After reviewing the excellent and insightful comments received by the NTIA All-Hazards roundtable, it should be clear that a comprehensive approach to improved emergency alert warning must deal with technical as well as administrative, legal, and economic issues. Within the technical area, consideration must be given to data communications in many different stages of the alerting process. We focus here on technical issues related to the final stage of the alerting process, i.e. the delivery of alert messages to the public.
0.1 Early Use of PSTN
A wide variety of approaches have been suggested and discussed for delivery of emergency alert messages. Of these, it seems that solutions based on the public switched telephone network (PSTN) show the most near term promise for reliable delivery of messages to the greatest percentage of the public in a given area. These systems make use of existing infrastructure, utilize equipment familiar to the public, and have a very high “market penetration” relative to other technologies.
0.2 Evolving to Wireless Broadcast
While PSTN based solutions offer great promise for the present, they have disadvantages for the future. First, they do not effectively reach the mobile population. This will become increasingly important as personal mobility increases and as the public depends more and more on wireless communications. Second, PSTN approaches do not scale well for delivery of messages to large numbers of people. For some types of emergencies, seconds count and the timelines associated with effective alerting could not sustain the several minutes that might be required to alert a large population using PSTN-based methods. One example of this situation would be alerting for earthquakes. Wireless broadcast methods solve all of these problems, offering the ability to reach mobile population, alerting large numbers of people with one message, and providing near instantaneous message delivery.
0.3 Not All Wireless Broadcast Systems are Created Equal
In considering the use of wireless broadcast systems for delivery of emergency messages, one must recognize that there are various advantages and disadvantages for each of the systems. The following issues should be considered. First, we should recognize that there is a wide range in the size of the coverage “footprints.” Continental coverage systems, such as the emerging direct broadcast TV and CD-quality sound systems may have too large a footprint to be effective for some types of emergency alerting. While there may be rare or infrequent emergencies within a local community, aggregating these over a continent may produce a data stream that begins to cut significantly into revenue bearing traffic. The extent of the difficulty in this area depends on the nature of alerts accepted for delivery in this way and the threshold of seriousness used to determine which messages are sent. Consider, for example, sending information about slippery road conditions and school closings for the whole continental US. At the other end of the spectrum, low power FM (LPFM) and cellular base stations cover a much smaller area. These systems would not generally be overwhelmed with emergency alert data traffic, but on the other hand, many transmitters may need to be activated to alert a large population. This is possible, but inconvenient, from a scalability point of view. A good compromise, in terms of area coverage, would be systems that cover a metropolitan sized area. Examples include standard terrestrial broadcast TV and FM, as well as NOAA Weather Radio (NWR).
A second area of consideration for wireless broadcast systems relates to effective coverage and reliability of a given system. By their nature, wireless systems do not generally have a high reliability of message delivery, especially for mobile receivers. Cellular systems combat this situation with many base stations, multiple frequencies, plenty of RF margin, and spatial diversity (multiple antennas). For emergency alerting, one way to mitigate the reliability problem is to use multiple systems. However, we must recognize that building receivers that must select among multiple systems has a serious impact on the cost of these receivers, especially if different RF front ends are required. Another technique for reducing the reliability problem is to make use of either frequency diversity or spatial diversity. Cellular and PCS systems use both of these techniques. Also, broadcast TV and FM make use of multiple frequencies from multiple transmitters, many of which are at different locations. These multiple transmissions, however, are in the same band and use the same waveform, reducing the potential costs for receive-end alerting devices.
0.4 Improved Standard Protocols
While existing emergency alert systems (EAS and NWR) make use of wireless broadcast media, these systems need to be improved. In particular, the protocol for defining the emergency threat must be enhanced to facilitate precise geographic definition of threatened regions. These definitions should be in terms of latitudes and longitudes, not in terms of counties or other political or administrative regions.
0.5 Summary
The use of PSTN-based systems holds promise for near-term delivery of emergency alert messages. However, these systems should be augmented with more modern methods that make use of wireless media, used in a broadcast mode. Many wireless broadcast systems might be used, but these have pros and cons in terms of footprint size and coverage reliability. Careful system design must take into account the strengths and weaknesses of these systems. Finally, whatever wireless system is used, a standardized protocol must be developed for the emergency alert messages. This protocol must allow for precise geographic definition of the threatened region.
Specific replies to specific comments are given below, in order of their appearance on the NTIA website.
MITRE has no reply at this time.
MITRE has no reply at this time.
The low percentage of people that can be reached at any one time, using existing systems, points to the need to use multiple methods for providing emergency alert.
