 |
|
|
When emergencies happen one of the first thing many people do is reach for their cellular phone. They may want to report a crime in progress, call for help, or just tell someone that they're okay. Unfortunately, when rare headline-making emergencies occur, calls do not always go through due to system overload. This was a problem during the short intense emergency at Ottawa's Dawson College as well as during the much longer emergency in New Orleans due to Hurricane Katrina.
Wireless systems are assigned limited frequencies by the Canadian government, and the amount of frequency that can be used in one place is also constrained by neighbouring cells needing to use part of the same block and the amount of radio equipment installed in the cellsite. With analog and TDMA/GSM cellular systems the number of users who can be supported in a single cellsite is a simple function of the number of transceivers installed but with CDMA it is more complex.
No matter what the technology, cellsite capacity is carefully engineered to ensure that the undesirable tone known as “Fast Busy” (2 beeps per second versus 1 for normal busy) is rarely heard, even during the busiest times of the day. This tone indicates a lack of resources for the call, usually radio capacity, although sometimes it reflects the lack of a connection back to the main switch or other network overload or failure situations.
The problem with engineering for busy hour is that according to several radio capacity experts I talked to, including Seattle-based consultant Elliott Drucker, emergencies can produce 5-10 times more traffic than normal. They are supported by the experience of Telus cited in the Montreal Gazette: Eleven times normal traffic during the recent Dawson College shootings. It is impossible for cellular carriers to have this much extra capacity in place. It is not just that this would increase the cost of cellphone communications several fold, something that consumers would not tolerate, but there simply is not enough frequency available in many locations, particularly in urban areas. Finally, while some emergency planning is possible, the location and intensity of the next emergency is, by its very nature, unknown. Every cellsite in Canada would have to be made ten times larger and even then it wouldn't always be enough.
Cellular system capacity is not the only limitation. When dialing 9-1-1 the limited number of trunks to the call answering center can also prevent emergency calls going through. A flood of emergency calls can overwhelm their capacity to respond, resulting in calls being blocked by a lack of trunks into the emergency center or people to answer them. More radio channels would only make this problem worse.
One way to relieve pressure on 9-1-1 resources is to bring 3-1-1 on stream. The CRTC approved the use of this short code in 2004 for non-emergency municipal government services. Several Canadian municipalities have implemented it, including the cities of Calgary, Toronto, Halifax, Halton, Gatineau and Montreal who originally petitioned for it. Important, but non-critical calls, such as reporting abandoned cars, burst water pipes and traffic light outages, can be placed to 3-1-1, relieving the burden on 9-1-1 operators, allowing them to deal with truly life threatening situations.
Cellular service often proves more robust than landline phones which rely heavily on overhead wires to customers. However, severe weather emergencies, such as the Montreal ice storm of 1998, can bring down cables carrying electricity, or even towers due to ice accumulation or violent winds. A cellsite may have backup power from generators or batteries, but if its ability to communicate back to the central office is compromised, its antenna system is damaged, if there is a fire, or the shelter is flooded, it may be knocked out of service. Surrounding cellsites can be adjusted to make up for the loss, but the overall capacity of the system will still be lower, just when it is needed most.
According to Richard Boyer, Director at the 9-1-1 Emergency Center in Montreal, their most important response to overload during an emergency is to ask Bell Canada to increase capacity in the critical area. One method to do this is to do the opposite of what might be expected - reduce transmit power in an overloaded cellsite causing it to shrink while increasing power in its neighbours, allowing them to pick up traffic around the edges of the overwhelmed cell.
 |
One way to increase capacity is to bring in a mobile cellsite, often called a “Cellsite On Wheels” (COW) or “Cellsite On Light Truck” (COLT). You will sometimes see these parked outside a major convention centre to add capacity. Telus, for example, has light trucks with cellular communications and satellite backhaul, which can provide self-contained emergency communications. Their main limitation is the time it takes to bring the COW to pasture. For short lived emergencies, of which the Dawson shooting was a good example, a truck might only arrive after the critical emergency period is over. COWs and COLTs are more useful during weather emergencies, forest fire evacuations and other longer lasting situations.
Increasing the number of mobile cellsites in Canada would require government funding as there are only limited commercial uses for them. This may be justifiable because they provide a benefit to all Canadians. The trucks often come with a supply of handsets so victims of a disaster who have lost regular phone service may benefit from them even if they do not normally subscribe to cellular phone service. Canadians also benefit when they increase the ability of emergency workers to communicate.
