GSM To Meet Or Exceed 1X Capacity, Backers Say
By Joseph Palenchar -- TWICE, 8/19/2002
Bellevue, Wash. — North American carriers adding GSM to their TDMA networks defended their technology choice during a recent webcast in which one market researcher called GSM "the most spectrally efficient voice system available today."
The researcher, Peter Rysavy, claimed GSM network upgrades will soon enable GSM carriers to match or exceed CDMA 1X capacity in real-world conditions. GSM voice capacity will be further enhanced with the introduction of EDGE, which was previously promoted only as improving data throughput. GSM carriers will achieve further gains by implementing additional upgrades, including 3G W-CDMA technology that lets GSM carriers reuse much of their existing infrastructure.
Rysavy's research was prepared for GSM advocacy group 3G Americas.
During the webcast, Cingular chief technology officer Bill Clift went as far as saying that a GSM network outfitted with EDGE and an adaptive multirate vocoder "is nearly as efficient" as W-CDMA. For data throughput, however, W-CDMA will have the clear advantage over EDGE. W-CDMA accelerates dataspeeds to 10Mbps in stationary applications (up from 2Mbps stationary, 384Kbps at pedestrian speeds, and 144Kbps at mobile speeds, following a spec upgrade). EDGE, in contrast, promises data at peak handset rates of 236Kbps, or up to a peak 473Kbps in the network, with the highest speeds not available in mobile use, the group said.
At 10Mbps, W-CDMA is " the fastest of any cellular technology," Rysavy told TWICE.
Clift noted, however, that EDGE enhancements can narrow the datarate gap, with compression and other techniques boosting EDGE's user-perceived throughput by 2 to 5 times.
Cingular will implement W-CDMA "when and if that spectrum is available in the U.S." AT&T Wireless, in contrast, has said it has enough existing 1.9GHz spectrum to implement W-CDMA.
By mid-2004, Cingular will complete its GSM overlay, with the "preponderance" of its network getting GSM by the end of 2003, Clift said. Today, "about 36 percent of the [Cingular] network is GSM/GPRS," he noted.
Not every analyst agrees with Rysavy's optimistic assessments, but Rysavy said his GSM voice-capacity estimates are greater than other analysts' estimates in part because he takes into account recent and soon-to-come GSM innovations.
Early GSM used the TDMA approach of time division multiplexing, Rysavy said. Since then, GSM has added such techniques as dynamic power control, which constantly adjusts a handset's transmission power to reduce interference and increase capacity.
More important, GSM networks have implemented frequency hopping so that conversations hop through multiple channels, averaging down the effects of interference, he said. That "allows a greater number of radio channels to operate in each sector," increasing capacity by 2 times, he continued.
Rysavy said he believes all U.S. GSM networks use frequency hopping.
On top of that, U.S. networks could add a technology called adaptive multirate (AMR) voice codec, which "dynamically allocates bandwidth to voice or error control" depending on the RF environment, he explained. That allows conversations to continue under adverse radio conditions, such as heavily loaded networks.
AMR increases fractional loading, or the percentage of radio channels that can be used at the same time, from the 20 percent to 30 percent range to the 40 percent to 60 percent range, he said.
Of all the current and on-the-horizon GSM advancements, AMR and frequency hopping "provide the greatest immediate benefit," he said in a white paper.
Cingular will begin offering AMR handsets "this fall or early winter" along with AMR service, Clift said.
GSM can boost analog capacity by 16 times if frequency hopping and AMR are used, nearly matching CDMA 1X's 17x capacity gain, Rysavy said. (See chart.) When DFCA (Dynamic Frequency and Channel Allocation) technology is added, GSM capacity exceeds 1X capacity, he said. DFCA dynamically assigns radio channels in an optimum way according to the current RF environment, he explained.
In particular, Rysavy said that if 50 percent of phones in a GSM network use AMR, voice capacity will increase by 50 percent, and with 100 percent of subscribers using AMR phones, voice capacity can jump by 150 percent. Adding DFCA to AMR boosts capacity an additional 20 percent to 25 percent.
AMR and DFCA are "emerging and proven," he said. Other technologies under development to further boost GSM-network capacity include EDGE's 8-PSK modulation. That will boost capacity by another 15 percent to 20 percent.
GSM's gains can come without "a significant increase in terminal prices," Rysavy noted.
Adoption by GSM carriers of the third-generation W-CDMA technology will boost capacity to 20 times that of analog, exceeding CDMA 1X's 17-times analog, he said.

















