 were caused purely by volume. )
GETTING THROUGH?: Whereas the 9/11 cell phone outage was the result of many factors this week's problems (though brief) were caused purely by volume.
Image: Courtesy of Alec Tabak, via Wikimedia Commons
More In This Article
-
The Best Science Writing Online 2012
Showcasing more than fifty of the most provocative, original, and significant online essays from 2011, The Best Science Writing Online 2012 will change the way...
Read More »
Anyone in the eastern portion of the U.S. this week who was forced to evacuate an office, home or school following Tuesday's magnitude 5.8 earthquake soon noticed that cell phone service was spotty or, in many cases, nonexistent. For New Yorkers herded outside of their skyscrapers and into the streets, it was a communication blackout reminiscent of (although of course not the same as) the 9/11 terrorist attacks. In both situations, mobile phone users were unable to connect to the cell network to communicate with loved ones.
Whereas the 9/11 cell phone outage was the result of many factors—including the downing of cell phone towers—this week's problems (though brief) were caused purely by volume. Countless cell phone users were fighting for limited access, leaving most without service. With Hurricane Irene bearing down on the East Coast, one is left to question the reliability of mobile phones in the face of serious emergencies.
Scientific American spoke with Andrea Goldsmith, a Stanford University electrical engineering professor and researcher at the school's Wireless Systems Lab, about why mobile phone users are often unable to connect during emergencies, as well as options for improving cell network performance when it matters most.
[An edited transcript of the interview follows.]
What is happening when, during an emergency, you try to use your mobile phone and can't get a signal?
There's a finite capacity that the cell phone system can support at any given base station. Say it's 10 users or 100 users talking on their phones. When the 101st user calls in, there's no channel available for that user to occupy. That's a little bit of a simplistic description, but that's effectively what happens. When there's an event like an earthquake that causes people to want to make more phone calls or want to access the Internet more, the systems just become overloaded.
How have cell-phone services changed since 9/11?
Cell phones are coming up now on their fourth generation of technology, so we've had two evolutions of cell phone technology since 2001. Most of the work in between evolutions has been aimed at getting higher data rates because people now also use their phones for accessing the Internet and for exchanging videos and pictures. There have been improvements in capacity, but not enough to keep up with just a tsunami of traffic across the wireless networks.
With the introduction of smartphones, many mobile phone users have the ability to connect over the cell network or via Wi-Fi. Shouldn't this provide additional capacity during emergencies that wasn't there years ago?
Cell networks and Wi-Fi both have limitations. Cell phone networks have a lot less bandwidth than Wi-Fi, but at least the cell phone service providers own the part of the spectrum over which their customers communicate. This enables the cell providers to guarantee a certain quality of service. Still, it's not easy for them to add capacity because there isn't a lot of spectrum available for them to purchase.
Wi-Fi has access to a large amount of bandwidth for connecting to the Internet but doesn't have exclusive rights to that space. Smartphones have to share the spectrum with a large number of different wireless devices, including baby monitors, cordless phones and Bluetooth gadgets. With Wi-Fi, when there are a lot of people trying to get on the network, nobody can get on.
So your ability to connect via the cellular network or Wi-Fi really depends on the people around you?
Yes. If you went into a Starbucks in Times Square, say, probably you wouldn't be able to get on the Internet via Wi-Fi. And if you were in the part of the city where the cell phone towers were not overloaded, you probably got a connection just fine.
What can be done to improve cell networks so they are less likely to be overloaded during emergencies?
Typically you design any system for average use. The problem is, if it's designed for average use and there's an event that's statistically unusual, typically you're going to overload the capacity of the system. One option is to just build in extra capacity, and one way of doing that is to make the cells smaller. Right now the coverage of the base station that your phone talks to is fairly large. These cell towers put out about 10 watts of power, so they can, for example, cover a fairly large region of New York City. If a large cell can support something like 1,000 users in Midtown and you make that cell smaller, so that it covers a couple of blocks, you can now support 1,000 users within a much smaller region.
See what we're tweeting about
11 Comments
Add CommentRealistically, how are these technologies more likely to be used: to make cell coverage more robust in a usage spike, or to allow providers to offer just-adequate service at reduced cost?
Reply | Report Abuse | Link to thisYou know, it's funny, when I was a kid, all we had were phones that were hardwired into the wall, and barring a car hitting a pole down the road, we had dialtone - ESPECIALLY in an emergency. Maybe the old folks weren't so far off the mark - "Let's hear it for progress."
