Breakthrough Ideas (Continued)
The Synchronous/Asynchronous Conundrum.
In trying to circumvent Qualcomm’s IPRs, the European’s carried over the asynchronous approach used in GSM into UMTS. To partially offset this ill-considered decision, the masters-of-FUD claimed that Qualcomm synchronized by using a military system,GPS. Admittedly, GPS has military uses. The U.S. uses GPS to improve the guidance of its missiles. No longer managed by the U.S. military, however, GPS remains a prime navigational/timing solution used internationally. For instance, among the many civilian uses of GPS around the world, nations cooperatively synchronize the world’s major grid of power plants using GPS.
In asynchronous operation, each base station has independent timing, which means that a mobile does not have prior knowledge of the relative time difference between its present and future base station. Because CDMA2000 synchronizes all base stations with GPS time, the mobiles can “decide” when the handoff will occur by comparing the relative strengths of the synchronized multiple pilot signals that it is receiving. Synchronized soft handoffs solved the fading problem that asynchronicity produced, but Europe hoped, but failed, to gain leverage for cross-licensing by introducing this GSM “improvement” into its “GSM-derived” UMTS standard.
The problem with asynchronous handoffs is that signals fade rapidly at precisely the wrong time and place—at the cell border when the old signal is fading before the new one has come up and where a handoff must occur or the call will be dropped. This creates a 2-5 second fading-hole. Fading-rate exponentially increases the power and processing resources required for a timing-search-solution. Thus, the search window widens rapidly at cell borders, increasing the power consumption used to search, consuming the battery’s power and, thus, its talk- and standby-times
The problem here is not that UMTS does not try to implement asynchronous soft handoffs. However, to do so successfully remains unnecessarily complex. Substituting asynchrony for synchrony in a failed effort to circumvent Qualcomm’s intellectual property created the needless complexity. The lack of design experience in spread spectrum RF systems further complicates it. Agreeing that the problem is complexity, in a sales brochure on their web site, Lucent Technologies advises:
“A comprehensive understanding of the way in which CDMA exploits the statistics of fading, location, velocity and usage is key to optimizing performance…the benefits in UMTS soft handoffs are critically dependent upon setting handoff thresholds correctly. If thresholds are too low, soft handover links that are not really needed will be added, thereby generating excess co-channel interference and excess call processing loads. If thresholds are set too high, mobiles that require soft handover links to maintain voice quality will be compromised. Field experience with soft handover is key to selecting thresholds that are optimal to morphology, subscriber distributions and RF environment…management of co-channel interference is critical…Strategies for managing co-channel interference are complex but have been carefully developed within IS-95 deployments…These techniques will be essential in realizing W-CDMA performance.”
lucent.com
Because UMTS is a totally new RAN using new set of RF spread spectrum parameters and design rules for its WCDMA airlink, there is no substitute for ascending its learning curves¾from chipset- to system-design. Whether you wish to understand its complexities or hope to optimize its performance, experience with spread spectrum is required.
On the one hand, quasi-political consortia, like 3GPP, failed to use a selection mechanism, like the scientific method, to avoid or solve technical problems. On the other hand, Qualcomm is using its vast experience to develop innovative and cross-tested solutions to WCDMA’s problems. Nonetheless, even Qualcomm’s best efforts using the flawed UMTS specifications cannot equal the performance that comes from their own optimized 1X specifications. Choosing the UMTS route pays the price in decreased spectral efficiency. Qualcomm just pragmatically accepts that not everyone can admit their mistakes, but wants to help those open-minded parties who can.
There are three potential hardware solutions to WCDMA’s handover problem. First, reduce power by adding four to five times as many new towers to reduce the size of asynchronous search windows. (On the Isle of Man, the UMTS FOMA test site required 28 towers for only 22 mobiles. This was not a promising trial-outcome.) Second, add more processing power to the timing filter so it can acquire a new signal within the present search window, but this would require the generation-after-the-next-generation processor. Third, add synchronization.
Qualcomm has suggested three synchronization schemes that could be implemented with no changes to the 3GPP standard. QCT has demonstrated that synchronous deployment of WCDMA improves performance¾up to 2.5 times longer handset standby times, improved paging performance, and enhanced system capacity. WCDMA system performance improved as a joint function of the number and accuracy of synchronized base stations. Frequent handoffs between synchronous and asynchronous base stations, however, enable the MSMs to maintain system timing. Thus, without including either in-building or tunnel coverage, which the European’s argued that asynchronous WCDMA can best deliver, even a partially synchronized system remains sufficiently robust to provide effective soft handovers.
The on-the-hook-for-3G-spectrum European operators have the discretionary power to implement some form of synchronized network design. The advantages for choosing Qualcomm’s synchronous solutions include superior technical and economic performance in voice networks, a near-term addition of 1xEV-DO to provide high data rates, and a universal world phone. This should be a no-regrets decision, but if Europe stonewalls, at least, any future UMTS carriers in Asia, free of European politics and hubris, will have no regrets when choosing a synchronous solution to the asynchrony problem.
Now given Qualcomm’s ZIF-enabled proficiency with multi-mode chips, is there a Trojan horse inside its 6xxx chipsets? Does its asynchronous UMTS also include integrated circuits for synchronicity, just waiting to be turned on by a BREW phone call? Is that secret seductively awaiting troubled operators at a weak moment when they need to maximize profits? Can Qualcomm, while doing its interoperability testing using its own MSM5200 or MSM6200 chipsets, not demonstrate excellence to prove the excellence of synchronized WCDMA solutions? |
