Breathrough Ideas (continued)
Unique Competitive Advantage (continued)
5. RF System Design. An RF system design is the architectural landscape plan for laying out an efficient RAN system for a specific cell geometry, topography, and demography that fully exploits the design features permitted by its air interface. The art is in the placement of antennae that simultaneously takes into account the properties of propagating radio waves, topography of the land (both natural and man made), the demography of callers and their usage patterns, and, sometimes, the amoebic-hexagonal geometries of cells (if, and only if, this is necessary to avoid adjacent cell interference).
CDMA RANs link mobile terminals to CDMA infrastructure equipment, which, in turn, can be linked either to a Mobile Telephone Switching Office (MTSO) or Public Switched Telephone Network (PSTN). CDMA RAN infrastructure requires two pieces of infrastructure equipment—the Base Station Transceiver Subsystem (BTS) and the Base Station Controller (BSC). The BTS includes a three-sided antenna for transmitting and receiving signals and performs the CDMA processing of all signals. The tri-antenna propagates alpha, beta, and gamma sectors, which can be further subdivided by adding additional equipment to increase capacity, with each new sector operating as an independent BTS. The BSC performs vocoding of the voice signal, routes calls to the MTSO, handles call control processes, maintains a database of subscribers, and maintains records of calls for billing. Qualcomm provides the operating system software required to operate its RANs.
The mobile-to-land call path proceeds from mobile phone through BTS, BSC, MTSO, PTSN, CO (local Central Office), to land line phone. The land-to-mobile proceeds in the reverse order. Within the same network, the call path proceeds from Mobile #1, BTS #1, BSC #1, MTSO, BSC #2, BTS #2, Mobile #2. Across networks, the call path proceeds from Mobile #1 to its BTS #1, BSC #1 to MTSO #1 to PSTN to the second network’s MTSO #2, BSC #2, BTS #2, to reach Mobile #2.
The design of the CDMA RAN is not only made more powerful by universal frequency reuse, but also is simplified by exploiting the soft, elastic, permeable borders between cells, permeable because each cell can run over the top of any other cell. Also, CDMA soft-soft handovers treat sector borders as permeable by establishing simultaneous communication with multiple sector-BTSs. Thus, with spread spectrum, cells and sectors within cells become organic: simultaneously self-governing and nested as communicating parts within the whole RAN system that is integrated by its shared synchronicity.
Rather than being synchronized as a system, GSM RAN designs, in contrast, maintain hard borders between sectors and cells, each remaining independent, with their own separate synchronization. In addition to the problem of having to allocate a new channel, this means hard handovers require a search for a new pilot signal just as fading increases exponentially near the border of a sector or cell. These firm boundaries are required by GSM architecture because it must prevent problems from worst-case interference that result from poor plans or inadvertent overlap from strong signals that create worst-case interference, acting “as-if” they were “adjacent.” As geometrical complexity increases, the areas of overlap proliferate exponentially, creating a need for exponential addition of base stations in GSM systems that were already less spectrally efficient. Converting voice channel to GPRS channel appears to increase the problem further.
Because subdivision is the principal method of increasing capacity within a crowded cell, within GSM RAN designs, this overlapping interference problem compounds. It becomes a significant problem because excessive interference may create feedback and burnout problems within the mobile itself. A second complicating factor for system designers is inherent in the physics of radio waves. The properties of radio waves cause system design to become more complicated when dense usage occurs close to transmitters. When compared to long distance free air propagation, small distance air propagation becomes more complex rapidly along several dimensions.
The need for GSM re-design proves pressing precisely where interference problems are greatest and most likely to occur, making re-design of frequency reuse plans complicated. The design for a problem site, which already may have increased its geometric complexity through earlier transmitter additions, must be precise to gain the intended capacity improvement while simultaneously preventing worst-case adjacent-cell interference. This requires careful analysis, system downtime to upload the new frequency reuse plan, and re-implementation if problems remain. Thus, placing new BTSs becomes a difficult and expensive operating problem. And, lacking an elegant solution, it’s make-do.