Emergency workers usually have their own private radio communication networks, but cellular phones have also become important supplements for them. Wireless Priority Service (WPS), a topic I wrote about in Wireless Telecom in 2002, is now available in Canada, which can make it more likely that cellular calls for a variety of people in important emergency response positions can go through. This includes not only emergency workers, but also political executives (such as mayors and senior members of the government), military officials, and even workers cleaning up the mess afterwards.
WPS is designed to be as unobtrusive as possible. The service never preempts established calls. Once a call from a non-priority user is connected, it will stay connected, and will not be disconnected by WPS. Users with priority do not get this preference on every call, but have to dial a special prefix. Since usage can be monitored, using priority for unimportant calls is unlikely, making it likely that it is only used when it is critically important that their call goes through.
WPS works by queuing mobiles when no radio channels are available. When a channel comes free it is assigned to the mobile in the highest priority queue that has been waiting the longest. This not only increases the likelihood of emergency workers being able to make a call but it also reduces the amount of time they spend redialing.
WPS not only provides an adjunct to private radio systems used by police, fire and ambulance services, but also allows communication between users of different systems, and allows people who would not want to carry two radios to have access to an emergency communications system in their regular cellular phone. A good example is the mayor of a small town who might play a critical coordinating role during an emergency, but is unlikely to want to carry a public safety radio around with him every day of the year, but is almost certainly never more than a few inches away from a cellular phone.
Some people might question why emergency workers might rely on cellphones for emergency communications when they have their own systems, but Drucker makes the provocative suggestion that they should become more reliant on cellular technology. “Proven commercial air interface technologies (e.g. cdma2000 and W-CDMA) are the way to go for public safety.” In other words, he suggests that public safety systems be built upon the same radio communications technologies that consumers use, although they could be located in different frequencies. One advantage of this approach would be that dual-mode radios would be more feasible.
Several observers noted that during the Dawson college shooting text messages went through even when voice calls did not. Text is a better form of communications for short messages (e.g. “Im ok”) because it uses no radio resources when the message is being composed, and only a tiny amount when the message is being transmitted. Even better, the messages are non-real time, so the phone can try several times to contact the system, and the system can queue the messages up when it is too busy to deliver them all at once.
Simple mathematics shows how dramatic the difference is. The largest text message allowed by most systems is 160 characters, about 1300 bits (and most messages are much shorter because most people don't want to type that much, particularly during times of stress). By contrast, voice communications uses digital voice coders that communicate at rates between 8 and 13 thousand bits per second. So even a big text message represents only a fraction of a second of a voice call.
A Montreal-based cellular radio engineer told me that voice calls can also drastically reduce available capacity because Canadians are not used to their phone systems taking a long time to put calls through. When someone makes a call in an emergency and ringback doesn't happen within a few seconds, they are likely to hang up and try again. This only makes a bad situation worse as all the capacity used by such abandoned calls is wasted.
Emergency Alerts are text messages that can be sent out to every phone in a geographic area. They are ideally suited for emergencies because they are geographic and, if done right, use even less resources than normal text messaging, because one message from each cellsite can go to most mobiles in that cellsite.
They don't allow people in the emergency to initiate communication, but would allow them to receive useful information.
The problem, as I described in a recent Wireless Telecom article, is that although pieces exist, there is no comprehensive solution that integrates the organizations that generate information about emergencies, cellular phone networks, radio interfaces and mobile capabilities. The United States is currently leading investigations into this area, and it is likely that technological answers will emerge over the next few years.
Assuming that broadcast emergency alerts were a reality, it is easy to see how they could allow emergency workers to communicate with people with cellphones. Not only could they be used for pending severe weather events, such as major storms and tornadoes, but they could broadcast a message during an emergency asking everyone to “Leave the mall by the south doors”, for example. In a weather emergency they could remind people about live wires down on a street. During a chemical spill they could ask everyone to stay indoors with all windows and doors closed until evacuated. In many cases, even if they do not reach everyone, they can still be useful, as people who receive the messages can inform others in the vicinity. It is obviously critically important that this system be designed securely, because people with malicious intent could obviously use it to cause chaos, just like yelling “Fire!” in a dark, crowded theatre.
Reverse 911 uses a similar concept, although it is more oriented to landline phones. An emergency system prepares a short recording and then sends it to an auto-dialing system that rings each phone in a geographic area and plays them the recording. For cellular systems this needs to be based not on a static list of phone numbers, but on real-time knowledge of the wireless phones actually in a specific geographic area. There is no way to truly broadcast voice messages on most systems, so the capacity used by this system is much higher. Reaching every phone in a large area may be impossible within any reasonable amount of time.
