The world has become a place where data is expected everywhere. As time progresses, that data is growing at a rate that mobile networks are already having trouble supporting. A good example of this might be a 45-second video of your kids at Disneyland taken with a smartphone. At 1080 resolution (the standard resolution now on modern smartphone video cameras), the video would be upwards of 100 megabytes. Want to share this immediately with your friends on Facebook? The cellular network burden (from a file size perspective) of that video is approximately the equivalent of 1,700 average emails. Not only are we producing more data, 3G networks have one million new devices connected to them every day.
Qualcomm held their analysts meetings in San Diego recently and we learned how they envision solutions for a predicted 1000x increase in cellular data traffic over the next ten years. The first thing to understand is the formula for wireless capacity.
Qualcomm is addressing all components of this formula.
Let’s first look at the W, which is Spectrum. In order to address the massive 1000x increase in data capacity needed in ten years, licensed spectrums may be pushed to their limits. In betweenlicensed and unlicensed spectrum is a huge amount of capacity (Wi-Fi is an example of an unlicensed spectrum). Unfortunately this spectrum is most often exclusively reserved for government/military. However there is a concept called “authorized shared access” (ASA),in which such entities will allow public use of reserved spectrum as long as there is a command and control system which can quickly yield spectrum to the incumbent. At such a moment all civilian users would need to be seamlessly transferred to the licensed commercial spectrums. Qualcomm is working on such a control system.
Qualcomm is also looking to increase the efficiency of spectrum. Multicasting is a technology they are developing which allows content to be broadcast. Instead of the current situation where users watching identical content each represent a unique data stream, with multicasting, these people would be latching onto the same data steam, with the benefit of an actual throughput gain multiple. So for example, if someone down the street is watching live evening news on his phone at the bus stop and someone else in the same cell is watching the same thing in the park, the network burden would only be a single data stream. The same could be applied to thousands of users who get an instant replay in a sports stadium or a 2 AM OS update. They all latch onto a single data channel instead of individual transmissions as is the status quo. This is Multicasting.
Let’s next look at the n in the formula, which represents the number of antennas and their coverage cells. The most prevalent antennas are “large” or “macro” cells provided by big antennas. Big antennas are the ones that are controversial in some communities for aesthetic and/or health-effect fear reasons.
As the number of subscribers grows these macro cells hit their capacity (“maxed out”) and operators are now adding overlapping macro cells to increase capacity.
Small pico cells
The second type of cellular base station is a “micro” cell or “pico” cell. Currently the realm of Freescale Semiconductor and Texas Instruments, pico cells are deployed in malls or train stations and could be deployed on business and condo rooftops without the level of regulatory hurdles that macro cells might require.
Pico cells can augment hard to reach locations or macro coverage dead zones. The can also be used to “fill in” areas that are adjacent to regulated “exclusion zones.”
Both macro cells and pico cells are generally “outside in” solutions; where the signal originates from outside your home or place of business. Outside in solutions may not be enough to meet the demands placed on mobile networks over the next decade. Qualcomm may have a solution with what is being called an “inside out” cell augmentation strategy. The technology behind this is the “femtocell”
Femtocells are very small cellular base stations connected to an ISP (Cable, DSL, etc.) and are currently available for customers who do not get good coverage in their home or small business. Femtocells are not much larger than a deck of cards. The neat little secret is that their signal leaks out into the street.
Network operators could fill in their dead zones and increase capacity in high traffic areas by using purposely placed or signal leakage of femtocells to augment capacity. Users in certain areas might be offered discounted broadband as an incentive to provide a mini coverage zone.
Of course large agreements with ISP may be needed as well as careful review of privacy, and the effects of cellular traffic on the home/small business network performance in question.
Something else to note is that Qualcomm often uses the term “small cell” which typically encompasses micro, pico, and femtocells, instead of femtocell. This could indicate they plan to compete with Freescale and Texas Instruments on other small cell solutions.
Signal to noise
The last critical part of this equation is the signal to noise ratio (SNL). If a smartphone is to perform seamlessly hopping in and out of competing macro, pico, and femtocell coverage, interference and handoffs must be managed with the utmost precision and intelligence. Although Qualcomm didn’t give specific details, they said they are working on technologies that will be able to address this issue.
All in all we were impressed with what Qualcomm is doing on many levels and so are their investors and customers. Qualcomm reaped over $6 billion in net income during fiscal 2012.
Mobile data is already being challenged from our experience notwithstanding the massive increases in demand on the horizon. Our 3G and 4G connections are not consistent, have widely varying performance levels, and are often a letdown depending on our location. Something needs to be done; Qualcomm is not sitting by idly on their hands. We hope that many companies will help address this challenge and little help from Uncle Sam and his overseas colleagues wouldn’t hurt either.