This is a real phenomenon, which highlights the need for precise geographic warning. Also, given a system that provides precise geographic warnings, system operators must resist the temptation to make the warning area too large. Doing so ultimately causes the public to lose confidence in the warning system.
Regarding interagency notification,
It should be clear from these comments that a standard protocol must be developed. This protocol, including all it variants, must be usable for all phase of alert notification, all the way from the sensors used to detect the threat, through the whole system, and finally to the public.
Any system that is developed must be able to deliver messages very quickly. However, this also implies the need for quick operational procedures as well as rapid delivery methods. Wireless broadcast message delivery would be faster than those based on wireline PSTN technology.
MITRE as a not-for-profit company, chartered in the public interest. MITRE might be helpful as part of this commission.
The delivery of messages over “any or all available communications channels” would certainly improve reliability. However, protocols must be defined that can be used on these channels. Some analysis will be required to evaluate each channel, its pros and cons, and the requirements and limitations it places on the protocol. In addition, the use of many different communications systems would have an impact on the cost of building receivers. If the receivers select the best of several different communications channels, this implies the need for multiple RF front ends. If each receiver uses only one of the possibilities, then economies of scale might be lost.
Please be aware of MITRE owned patent # 5,636,245. As a not-for-profit, MITRE will sometimes, under certain conditions, licenses its intellectual property (IP) at no cost to companies that intend to use the IP in the public interest. Contact Gerard Eldering at 703-883-7132 for more information.
Cell broadcast techniques seem like a promising component of emergency alerting.
Receiver equipment manufacturers must take into account the needs of hearing-impaired and sight-impaired people. These needs should be addressed in the development of new standardized emergency alert protocols, as well as in the development of equipment.
This is an important feature to add to any new standard protocol. Remember that providing plenty of information from multiple sources enhances the credibility of the warning. On the other hand, some hazards will require rapid response and there will not be time for the public to check other sources. In these cases, the alert warning should not include a message telling where to tune. Instead, the message should emphatically state that people should immediately leave the area, or take whatever other action is required.
MITRE has no reply at this time.
It may well be true that the public is not generally willing to pay for emergency alert information, until after a serious emergency. However, there may be other ways to finance the emergency alert system. One of the tasks of the All-Hazards roundtable should be to discuss economic models that might be used to pay for any emergency alert system. This model must address system development and deployment, as well and operation and maintenance costs.
MITRE does not agree that wireless systems, in general, are not well suited to carry the call volume associated with delivery of emergency alert messages. On the contrary, wireless systems, in general, are extremely scalable, because of their broadcast nature. On the other hand, paging systems may not be well suited for emergency alerting, unless they can be tailored to deliver a single message to a large group of people, with a single transmission.
Regarding the complex flow of interagency information required for public alert:
This complexity can seriously impede the timely delivery of EA messages. Mitigating this problem will require both technical and administrative solutions. The technology certainly exists to alert the public quickly. Administratively, it may be advantageous for messages to be delivered without interruption from the sensor directly to the public. Addressing the jurisdictional and administrative issues associated with emergency alerting should be a topic for the NTIA roundtable on All Hazard Warning.
In addition to the information listed by NCTA, many other data fields will be required in an effective EA system. In particular, the precise location information of the hazard must be provided.
Regarding weak signal or poor coverage
The possibility of weak signal and/or poor RF coverage emphasizes the need to use multiple channels. On the hand there are disadvantages of using different systems, since this increases the number of RF front ends required in an emergency alert receiver, and therefore increases the cost of any system. A good compromise is to design the emergency alert message to ride on a system with multiple transmitters using multiple frequencies, but within the same system. Broadcast FM and TV are good examples. Channel diversity will greatly enhance system reliability. Receiver-end equipment could automatically tune to the strongest signal. If no signal was detected at all, the receiver could produce a low-level alarm, alerting the user that something is wrong with the system.
Emergency altering requires moderate data rate, for relatively short times, but used infrequently. Paging and other messaging systems may not have sufficient data rate capacity, for individual addressing, but group broadcast capability may solve this problem.
Priority queuing will definitely by required for EA messages.
Note that 60-90 seconds may be too slow for some warnings. Accurate timelines for various emergencies must be constructed. These timelines must take into account the operational delays, as well as transmission and signal processing delays associated with the system. An overall time delay budget must be constructed for each emergency. In some cases, the budgets will be very tight.
MITRE has no reply at this time.
Community Notification Solutions, if they are based on PSTN message delivery, are not well suited for reaching the mobile public. Also, such systems tend to be too slow for some critical emergencies, such as earthquakes, where seconds can matter a great deal. Finally, PSTN-based solutions do not scale well for large populations.