Though their circumstances differ in the details, mobile operators worldwide share common factors that are driving them to make their networks more spectrally efficient. The issues that are most prominent in operators’ minds include:

• Increasing voice and data traffic
• Ongoing financial pressures
• Government restrictions regarding spectrum availability and cell site locations
• The need to “future proof” network technology decisions

These demands continue to lead operators to the cdmaOneTM and CDMA2000 family of technologies that offer the most voice capacity, fastest data throughput, and greatest overall spectrum efficiency of all commercially deployed air interface technologies.

CDMA Leadership in Capacity Solutions

Expert analysis and experiences of operators around the globe prove that cdmaOne and CDMA2000 are without question the most spectrally efficient technologies available today. And new approaches on the horizon will further enhance the capacity of CDMA networks, providing even more powerful benefits to current CDMA operators and those migrating from other technologies to CDMA2000.

The capacity superiority of cdmaOne (IS-95A) in comparison to other 2G technologies has long been acknowledged. For instance, Deutsche Bank has compared voice and data capacity of various air interface technologies in 5 MHz of spectrum. On the voice side, cdmaOne (IS-95A) was shown to have higher voice capacity than TDMA, GSM and even GSM with AMR, the vocoder technology developed to increase GSM capacity. Deutsche Bank estimated that the only technologies with expected higher voice capacity than cdmaOne (IS-95A) were WCDMA and CDMA2000 1X. In fact, CDMA2000 1X was estimated to have the most capacity of all other air interfaces including WCDMA.

Number of users per 5 MHz of Spectrum

*Source: Deutsche Bank, May 2002

In addition, the dramatic, high-speed data capabilities of CDMA2000 1X are well documented. CDMA2000 1X’s impressive data delivery results do not come at the expense of overall network capacity – as one would see with GSM technologies such as GPRS and EDGE – but instead accompany doubled capacity on already spectrally efficient cdmaOne networks, enabling at least 80 Erlangs in 5 MHz of spectrum. Operator SK Telecom in South Korea reports its commercial use of CDMA2000 1X has boosted network capacity by 70%, to more than 231 Erlangs in 10 MHz of spectrum, and a U.S. operator has achieved 80% capacity gains on its CDMA2000 1X network. Such results on commercial networks allow CDMA2000 1X operators to provide quality service to high-density areas and build capacity now for anticipated traffic while keeping a handle on capital expenditures.

While the CDMA2000 1X success story is retold by every operator who deploys this spectrally efficient technology, even more exciting developments are on the horizon. Solutions such as Selectable Mode Vocoder (SMV), mobile diversity antennas for handsets and smart antennas for networks will take CDMA’s long-standing tradition of spectral efficiency to even greater heights.

SMV is a breakthrough technology providing capacity and quality gains on cdmaOne and CDMA2000 networks. SMV, a standard released by 3GPP2, is designed to let wireless operators gain up to 75% more system capacity when using SMV’s lower encoding rates while preserving high-quality voice delivery.

SMV selects the optimal encoding rate based on input speech characteristics and therefore ensures voice quality throughout a phone call. There are four modes of SMV operation, each offering a different percentage of frames at different coding rates. Operators can select their preferred operational modes to achieve their target average bit rate and voice quality. Further, SMV modes can be selected dynamically on a frame-by-frame basis. Backward-compatible CDMA SMV will easily complement technologies such as CDMA2000 1X and CDMA2000 1xEV. Products offering SMV are expected to reach the market during the second half of 2003. Qualcomm has already begun developing its MSM-6100 chips to support SMV.

Independent listening tests conducted at Dynastat Labs in October 2000 confirmed SMV’s higher sound quality compared with Adaptive Multi-rate Vocoder (AMR), a new technology being developed for GSM networks, even when lower transmission rates are used. Dynastat concluded that SMV is a more efficient vocoder compared to AMR because SMV delivers similar or better quality than AMR multi-rates at lower encoding rates. For those who want to learn more about SMV, the CDG offers an online SMV demo at the CDG website (www.cdg.org); the test is also available by phone.

• SMV is a more efficient vocoder compared to AMR as it delivers similar or better quality than the AMR multi-rates at lower encoding rates.
• Encoding rates shown assume no voice activity detection for either AMR or SMV and represent the active speech coding rate both AMR and SMV.

The CDMA community is also aggressively targeting technology based on mobile diversity antennas, a handset-based smart antenna methodology that requires no network modifications but will increase capacity by a factor of two. The idea behind this approach is to combine signals from two separate antennas to reduce the impact of spatial variations in signal strength and thus increase the average data rate available. Mobile diversity antennas can also be used when one needs to put antennas for two unlike services – for instance CDMA for voice and data and GPS for tracking – into a single device.

The CDMA community is leading the way in mobile diversity solutions, and companies such as Qualcomm will introduce commercial ASICs supporting diversity reception in 2003. Likewise, semiconductor startup InnovICs unveiled its diversity technology in May 2002 and aims to improve coverage by 50% and increase cell capacity four times for high data rate signals.