According to Richard Boyer, of the 9-1-1 Emergency Centre in Montreal, they had tried this system, but found that it tended to create more panic. Being automated, there was no way to answer the questions of people who received the messages. Boyer noted that principles of emergency communication are to avoid adding more communications to an overloaded system, and to be very careful that communications does not add to feelings of insecurity.
The future of wireless communications is that data will become more important, but the two main forms of communication will continue to be talking and writing. Voice may be treated by carriers and technologists as just another form of data, but consumers will still be picking up their phones and talking when they have something important to say (and sometimes when they don't!). Text messaging may become more sophisticated, but with a small device people will probably still limit themselves to short messages made shorter by using the ascii slang that teenagers everywhere have helped develop.
IP (Internet Protocol) communications are being enthusiastically embraced by carriers because it promises them one network to carry voice, data and signaling (internal communications) whereas traditional telecom systems used up to three different networks.
IP, however, is not without its flaws. Without considering the problems of spamming, phishing, denial of service attacks and other security issues, IP systems are not as rigidly structured making it harder to avoid overload (although raw capacities are generally much higher). While traditional telephony systems divide capacity into a fixed number of slots, of which one slot can carry one call, IP systems (like some more modern radio systems) share their capacity among all users. This makes it easier to mix a wider variety of traffic together, but it makes system engineering much more difficult. Engineering is based on statistical assumptions about traffic patterns, assumptions that are likely to be invalid during emergency situations.
The worst possible situation could be that an IP system allowed more users to access the system than it should, resulting in low Voice-over-IP (VoIP) call quality, meaning that few could communicate even though the system was happily carrying packets around, but dropping some and delivering others so late, that real human communication becomes impossible.
Clearly, emergency overload situations are going to have to be examined as new generation systems begin to make it into the field. System performance during unusual situations needs to be superior to justify their existence, just as their performance during usual situations has to be.
It is always useful to think about what you might do in an emergency situation before you get there and the adrenaline makes rational thinking difficult. One of the reasons that emergency workers are so much calmer and more focused than the rest of us during an emergency is because of extensive training through simulation as well as experience in many real life emergencies. They've been through the stress and know how to control it. For others, even a little bit of forethought can help you be part of the solution, and not part of the problem.
The aftermath of every emergency should be taken as an opportunity for future improvements. No systems are perfect, but all can be improved. There is no point in planning for a repeat of the exact same emergency, because it almost certainly won't happen. But problems with communications can indicate places where common sense improvements can be made. Perhaps this can result in slightly different approaches to cell engineering, an educational campaign to consumers, or lobbying the government for more funding for emergency preparation and reaction services. Communications between cellphone company employees and emergency workers may not have been as smooth as needed. Perhaps the backup power system didn't work as well as hoped, or you need better diversity of backhaul, perhaps providing both microwave and underground fiber connections to some cellsites.
It is also important to think about what did work well. Make sure that the reaction to an emergency that might have hit the headlines doesn't persuade you to make changes that will just cause different things to go wrong next time. When doing a postmortem a variety of people in the company should be involved, and selected outsiders as well. Think, discuss, think again, and then make a proposal to implement changes.
If the conclusions are that nothing could have been done differently, that's a reasonable answer. But it is important to act responsibly and do a thorough analysis, even if you think that nothing could have been done better (which is unlikely imho).
Blame games should be avoided, but constructive changes can sometimes come from destructive events. Above all, be honest with the public. If your system did not work flawlessly explain why and explain where improvements can be made in the future and where the limitations will remain.
| Topic | Website |
| 3-1-1 Service | crtc.gc.ca/PartVII/eng/2003/8665/c126_200315699.htm |
| 9-1-1 Service | nena.org (US based, but with chapters in Canada) |
| Emergency Alert Service (EAS) | fcc.gov/eb/eas (USA) |
| Emergency Preparedness | safecanada.ca/topic_e.asp?category=4 |
| Wireless Priority Service (WPS) | strategis.ic.gc.ca/epic/internet/inet-tdu.nsf/en/h_wj00016e.html |
David Crowe is a wireless standards, technology and numbering resource consultant based in Calgary. He can be reached at David.Crowe@cnp-wireless.com.
© Copyright Monday, November 27, 2006
: Cellular Networking Perspectives Ltd.