of small runtime applications for CDMA handsets, Qualcomm has released a free BREW SDK (Software Development Kit) that includes Windows-based C and C++ development and testing tools. The company has been busy signing partnerships with software developers in the games, position, location/determination, entertainment and mass media and mobile commerce sectors.

A further Qualcomm innovation destined to benefit 1XRTT and 3XRTT handsets is radioOne, a new Qualcomm zero intermediate frequency (ZIF) radio frequency (RF) architecture. Unlike current CDMA radios, which convert radio frequency signals to and from baseband in multiple steps, converting RF to or from intermediate frequency (IF) and then converting IF to or from baseband, ZIF radios convert incoming RF signals directly to or from baseband analog signals. Elimination of the intermediate frequency conversion reduces the number of filters and RF chipsets needed in the handset. In turn the elimination in components improves battery time, reduces costs, and enables smaller, more flexible handset designs. Qualcomm expects to incorporate radioOne technology into its next generation 1xMC/EV/GPS commercial products at the end of this year.

Such Qualcomm-backed innovations are key. Qualcomm's chipset division, for instance, dominates the CDMA handset market with a market share of more than 70 percent. In the past, Qualcomm offered a very limited chipset portfolio that many said held back the efficiency and design of cdmaOne handsets. However, more recently Qualcomm has expanded the depth and breadth of its CDMA expertise in order to persuade wireless carriers that CDMA is the best choice for next generation services. The company is now offering or developing chipsets that cover the full range of CDMA air interfaces with varying levels of content. The company wants to provide a core platform upon which handset vendors can add more specific functions and applications.

For instance, Qualcomm's new MSM3300 chipset offers multimedia, internet and Bluetooth features. Qualcomm claims to have already shipped about 500,000 of its first generation 1X chipset, the MSM5000. There are a number of products still under development including a second generation 1X chipset that supports higher data rates, and a 1XEV chipset.

More detailed information about CDMA2000 handset plans are harder to come by. The handset vendors we spoke to were vague about their intentions, citing competitive advantage. Samsung, for instance, said it had limited availability of 1X handsets and is currently trying out the technology in Korea.

Meanwhile, Japanese operator KDDI has been trialling CDMA2000 1XEV-DO technology for cdmaOne and CDMA2000-compatible phones. The wireless carrier claims it has transferred data at an average speed of 600kbit/s reaching up to 2.4Mbit/s in the downstream direction in a standard bandwidth 1.25MHz channel. It said sector throughput was 1.2Mbit/s.

KDDI has been running CDMA2000 1XEV-DO trials in cooperation with Hitachi, Sony and Kyocera. "Based on the results from this trial, we will be enhancing our existing CDMA handset line with Qualcomm's MSM5500 dual-mode CDMA2000 1X/1XEV-DO integrated circuit," says Yukio Fukumura, general manager at Kyocera's R and D center.

Despite CDMA's handset vendors reluctance to be specific, it is possible to foresee eclectic and ambitious product strategies. Most models will provide screen resolutions some four times better than at present, and will come with basic BREW and Java capability, integrated CMOS digital video camera and some 2megabytes of onboard memory, providing the capability to handle MPEG4 video files and MP3 digital music. Many models will have slots for Compact Flash, Sony Memory Stick, or Panasonic SD cards, allowing users to store music or other files.

For the initial generation of 3G phones, battery life will be reduced since the additional memory, improved displays, video cameras, and processors will burn through more wattage. Handset vendors claim this should quickly improve.

On the downside, the 1X and 3X handset market will experience intense competition and declining margins. Shortage of key components could also the hinder the fast and efficient introduction of next generation devices. Wireless operators are showing an increasing unwilling to subsidize handsets or encourage growth through generous prepaid offers. The technology may be rolling out now but, ultimately, we are unlikely to know whether CDMA2000 handset vendors can fully deliver on their promise until this time next year.

This article first appeared in CDMA World Focus June 2001 published by Informa Telecoms. For more information see