As voice/data usage increases, three-sector cells routinely will become six-sector cells. Advantages like coded spread spectrum, power control, synchronized pilot signals, and permeable borders mean that a CDMA picocell can be placed temporarily or permanently anywhere it is needed to increase capacity or coverage at any time. Thus, subsection capacity-addition and its implementation are greatly simplified in CDMA RAN design because of its spread spectrum architecture. If an unexpected problem arises requiring a temporary increase in capacity for a hotspot, a mobile truck-picocell could go to the trouble spot, turn on its microwave link and bloom a picocell, say, within an hour. The system design is versatile and flexible. As another example, by using the tradeoff between capacity and coverage, Nortel offers a CDMA rural cell, called a “boomer” cell, with a radius up to 180 km, 10 times the typical range.
Because reuse is universal, the CDMA RF design solves adding capacity elegantly, without any re-planning, requiring no system downtime to upload a new frequency reuse plan. Because CDMA designs have systemic advantages in spectral efficiency, they require fewer cells both initially and subsequently. This creates significantly lower sunk and operating costs. Thus, spread spectrum’s unique code division mode of boosting access simultaneously increases the CDMA RAN’s degrees of freedom.
Dr. Jacobs likes to say that CDMA is like a U. N. cocktail party where everybody is in the same room, with every pair talking at once. If each pair of guests uses a different language to communicate, it becomes easy for the listener to tune in to what their partner, who speaks their language, is saying. They do not shout in hopes of being heard, thereby interfering with other conversations; nor does each pair have to wait for their turn to use a time-slot when everyone else is silent so that they might be heard. Simply put, Qualcomm’s CSMs and MSMs are highly intelligent systems that efficiently converse with oneanother because they seek, find, hear, and understand each other’s unique language.
Therefore, Viterbi’s breakthrough idea for advanced power control of a uniquely encoded spread spectrum signals created CDMA’s discontinuous advance in wireless architecture, enabling it to leapfrog the functionality of FDMA 1G and TDMA 2G systems.
How This Breakthrough Idea Translates Into Specific Competitive Advantages
This breakthrough idea was transformed through the micro-refinements of superior engineering into many competitive advantages. How do aspects of this IS-95 architecture translate specifically into discrete competitive advantages? Let’s trace the links: (1) Coverage—increased between 1.7 and 3 times that of TDMA by: (a) power control that dynamically expands the network’s coverage, and (b) channelization coding and interleaving that increases coverage for less power per unit covered; (2) Capacity— depending on the vocoder setting, increased from 10 to 20 times that of AMPS and 4 times that of TDMA by: (a) universal frequency reuse, (b) separation by codes, not times/frequencies, (c) power control minimized interference to maximize capacity, and (d) soft hand over also minimized power; (3) Clarity—increased to wire line quality by: (a) rake receiver reduced errors, (b) variable rate vocoder reduced amount of data transmitted per person, which reduced interference, (c) soft handoff reduced power requirements, fading, and interference, (d) optimal power control reduced burst errors, (e) summed wideband signals reduced fading, and (f) encoding and interleaving reduced fading errors; (4) Costs—cost per subscriber steadily declined since 1995 by: (a) increased coverage per BTS, which reduced build-out and management costs, and (b) increased spectral efficiency that permitted increased services to customers; (5) Compatibility—early on CDMA phones were made compatible with analog AMPS and became increasingly enabled to work in multi-band and multi-mode networks; and (6) Customer Satisfaction—increased by: (a) superior voice quality, (b) longer battery life from reduced power consumption, (c) reduced mobile size, (d) eliminated cross-talk because of unique coding, and (e) assured privacy from spread spectrum channelization and other coding.
These significant competitive advantages¾the six C’s of coverage, capacity, costs, clarity, compatibility, and customer satisfaction¾generated lower costs while offering superior technical differentiation. They. The paramount economic principle for a carrier is: maximize the number of customers by providing the best services to satisfy their needs and priorities at the lowest cost-per-bit transmitted. Qualcomm’s conceptual advances in CDMA technology provided the largest delta between what the customer is willing to pay for and an increased supply of capacity at the lowest cost-per-bit, with the lowest sunk and operating costs.
(Some of the above technical information can be found in an Online Introduction to CDMA at the Qualcomm web site.) qualcomm.com
Thus, by deep understanding of the theory of spread spectrum, Qualcomm commercialized CDMA to take advantage of its technical merits. Elegant engineering permitted advanced power control, solving the conundrum of the near-far problem, and introduced universal frequency reuse. An architectural design for ASICs that benefited from spreading the spectrum, power control, rake receivers, and soft handoffs in a synchronized RAN design yielded strong and sustainable competitive advantages that are protected by Qualcomm’s patents. |
