Building the Nationwide Public Safety Broadband Network

Part Three of a Multi-Part Series

This segment of the series on building the Nationwide Public Safety Broadband Network (NPSBN) will focus on user devices, network access from the field, and the type of services I believe the network should provide on day one as well as into the future. The success of this network will depend upon the types of devices, how easy they are to use, the types of applications that will run on them, and fast and reliable access to the network.

Network Access

Starting with network access, the stated plan is to include partners to help fund the network construction and operation but to give Public Safety true pre-emptive access to the network at all times. This is a worthy goal, however, there are some technical issues that will need to be addressed in order for this type of sharing to occur and still give Public Safety immediate access to the network on a 24/7 basis. Pre-emptive access means that a Public Safety user who needs bandwidth will be given that access even if it means terminating an active session originated by a user with secondary status. As I understand it, this secondary user will first have bandwidth and data speeds throttled back, and then if necessary, will be kicked off the network. This of course would upset the secondary user but even today calls are often dropped or terminated on commercial networks.

The primary technical challenge with this approach is that in order for Public Safety users with pre-emptive rights to gain access to the network, they first must gain access to the network’s signaling channel so that the network knows they are trying to establish a session. If the signaling channel is overloaded with requests for service, there is a possibility that a Public Safety user will not be able to register on the network and, therefore, the network will not even know that a pre-emptive request has been made. When is the signaling channel likely to be overloaded? The answer is when there is an incident, usually within a single cell sector, in which the public and the press also have an interest.

We have this situation today with commercial networks. You might recall the East Coast earthquake a few years back. It was centered in Virginia and was not large enough to cause any real damage, but it did disrupt voice, data, and texting across all of the commercial networks. The networks were up and running (I don’t know of a single failure during or after the earthquake) but everyone was trying to access the networks to make calls, send messages, and send pictures, thus the networks’ signaling channels were overloaded. As a consequence, access to any of the networks was hit and miss.

Another example is the Wireless Priority Services (WPS) phone service across the wireless networks. This system is designed to give priority voice access to those who are registered and need it. However, the WPS system is very rarely used because if you cannot access the network, the network has no idea that you are trying to make a priority call.

Is there a solution for this? So far I have not heard of one and I have talked to a dozen or more very smart LTE engineers. Hopefully this is a recognized issue and is being addressed. Perhaps one solution might be to restrict the number of secondary users on the network or to come up with a way that when needed, the commercial, secondary users of this spectrum are routed back to their home networks and taken off the Public Safety network until the spectrum is made available again. How often will this occur? That is anybody’s guess at this point, but it will happen during incidents that are confined to a single cell sector or perhaps several sectors in major cities, and it will be limited to those cell sectors. However, during a major storm or other disaster, the problem could be much more widespread. So for now, pre-emption of the network is not a 100% certainty.

Next is the type of services that will be offered when the network is first turned on. When we were walking the halls of Congress during our campaign to obtain this spectrum, our premise was that this network would AUGMENT and not replace current Public Safety voice networks. This National Public Safety Broadband Network was intended to be used for data and video services to assist first responders in being better prepared, having more access to information they need, to be able to track and solve crimes faster, coordinate fire responses, and for the EMS community to be able to provide more in the field life-saving services. However, what has happened since the law was passed and FirstNet was formed is that many elected officials, who use cell phones every day and who have never used a two-way or land mobile radio, believe that this network will also become the only network Public Safety will need for voice, data, and video services.

Unfortunately, a number of equipment and network vendors have jumped on this bandwagon and are touting voice solutions that they claim can be incorporated into the network on day one. Add to this the fact that Public Safety Communications Research (PSCR) in Boulder has engaged with the LTE standards body to make some changes to the standards to provide additional voice services, and what we have ended up with is the expectation that this new broadband network will be all things to all people—first responders, secondary users, and secondary users’ customers. The good news is that while the PSCR is, in fact, working with the LTE standards body, it realizes that there is a difference between affecting a standard and having field proven voice communications, and that these two are a long way apart. Many within the Public Safety community are working on a simple, easy-to-understand document that will point out why voice across this network won’t happen for Public Safety use (it will not be mission-critical) for a very long time.

