The Reality of Emergency Notification
- By Hal St. Clair
- Oct 22, 2009
Since 9/11, companies of all types and sizes have diligently worked to better prepare for the “what ifs” of tomorrow. They have honed their business continuity/disaster recovery (BC/DR) plans to include worst-case scenarios like terrorism and pandemic influenza. And, many have taken the next step, adopting emergency notification technology to carry out the increasingly complex communications strategies of their respective organizations.
While these efforts are to be commended, it is important for contingency planners to realize there are oftentimes certain unmentioned limitations that may impact the overall speed and effectiveness of these deployments. In other words, the capacity of an emergency notification system or service is not simply the product of the number of telephone lines used or the speed of a network. Instead, it is the result of a combined number of physical and/or technological factors affecting landline/cellular telephone, SMS (text messaging) and e-mail usage, which ultimately define the volume of deliverable messages.
Even though other technologies, such as SMS and e-mail, have gained recent popularity, most organizations continue to rely on the telephone as the single most common method of communication. Reason being, the intricate, yet highly stable networks associated with landline usage ensure that under normal circumstances, anyone who chooses to place or receive a call can do so without issue. Yet telephone switches, like TDM (time division multiplexing) circuits do have finite capacity, meaning surges in call volume such as those created by emergency notification technology, can potentially overtax the network(s).
Other capacity issues affecting both the public and private sectors’ use of telephones include PBX, Community (Local Exchange Carrier End Office switch) and Tandem office limitations. How? PBXs (Private Branch eXchanges) are often limited in the number of inbound telephone lines they can handle. Community switches and network connections are chosen to support certain usage patterns. And, tandem switches have clear limits on the number of calls they can accommodate along a given route.
Unfortunately, many tend to think of a notification in terms of the number of calls that must be made. But in practice, each call has duration, making it considerably more accurate to think in terms of call volume generated (the number of calls multiplied by the average duration). However, even when the network imposes fixed capacity constraints, it is still possible to meet specific crisis communications goals by 1) shortening the length of the message, 2) decreasing the total number of calls to make and/or 3) simply allowing more time to complete the notification.
Also know unnecessary or inaccurate calls will waste valuable time in an crisis. Therefore, be sure the contact data utilized by the emergency notification system is current and duplicate numbers are eliminated. Establish a program to update this data regularly and perform random testing as often as possible to verify its validity.
Cellular service capacity also has limitations superficially very different from landline-based telephone service, yet similar in practice. In its simplest form, the maximum number of simultaneous conversations possible is fixed within a “cell,” but the number of persons in that area is generally much greater. This is because the tower(s) serving a given cell area can only handle a finite number of channels at any given time. Interference can also sometimes cause coverage to change over time, despite a subscriber remaining stationary. Even if this interference does not cause a call to fail, it can dramatically reduce the quality of the call so that an emergency message, or recipient’s touch-tone response, can become unintelligible.
Further complicating matters, today’s cell phones are much more than just phones.
They are, in most cases, fully capable computers providing such capabilities as email, web browsing and SMS/MMS messaging. While these features are beneficial, some actually use cellular network bandwidth (channels) which can lead to problems in the context of emergency communications. Due to these concerns, the FCC cited on its ruling of the WARN Act to define a national “commercial mobile alert service.” This law establishes a nationwide notification service for the distribution of public safety text messages to cellular and other wireless service providers.
To mitigate the risks associated with cell phone usage, consider the bandwidth (actual number of towers) in your usual notification area(s); reduce the number of people you must notify or alert them in phases; and keep the message short. More importantly, never include a URL (web address) in an email notification sent to a mobile device, as prompting so many people to access their browser(s) at the same time could negatively impact communications or create an outage.
Another area of risk is SMS (text messaging), which, for essentially the same reasons, can create service disruptions within a specific cell. Most service providers have limits on the number of SMS messages their network can accommodate, so it is rare an emergency notification system can successfully generate enough traffic to create a problem. However, there is no such limit on cell phones.
Text messages enter the cell network in a number of different ways, including email, direct SMS/SS7 (a special type of telephony signaling protocol) and SNPP (Simple Network Paging protocol). Perhaps the most important thing to understand about these technologies is that messages are accepted by the network for delivery rather than being sent directly to the recipient’s device. This means the management of text message traffic and delivery can vary significantly from one region to another. Compound this with difference between coverage areas (i.e., tower density and channel configuration), and the success rate can differ all the more, even when both locations lie within the same city.
The limit on messages per second places an essential absolute ceiling on the number of text notifications that can be sent by all external sources. This cap is determined by the cellular service provider and might be measured in hundreds of messages per second. While this number sounds high, it is important to remember that all message originators are competing to send messages. If an emergency notification system is attempting to send alerts, it will only have access to a fraction of this capacity. The portion it ultimately receives will depend on the number of other messages being sent at the time.
To improve the effectiveness of any crisis communications strategy, it is best to supplement text messaging with other notification methods such as landlines/cell phones and email.
As for e-mail, the rate at which messages may be sent is subject to the “bandwidth” of the network connections between each of the clients and servers it utilizes. For example, if the sender creates an email message with a 1 MB attachment and sends it to 100 people, his email server must send 100 messages, each 1MB in size, for a total of 100MB. If the server has a 1MB per second connection to the internet, each message will take about ten seconds to send. Sending 100 of these will take about one thousand seconds (almost 17 minutes), assuming there is no other traffic on the network and each of the target servers can accept data at that rate.
E-mail filters sometimes destroy legitimate incoming messages as spam, and will generally do so without notifying either the sender or the recipient. Even if the message is not destroyed, it may be delayed, rendering it useless in emergency situations. Moreover, many email servers monitor “blacklists” when determining which mail to accept. If a notification system’s network address is blacklisted, these mail servers will ignore all messages coming from it. For this reason, it is very important to use discretion when sending e-mail messages via an emergency notification system.
To maximize e-mail delivery rates, organizations should again keep their message(s) brief. By limiting the size of an e-mail message, they reduce the amount of time required by e-mail filters and firewalls to inspect and process it for delivery. Leaving out images and links to web addresses also lessens the likelihood the message will be rejected.
Organizations should also set up email distribution lists if the group of people to notified is relatively predictable and static, or easily maintained by the company itself (e.g., all employees). Notifications can then be sent to the “head address” for these distribution lists and the receiving email server will take the responsibility of distributing the message company-wide. This will ultimately free up network bandwidth and reduce the chance of individual messages becoming lost or delayed.
When organizations first implement an emergency notification system or service, they naturally have certain expectations for its performance. For most these are based on the infrastructure and statistics provided by the solution provider. For others, they may be founded on past experience with manual or single-modal methods (i.e., text messaging) of communication. Either way, organizations may find their expected notification capacity significantly diminished in a real-life crisis if certain mitigation strategies are not implemented ahead of time.
Planning for communications capacity is a vital and complex step in system implementation. Just remember to factor in the drawbacks associated with landlines, cell coverage, text messaging and email, as any of these may negatively impact overall performance. By understanding and employing certain mitigation strategies, organizations can adjust, measure and re-adjust accordingly. After all, emergency notification technology should ease the stress of notifying people in a crisis, not add to it.