WARNING

This is a test of the U.S. Emergency Alert System. If this were a real emergency, you'd be notified via your handset and instructed to take appropriate precautions. That is, if your handset is turned on, or if you're in your home area, or you have the right digital phone, or ….

It was after supper in a small town in Middle America. Just as everyone was relaxing with a favorite TV show, the screens blackened and were filled with large letters announcing “Emergency Alert: Please stand by for an audio message.” The shocked audience was then treated to a 30-second pitch for Bud's Realty, telling them to call Bud if they were in the market to buy or sell a house. It assured them that he had the best deals in town. Then, their TVs returned to normal programming, leaving the watchers scratching their heads.

The town had implemented a secret phone number from the mayor's office to the local cable TV company that would automatically put emergency messages on the air. Nobody but the mayor and his close assistants knew this number. The real-estate company had not discovered this number; it was just “war dialing” through an entire block of numbers, playing its recorded sales pitch to anyone who cared to listen. The secret number just happened to be in the middle of one of these blocks.

Some people have proposed that emergency messages be extended to wireless phones through a technology known as broadcast SMS. Warnings of impending hurricanes or tornadoes or instructions about what to do following an earthquake could be sent directly to the screens of wireless devices. This seems like a great idea. But what are the advantages of doing this, and are there any unintended consequences?

Short-Message Service

SMS normally operates in a point-to-point mode, where a message is sent from one point, such as an e-mail address, to another point, such as a mobile phone. This is not efficient for messages that need to get to a lot of people. For one thing, you cannot determine from a phone number where a handset is without doing an HLR query. Sending SMS messages to all people who subscribe within the geographical area affected by the emergency would reach some people roaming elsewhere and would miss everyone who is currently roaming in that area but subscribes elsewhere.

Broadcast SMS (also known as point to multipoint) is more efficient and more appropriate for emergency messages. One message can be sent out from each cell in the affected area and can reach every compatible phone in the area. Unfortunately, although better, it's not perfect, and there are a number of limitations that seriously impede its use.

Broadcast SMS is an inherently unreliable service; there is no way of knowing whether a mobile phone received the message. A phone that is temporarily unavailable when the message is sent out would never know that the broadcast had been sent; consequently, messages would have to be broadcast several times to ensure wide coverage. Messages also may have to be sent in a point-to-point fashion to all mobiles that are in conversation. Fixing this problem by making broadcast SMS reliable would reduce its efficiency because every phone receiving the message would have to send an acknowledgement.

Broadcast SMS also is not available in many phones. For example, no analog cellular phones (or digital phones operating in analog mode) and only the more recent versions of digital phones support broadcast SMS. CDMA standards first supported this in IS-637, which was published at the end of 1995, but TDMA phones did not support it until TIA/EIA-136 (www.tiaonline.org) Revision A (www.eia.org) was published at the end of 1999. Even when standards support the capability, there is no mandate that forces phone manufacturers to implement it, nor is there one that forces carriers to purchase the capability for their base stations.

Another limitation of SMS is the length of the message that can be transmitted. This often is kept quite short (e.g. 160 characters) because of various network and radio interface limitations; it may not be enough to get the full emergency announcement.

The Emergency Alert System

The U.S. Emergency Alert System (EAS) currently is oriented to audio alerts broadcast on radio and TV stations. This has been criticized by some deaf and hard-of-hearing advocates, and it may be extended to provide a text version of every message. Not only would this be useful for the disabled (allowing the text to be broadcast by TV stations simultaneously with the audio), but it would be necessary to get broadcast SMS working for emergency alerts. The FCC (www.fcc.gov) released a notice of proposed rulemaking (EB Docket Number 01-66) on March 20, 2001, that includes a proposal to standardize text-message transmission.

Connecting to EAS would require that wireless carriers install equipment to receive EAS messages and be able to rebroadcast them by stripping off the relevant information and eliminating messages entirely that should not be broadcast (e.g. test messages). This would have to be done as a public service; it is unlikely that they will receive any funding or income from this.

Carriers that want to implement this service can receive some guidance from TIA/EIA/TSB-114, which was published in December 1999. This does not describe the exact process for taking an EAS message and broadcasting it, but it provides a list of requirements for this process, as well as a list of its limitations.

Is It Worth It?

It's certain that broadcast SMS would provide emergency information to some people who would otherwise not see it. In an increasingly mobile society, people are spending less time in front of the TV, less time listening to the radio and more time on their wireless phones.

But why is there an emphasis on providing these alerts to wireless phones and not to people sitting in front of computer screens? Computers are increasingly connected to the Internet through permanent DSL or cable-modem connections and are more capable of supporting EAS capabilities, whether the message is sent as a recorded message or text. And, even the longest EAS text messages could be displayed on all but the most primitive computer screens.

One of the reasons that the use of wireless networks for the EAS has been pushed strongly may be financial. Its most avid promoter, Douglas Weiser, was awarded U.S. patent 6112075 (method of communicating emergency warnings through an existing cellular communications network, and system for communicating such warnings) for the concept in August 2000 and stands to benefit financially if this concept is implemented.

It would make sense to have a realistic field trial of a wireless EAS before implementing it widely. There are many questions that need to be answered. Will users pay attention to these messages? How many phones actually support broadcast SMS? If people are making a call, will they be aware that an important text message has been received? Are there any security holes that would allow false-alarm messages to be broadcast? What are the costs of providing this service? Should all EAS messages be broadcast, or just the most important?

If the service proves to be useful, it may become obvious during the trial that default handling of broadcast SMS is not suitable for emergencies; further modifications to radio or network standards, or more detailed guidelines for the process of converting the EAS protocol into SMS messages may be needed. The originators of the messages also may need to be aware that they have a new audience and make adjustments to optimize their transmissions for this new channel.

Until these questions can be answered, the wireless EAS will remain on paper — a patent, one whole standard, bits and pieces of others and a bunch of articles, including this one.

---

Crowe (crowed@cnp-wireless.com) is a wireless-standards consultant and editor of Cellular Networking Perspectives, a wireless-standards and -technology bulletin.