Mobile diversity solutions can be used alone or in combination with smart antennas in the network infrastructure to provide even more impressive capacity gains and improved customer satisfaction through better overall voice and data call performance. Network smart antennas, which do not require any changes to end-user mobile devices, can be embedded in base stations or added to existing networks as appliqué solutions. Such antennas can be deployed on a per-cell-site or per-sector basis in order to alleviate specific or localized capacity issues. Both large and small vendors offer smart antennas for the CDMA marketplace.

Capacity Enhancing Technologies for GSM Networks

The GSM community has also developed a new technology to increase capacity called Adaptive Multi-rate Vocoder (AMR). This technology is being embraced by North American GSM operators whose current networks are facing capacity strain. AMR, originally selected by 3GPP in 1999 as the mandatory speech codec to improve GSM voice quality, is slated for initial handset deployment in 2003.

AMR is not a variable-rate codec like SMV and operates at one of eight available data rates, which provide variable tradeoffs between speech quality and system capacity as a function of traffic loading. Groups backing AMR claim it will offer capacity four times that of TDMA when used in half-rate channel mode, which is a new technique wherein GSM time slots can multiplex two voice conversations. Even supporters of half-rate channel mode admit that this approach sacrifices voice quality for capacity gain because changing channel conditions can greatly impact an AMR voice call and AMR does not dynamically readjust when faced with poor channel conditions, which might be caused by changing proximity to a cell site, loading or geographical encumbrances.

Along with AMR, the GSM community advocates the simultaneous use of another emerging technology called Dynamic Frequency Channel Allocation (DFCA) as well as 1/1 frequency reuse, the latter being a capacity-enhancing method that is already used in CDMA networks. This level of frequency reuse has only recently been envisioned for GSM networks that were not designed with this approach in mind and are not likely to achieve the same high level of loading found in CDMA networks.

Another capacity-enhancing technique being developed for GSM, single antenna interference cancellation (SAIC), requires GSM base stations that are fully symbol synchronized. SAIC is envisioned as being able to predict and cancel signal interference. This approach is quite immature and it is still far too early to establish reliable capacity gains from it.

Members of the GSM community claim that the simultaneous usage of AMR, half-rate channel mode, DFCA and 1/1 reuse will enable GSM to achieve a theoretical 170 Erlangs in 10 MHz of spectrum. Not only are these approaches for the most part unproven, particularly when taken together, there are critical questions regarding whether such dramatic results can ever be achieved in bona fide commercial deployments of GSM networks.

A widely publicized study touting increased capacity results from the use of these evolving technologies in GSM networks included numerous questionable assumptions and was modeled on usage in a simple pedestrian environment rather than the mixed pedestrian and highly mobile environments in which CDMA2000 1X already successfully delivers doubled capacity. Computer simulations have shown that a more realistic, though still optimistic, capacity increase from deploying AMR and related technologies would be on the order of 100 Erlangs in 10 MHz, and that would come with a notable impact on overall voice quality. As such, GSM operators will be challenged to swap acceptable end-user experiences for increased network capacity.

And while it has been claimed that DFCA can increase voice capacity by 50%, that technology requires synchronization of all time slots in all GSM base station voice channels to achieve continuous handoffs in the network. GSM networks were never designed to operate in a synchronized mode, and adding the necessary software and hardware to make them do so will likely be an expensive and complex process.

EDGE, being marketed as an evolutionary upgrade for GSM operators, will present other issues for GSM network operation. EDGE Release ’99 expected to be deployed by some operators in 2003, requires the addition of new network elements and initially will be merely a data solution providing approximately three times the data throughput of GPRS networks, that today average 20-30 kbps in commercial deployments. The next evolution step for EDGE is Release 5. Analysts question whether Release 5 equipment will even work with Release ’99 EDGE infrastructure. Deutsche Bank analysts have written that operators deploying Release ’99 infrastructure “will require an entirely new core network (perhaps a RAN as well) if they want to offer Release 5 services (real-time multimedia services, VoIP).” Similarly, researchers for Morgan Stanley have written, “Many technical and performance aspects of EDGE are still unclear.”

It is important to note that the use of EDGE and its new 8-PSK modulation scheme, in an effort to increase GSM data speeds, will require operators to make some decisions regarding their overall technology mix. High-speed data over EDGE will demand very benign channel conditions with no interference, no fading and a carrier-to-interference ratio (C/I) that is greater than 20dB. However, those conditions will be impossible to achieve if there is 1/1 frequency reuse. Further, AMR will not be compatible with VoIP via EDGE as EDGE is supposed to be implemented with 4/12 frequency reuse for high data rates. To implement EDGE for voice, operators will have to choose between AMR and low-frequency reuse implementation.

Impending technological innovations will enhance capacity for both GSM and CDMA networks. The CDMA community has already benefited from the strong tradition of superior capacity, voice quality and CapEx savings that was set by the original IS-95A standard and has continued with CDMA2000 1X and CDMA2000 1xEV. Emerging technologies that build upon the already spectrally efficient CDMA2000 1X platform will bring even greater benefits to operators and users of CDMA2000 technology for many years to come.