Today the only voice over LTE is being provided by Metro PCS while Verizon and AT&T are still working on solutions. One issue that has been discussed is that after adding dial-up voice to the device, its battery life is really poor. It is improving, but if we want devices that will last for a full shift and beyond, moving forward with voice services before the battery issues are solved would be a mistake.

One more aspect that must be taken into consideration is that when and if voice becomes viable across this network, dial-up voice will require absolute priority[1] on the network because it is VoIP and therefore the packets must be delivered in a timely and consistent manner. This means that the more dial-up voice there is on the network, the less network capacity there is available for data and video, which is the primary reason for the NPSBN in the first place. If we add push-to-talk (PTT) and there is a standard for which PTT technology is added, there is also a question about the amount of bandwidth taken up keeping live PTT units on the network. Some of the technologies being touted, which at this point are all proprietary, require each device that is active to be assigned some bandwidth even when not transmitting, which could also diminish the amount of capacity for data and video services. Therefore it is my conclusion that when the NPSBN is first deployed, no voice of any form should be allowed.

We don’t really know how much capacity will be available for voice or even for secondary customers on the network. If we assume that the network will be able to handle whatever is put onto it, we could end up with a network that is often overloaded and, therefore, not as reliable as Public Safety wants and needs. If we go back to the primary purpose of this network, which is for data and video services to augment Public Safety LMR voice, logic says that we should first build the network for only data and video services. After we have some experience with network loading, applications, and video management, we can then look at what other services can be added to the network without impacting its primary function.

Devices

Many vendors are working diligently at designing and building devices for use on this network. The original FirstNet vision is that any device on the network must be capable of:

1. Public Safety LTE network
2. Fallback to LTE/3G on four commercial networks
3. Fallback to satellite where none of the LTE network is available

Each device should also have:

• GPS
• Bluetooth
• Wi-Fi (suggest 4.9-GHz Public Safety licensed spectrum only)

I am not in agreement with this type of device for a number of reasons. First, one of the main purposes for choosing LTE was to provide devices that are more affordable than those currently in use on LMR voice systems. If we require each and every device on the network to include all of the above, we will end up with special devices. This will add to the cost, cut down the number of vendors that will build them, and prevent Public Safety from having the best choices for the first responder population. My belief is that we should leave at least the satellite portion out of the radios. We can provide access to satellites by mounting cheaper satellite units in vehicles where needed, and use 4.9-GHz Public Safety spectrum to communicate between the device and the vehicle.

Satellite services will be used only where it is not practical to build out the LTE network, such as in rural parts of the United States, but they won’t be practical in devices used in major cities where there is no clear view to the sky. I am skeptical about having five terrestrial networks built into each device, but perhaps it makes sense if we can acquire devices that are priced right. One issue I have with this is that many cell sites host all of the networks, so if that cell site is down all of the networks within that coverage area are down as well. Further, I am not sure how handoffs will occur from one network to another while maintaining data sessions. I have been told that this is possible but that it makes the Public Safety flavor of LTE even more non-standard. Again, one of the main selling points for LTE was to conform to worldwide standards.

Today the PSCR is testing a variety of devices. It is great that we have an independent organization testing devices to make sure they will operate on the network as it is deployed. I have seen many different types of devices including mobile modems for use with existing laptops in vehicles, mobile modems capable of two networks (one commercial and the Public Safety Broadband Network), a number of hardened tablets, which I believe will be ideal for incident commanders and other commanders in the field, and now a new generation of handheld devices.

The model for these handheld devices appears to be smartphones already on the market using the Android or Apple operating system. This makes sense because these devices are built by the millions, but these devices are lacking for Public Safety service in at least one major way. In order to access and send data and video, two hands are needed—one to hold the device and the other to tap on the screen or use a stylus and tap on the screen. We have conducted some preliminary research in this area and talked with many first responders. While many of them use smartphones in the field today as a secondary or personal device, none of those we talked with would even consider using a two-handed device while on an incident. Law enforcement, fire, and EMS personnel will not tie up both hands to make use of these devices. They must be able to be used one-handed, at least for when they are being used during an incident.

