In terms of making these cells smaller, would these be public cells, or would people start carrying around their own mini cell towers with them?
Small cell towers known as picocells and microcells are public. They're part of the architecture laid out by the cellular operators. Femtocells are desktop cell phone towers that a company or an individual would buy this for better cell phone coverage in their office or home [they can enhance a signal within a radius of 10 to 200 meters around the device].
In the case of an emergency like the earthquake or hurricane, would you have been more likely to get a signal if you had your own femtocell?
There are a number of challenges with femtocells. Earlier femtocells didn't manage their interference with one another well and weren't able to hand off to the larger cell towers outdoors. So a femtocell might have helped if you made a call indoors and stayed indoors. The question is how to do dynamic optimization of all the picocells, microcells and femtocells so that you're adapting not just to traffic changes under normal conditions but in particular under emergency situations. If an earthquake or the 9/11 attacks take out a number of towers the way the cell phone towers are configured now, service providers have to manually reroute cellular traffic; they can't adapt to dynamic changes in traffic.
The ability to do this dynamically is referred to as self-organized networks, or SONs. Combining that capability with small cells has tremendous potential to increase both the capacity and the ability to adapt to changes that are unpredictable.
How would self-organized networks, or SONs, be implemented into a cellular network?
It's a software network management tool for base stations. It can sit in the cloud, or it can sit on the mobile gateway that's managing all of the cell phone base stations.
What are some other options for improving wireless network performance?
Obviously, if capacity is a problem, throwing more spectrum at it is a way to increase capacity. Given that there just isn't a lot of spectrum available, another option is cognitive radio, which can take spectrum that's already occupied and use it more intelligently. In this way users sharing the spectrum, as they do with Wi-Fi, don't interfere with the licensed users of the spectrum, such as broadcasters. You would have to have a cell phone equipped with cognitive radio software that can find spectrum holes. For example, if you're trying to make a mobile phone call and can't get a signal, your phone would switch into cognitive radio mode and scan not just the cell phone bands but other bands for some piece of spectrum it could use.
That's a promising technology, but I'm skeptical that in the near term it will solve the capacity problem or even be viable politically. Spectrum is so precious to broadcasters and other spectrum bandwidth owners that they don't trust that secondary users can share their space without interfering.
Certainly, this distrust has been a big part of the controversy over the use of so-called white spaces between licensed bandwidth by wireless gadgets. What impact will the expansion of 4G networks have on these emergency bottlenecks that we currently experience?
Newer, 4G networks are optimized for data in the sense that they use the same network that the Internet is using. Telephone operators might just leave voice traffic on their 3G networks, which are still fairly new and not going away anytime soon. Voice requires a lot less bandwidth than data, which would then be pushed to 4G networks. In that case, things will be better in an earthquake because when you walk outside and try to make a phone call you won't be competing with people using their smartphones to send data across the same network.