Reply | Report Abuse | Link to thisthe land line can overloaded also if too many people call at once try calling oversea at chrisimas time
Reply | Report Abuse | Link to thisAn idea that popped up years ago still has great promise for emergency communication: in the absence (or overloading) of working cell sites, phones could "fall back" to a self-organizing mesh network for text-only messaging. Messages could pass from phone to phone (silently and securely) and eventually reach their target, even if the network of cell towers was down.
Reply | Report Abuse | Link to thisThere will be technical barriers doing this: cell phones can't have enough power for a mesh network. But it is not insurmountable. This is an interesting idea. I wonder why it didn't take off.
Reply | Report Abuse | Link to thisIt has been awhile since I have left the technical field but the sentence saying data requires more bandwidth than voice is interesting. Since a data stream is a constant value and one bandwidth, I thought a voice bandwidth would require a wider, varying bandwidth? Am I wrong?
Reply | Report Abuse | Link to thisA key question: In a site designed for 100 users, what happens when the 101th user signs on? Will the system collapse, or will it service the 101th, 102th... albeit at a lower performance? Experts call the latter option "Graceful degradation".
Reply | Report Abuse | Link to thisCommercial mobile systems now mostly use TDMA or CDMA technologies; it is well known that Frequency Hopping Multiple Access (FHMA) can adapt about three times more users for the same spectrum allocation, and offers graceful degradation. Until now, FHMA was not used in cellular because of technical limitations of a small, little known, but essential component: the Phase Lock Loop (PLL). Every mobile phone has several PLLs; base stations have a multitude.
Until now, PLLs used the same basic architecture devised more than 60 years ago, which uses just a small part of the information available in the PLL - the unibit PLL. A new technology, the multibit PLL, makes use of more info in the PLL to dramatically improve performance: fast settling time, low phase noise.
The new PLL can improve wireless three ways:
* Increase the cell size, due to lower phase noise. Most of the users are located at the periphery of cells, so just a small increase in cell size can service more users, reducing No Service areas.
* Improve the effective bandwidth of each channel
* Allow the implementation of FHMA at reasonable costs.
All the above can be implemented as an improvement to existing infrastructure; new and old PLLs can coexist as the Phone Networks are gradually upgraded.
Eng. Marc Zuta.
marc@patent4u.co.il
Having been born in the ENIAC era, this is not my best subject so I have a question. Can some bandwidth be dedicated to 9-1-1 calls only. Or do they go through no matter what already?
Reply | Report Abuse | Link to thisIn 1970 it took my wife 4-hours to get through to her mother in Texas during Hurricane Celia. We were living in Monterey, CA. During her attempts to call we had a 5.4 earthquake.
Reply | Report Abuse | Link to thisLandlines are not that secure either. Usually the phone lines go down with the power lines.
In Israel, where there have been frequent terrorist attacks, we know that cell phone networks reduce or cancel service at the time of an attack (or other widespread emergency).
Reply | Report Abuse | Link to thisThe downside is that people in the area of the event cannot communicate with loved ones; the upside is that the emergency services can make use of the the networks. The periods are short but before the system was implemented the types of problems described in the article were widespread.
First on all the cell phone industry is just there to make money, period. Cheaply made equipment and minimal infrastructure. Most of what you hear about it is hype. During 9/11 only about one out of every eight cells phones across the river in NJ got service. The system couldn't take the load. Their immediate response to complaints? "It was because the towers went down." (What?) The system hasn't been improved much since it still hasn't kept pace with the number of users being added on every year.
Reply | Report Abuse | Link to thisHaving 911 calls assigned higher priority over regular calls is a good idea but that can be overloaded/misused too. How do you explain to someone looking for a relative that they are getting kicked off because an emergency call must go through?
Part of the article was amusing because it talked about smaller "cells". This was one of the first concepts of cell phones, each cell covering about four city blocks with a central microwatt transmitter with overlap of surrounding cells. Never happened probably because it was too expensive (less profit).
Most importantly, there was one major factor that the article didn't address; the user. One of the most reliable communication forms during wide area disasters is, you'll probably never guess, old fashioned ham radio. These people are trained to use the radio, many even volunteer with their own time and equipment (including generators) to pass emergency messages between government and private agencies during a disaster. These messages are prioritized into emergency (which come first), logistics and then "health and welfare". If an emergency "network" gets on a frequency, others will move off or just shut down. Can you see the average cell phone user doing this? You can't even get some to turn them off in theaters, much less not use them during a movie (or driving). How much bandwidth and backup battery power was used up by chit-chat during the hurricane? Educating the cell phone user is the easiest and cheapest fix to the problem. BTW, the whole cell concept was actually invented by a group of ham radio operators trying to improve their repeater coverage in Chicago.
I would not rely on my cell phone in a wide area disaster except as a "nightlight" so I could see in the dark. Anything else, depends....