Globally, more than one million wireless basestations are now deployed, providing service to more than one billion people. Nowhere is the growth as dramatic as in China. China Mobile has nearly 150 million subscribers, while China Unicom has 83 million.
Wireless customers expect and even demand that their mobile-phone performance be comparable to landline quality. But the expansion of wireless services and traffic has led to an associated rise in radio frequency (RF) interference, directly contributing to a greater percentage of dropped calls, blocked calls and origination failures.
Our common design experience for enhancing basestation front-end technologies in the world's largest and fastest growing user population can be a beacon for others doing business in China.
In the past, wireless operators addressed network capacity and quality-of-service (QoS) by adding new infrastructure. Capital expense budgets have been trimmed, however, and finding real estate for new basestations is both expensive and difficult. Carrier adds can also be expensive as, on average, for every underperforming basestation requiring a new carrier, up to three additional surrounding basestations will also require a new carrier for hand-off and network-management purposes.
For these reasons, operators now are deploying alternative solutions, such as smart antennas, repeaters, high-selectivity filters and amplifiers. While these solutions might address either network sensitivity or selectivity, none solves the problems and issues associated with both simultaneously.
A technology now widely accepted for enhancing wireless network performance through both improved sensitivity and selectivity in one package is the cryogenically cooled receiver front end (CRFE). These systems consist of two key elements: a superconducting pre-selector filter and a cryo-cooled low-noise amplifier. By rejecting unwanted out-of-band interference and enhancing signals in the frequency band of interest, these CRFE systems typically demonstrate between 4 dB to 8 dB performance improvements when deployed in a basestation. They offer noise-figure improvement and eliminate intermodulation-induced noise. Field trials in operational networks show improved quality of service, capacity utilization, range extension and higher data rates.
Unlike some alternative solutions, these "plug-and-play" CRFE solutions are up and running in a couple of hours, and are lightweight and easily installed. They consume minimal power and have high reliability.
Anticipating these QoS problems as far back as 2002, China Unicom became interested in network enhancement products to meet the demands for network quality and performance. The company evaluated different network enhancement products that promised to increase capacity, improve in-building service penetration, increase spacing between future basestations and reduce mobile transmit power. At that time, China Unicom began discussions with Superconductor Technologies Inc., a U.S.-based company that develops, manufactures and markets high-performance products for wireless networks, including its flagship product, the SuperLink Rx 850 CRFE.
China Unicom and STI partnered in 2003 to conduct two field trials in Shenzhen in the province of Guangdong. These trials measured network performance and the typical China Unicom user experience before and after deployment of STI's SuperLink Rx 850 units in live China Unicom basestations.
Guangdong Unicom has historically been the test-bed for new technology for the China Unicom network and offered a stiff test for the high-temperature superconducting products. The Guangdong Unicom branches are responsible for about 25 percent of the Unicom subscriber base and 20 percent of the network's basestation deployments. Shenzhen Unicom serves suburban, urban, and dense urban areas in its network.
Calls on trial
The first trial was conducted in a five-site cluster in the Futian District network, a dense urban environment, with most basestation sites less than 350 meters apart. These basestations carry the most traffic in all of the Shenzhen and Guangdong Unicom networks and typically have dropped-call performance worse than the average for the network.
The second trial was conducted with two sites in the Nanshan district, a suburban high-tech office park with basestations separated by approximately 1 kilometer.
The Futian deployment was chosen to measure the benefit of the SuperLink Rx 850 in a demanding RF and network usage application. The suburban area was selected to show the performance enhancement achievable in an "average" Shenzhen Unicom network basestation.
The deployments were tested for four weeks under varying conditions, user patterns and parameters. The network performance data that was supplied by Shenzhen Unicom was reviewed and evaluated by the STI and Unicom engineering teams. The results with SuperLink Rx activated significantly exceeded all trial goals in dropped calls, ineffective attempts (blocked calls) and mobile transmit power.
As shown in the accompanying figure, the SuperLink Rx sectors showed dramatic performance gains, with up to 55 percent improvement in sector dropped-call rates and up to 30 percent improvement in ineffective-attempt rates. The overall results exceeded the established goals of 25 percent and 10 percent respectively. Unicom engineers determined that these improvements translated into the elimination of 250 dropped calls and 1,600 ineffective attempts per day.
These QoS improvements were achieved even though the Nanshan sites carried almost 20 percent more traffic once the SuperLink Rx units were operating.
Drive tests showed that the uplink mobile-transmit power was improved (lowered) by approximately 8 dB in the Futian cluster sites and by 10 dB to 12 dB in the areas covered by each Nanshan site. Histograms of the mobile-transmit power levels showed that the power reduction was consistent throughout the handset power range. To ensure that no changes occurred in the RF environment between the two drives, the downlink parameters (Ec/Io and FFER) were compared and found to be consistent.
Interestingly, the mobile-transmit power was reduced throughout the Futian cluster, thus providing more overall network benefit in mobile-transmit power than from an additional basestation, which only provides a reduction in the immediate surrounding area.
Moreover, performance and mobile-transmit power improved so significantly on the test sites that performance of the neighboring sites improved as well, although somewhat less than the sites directly involved in the trials. The reason: co-channel noise is reduced due to the lower mobile-transmit power in the SuperLink Rx sectors, thus improving the neighboring site performance.
By applying the mobile-transmit power data gathered from the field trials, engineers in Shenzhen Unicom's network optimization team can predict improvements in network capacity, quality of service, in-building penetration and range extension.
Liu Jian is general manager and Wu Libin is deputy general manager in the China Unicom Shenzhen branch (China). Robert B. Hammond is senior vice president and chief technical officer of Superconductor Technologies Inc. (Santa Barbara, Calif.).
See related chart