One example we have been given is a simple traffic stop. Officers pull over a car for some infraction, use their voice radio to call in the plate, type of car, and location of the stop, or perhaps type one-handed on a mobile data terminal in their car, keeping their eyes on the stopped car and the driver. Once it is confirmed that there are no wants or warrants, or that the car is not stolen, the officer will approach the vehicle, tell the driver to keep both hands where they can be seen, and will make sure his or her gun hand is not encumbered. The officer will use one hand to look at the registration, license, and proof of insurance and then, generally, return to his or her own vehicle to write a ticket, again keeping one eye on the driver and the car.

In this case, officers won’t use a handheld device to record the plate and car information while standing at the driver’s door, nor will they encumber both of their hands while out of their own vehicle. Those who are and will be building these handhelds need to understand that these devices have to be simple to operate, capable of one-handed operation, and hardened to withstand the elements. I don’t believe that today’s smartphones are the correct model for these devices, or perhaps the usage model has to be different when officers are on an incident. It will be great when an officer in the field will simply swipe a driver’s license through a device and have it populate a form for a ticket or a report, but that probably will not be done while standing next to the stopped car.

Fire personnel wear gloves much of the time they are at a fire so most of the smartphones on the market today won’t be of much use to them and the odds of them removing a glove to touch a screen are slim to none. Perhaps the incident commander using a tablet will be comfortable using both hands to track the fire, track vehicles at the scene, respond to others, and see live video of the back of a burning building, but that remains to be seen.

We also have to be realistic about calling for help when needed. I don’t believe that law enforcement, fire personnel, or EMS providers who find themselves in trouble and need immediate assistance will turn to their LTE device to summon that help. Instead they will use the microphone on their LMR radio, or if they are unable to do that, they will push the red button on the radio to summon assistance. Once the alert has been sounded, those responding locate the person in trouble by using the GPS built into the LTE device, but that does not require any action on the part of the person needing help.

I think this first generation of handheld devices will end up being used in the field for routine information queries, to view information sent to them by dispatch, to complete reports during downtimes, and to make better use of their time in the field. I believe that in the beginning these devices will be used to receive updated information regarding an incident, and perhaps to shoot a picture or video and send it to dispatch and/or other responding units, but both of these functions will have to be able to be performed with one hand if these devices are to be used during an actual incident.

It is important for those building devices or intending to build devices to do the research to determine what types of devices are needed, what types of applications, and how these devices can be used to assist those in the field. I am a big believer in ride-alongs with Public Safety. They can prove to be very beneficial for both parties. There are significant differences in how these devices will be used during routine patrols or responses during relatively quiet periods and how they will be used on Friday and Saturday nights when there are multiple responses with multiple units involved. We have found in much of our research that talking about what someone wants and needs and giving them multiple devices they can touch and feel will yield totally different results. When working with corporate customers, we always recommend a series of beta tests with different devices, and with a broad range of field personnel. The information gathered during these tests is invaluable in building better devices that will serve the purpose for which they are intended and attention must be given to each of the three “F”s: Form, Function, and Fit.

While a great deal of work is being done on handheld devices, I am in favor of starting out with vehicle-mounted systems, progressing to tablets for incident commands, and slowly populating field personnel with handheld devices as we learn what types of devices best fit the various groups within the first responder community. This will be a challenge for both FirstNet and the device vendors as I don’t foresee a single handheld device—a one-size-fits-all device—across the diversity of the first responder community. This means that the quantities of each type of device might be more limited and, therefore, prices may be higher. One misconception I have run into concerns pricing expectations for the handheld devices. Most commercial devices on the market for consumers are subsidized by the network operator. For example, an iPhone from AT&T or Verizon would cost:

iPhone 5 (LTE) 16 GB device with two-year contract         $199.99
iPhone 5 (LTE) low-end with no contract                         $649.00

Since no one is about to underwrite the cost of a Public Safety device, we have to figure that these handhelds will be in the $700-$1,000 range. If we add satellite back into the mix, prices will be even higher. While these prices are lower than many LMR radios, they are still higher than some expect them to be. Bottom line is that we need to be careful about setting cost expectations. The fewer devices built, of course, the higher the cost per device. If FirstNet does enter into partnerships with network operators, that should increase the number of devices and, if operators are still offering subsidies, this might help. However, there is a trend being led by T-Mobile and some of the smaller network operators to do away with device subsidies. In reality, we won’t have any idea of the device pricing until we see what FirstNet does as far as partnerships are concerned.

It is clear that the cost of the devices and the monthly fee for using the network will be borne by the local agencies. If the monthly cost paid to FirstNet is $40 per month per device (a guess on my part), there is also the issue of how many devices a local organization puts on the network. For example, if a police car has a notebook computer onboard, that is $40 a month. If the officer in the car also has a device, the total monthly cost for that vehicle will be $80 per month, or $120 per month for a two-person car. The same will apply to fire and EMS vehicles and devices for individuals. This is another reason I think that most agencies will deploy vehicle-mounted devices first, add tablets for some senior field people, and only add on-person devices over time. This fee will probably not include the cost of using commercial networks where the Nationwide Public Safety Broadband Network is not available (to start) or if the NPSBN is congested and some traffic is off-loaded onto a commercial network or two.

Conclusions

Even with all of the issues I have detailed in this and the first two installments, this network is needed. It will cut response times, save lives, and make fieldwork easier with less desk time at the end of the shift. I am confidant that the FirstNet team will find solutions to these issues and that the result will be what we were after in the first place: a truly nationwide, fully interoperable data and video network. At some point in the future, administrative-grade voice services may be added, later perhaps mission-critical voice services, and even later still, perhaps off-network voice services. In the meantime, Land Mobile Radio systems are vital to Public Safety and must be kept in place, up-to-date, and even expanded. The Nationwide Public Safety Broadband Network, in my estimation, is designed to augment LMR voice services and not replace them, at least not for a very long time.

There are obstacles to be overcome, but most of them will be overcome. Public Safety LTE is not only about the United States anymore. Public Safety is promoting LTE systems in most areas of the world including Canada, Latin and South America, Europe, Asia, Australia, and New Zealand. Each new system will bring more savings to Public Safety LTE, new devices, and certainly new applications. Within ten years the Public Safety community will look back and wonder how it ever managed without this network, and as LTE is refreshed, refined, and enhanced, we will see even more advances that will benefit all Public Safety organizations around the world.

One thing we have to remember, especially when it comes to voice over LTE, is that there are two aspects to all of this. The first is the advances in technology that will be required to enable Public Safety to have a single network that will provide for data, video, AND mission-critical voice. Many within the technology community are bullish on voice over LTE but it is still a very long way from being a reality. Standards need to be set, the proper release of LTE needs to be made available, it needs to be vetted, tested, and re-tested, and it needs to be field proven before anyone should even consider pulling the plug on their LMR voice networks.

The second part of this is the most dangerous for Public Safety and that is the perceptions of non-technical people who are elected to positions in the federal, state, and local governments as well as those appointed to regulatory organizations. The T-Band giveback requirement in the law is a perfect example of what can happen when these elected officials want to believe something but don’t have the grounding or understanding to accurately assess the consequences of passing a law or making a ruling that would turn off funding for mission-critical voice networks. Because of their assumptions (not based on fact) that voice over LTE is in the immediate future for Public Safety, the LMR channels could be returned to the government for use by others, or to be converted to broadband spectrum.

We must be very careful that the technology community does not make blanket statements that can be wrongly construed. My advice for the technologists is to continue the good work they are doing but to tone down their technology assessments and predictions with the reality of the time involved and all of the processes and issues that must be addressed and solved, assuming they can be solved. Until then, the Nationwide Public Safety Broadband Network will remain about data and video, while today’s Land Mobile Radio networks are about mission-critical voice. Both networks will be needed for a very long time to come.

Andrew M. Seybold