Building the Nationwide Public Safety Broadband Network
02.19.2013 by Andrew M. Seybold
Public Safety, on the other hand, will most likely turn this model around. That is, there will be more data, video, and information sent from the devices back to command centers, Incident Commanders, and others while the amount of data and video sent to the devices will be less, and less often.
Part Two of a Multi-Part Series
FirstNet held its most recent Board of Directors meeting on February 12, 2013, in Boulder, Colorado. This meeting included a press conference and many more assurances for the Public Safety community that this network will be for Public Safety first and foremost, and that the requirements of the Public Safety community will take priority in its construction and operation. I was pleased to hear the FirstNet Chairman address something many within the Public Safety community have been concerned about. He stated emphatically that while the network would be built out on a nationwide basis, it would be designed to provide local agency control. This is a very important message for Public Safety because while the network will be fully interoperable nationwide, it will be used locally every day.
After publishing Part One of this series I received many comments, 99% of which were positive, including several from FirstNet members and people who work with FirstNet. I also received numerous suggestions for topics others feel I should cover in this series. These ranged from how we will afford this network to more about devices, applications, operation, and cyber security on this network. I received multiple comments about how secure this network will and must be, especially in light of the fact that so many government and business networks have been hacked recently. Information that will be transmitted over this network must comply with federal and local standards, and it must be secure and trusted since if the network were to be hacked, it could have catastrophic consequences. So as I determine what to include in this series, I will, in this part, concentrate on how the NPSBN can be built with the funding available, funding that may become available, and with partners that want to assist Pubic Safety in these efforts. I will save the cyber security discussion for a future topic.
Today, FirstNet has a loan of $2 billion from the U.S. Treasury with a promise of $5 billion more from future spectrum auctions. The timing of these auctions is in doubt as is the amount of money they will raise so it is unknown when, or even if, this additional $5 billion will become available to FirstNet. In the meantime, FirstNet is gearing up, working on network architecture, talking to potential partners, and beginning to meet with the states, territories, and tribes as mandated in the act. FirstNet also stated that it will work with local first responder communities to gather information, although I am not sure if this will be a duplication of the states’ mandate or if it will focus on other issues. For now, let’s take a look at the network and the funding that is available and that could become available.
FirstNet has to start with the fact that it only has $2 billion to begin with and some of that must go to building its organization, hiring employees and consultants, and possibly a project management firm. Building this network is a huge undertaking that will take many smart people who know about building cellular networks, and it must include people who understand both the politics and requirements of Public Safety. When you put this all in perspective, building this network is 25% about the technology, 50% about the funding, and 25% about the politics on a federal, state, and local level. All of these things must be taken into consideration.
One way to become really discouraged about the lack of sufficient funds is to look at what the commercial network operators have and are spending rolling out their own LTE networks. Verizon has been at it for several years, AT&T started a little later but is making great progress, and Sprint and T-Mobile are fast on their heels. All of these operators are building out their LTE networks using existing sites, power, and back-up systems, then adding either high-capacity microwave or fiber at each site in order to handle the increased capacity requirements. AT&T and Verizon have both stated that even with existing sites they have spent many $billions more on their LTE network roll-outs than what is available to Public Safety. This points out a couple of things that FirstNet is aware of and looking at: Who can Public Safety partner with that can bring money and in-kind services, locations, and more to the network? What is the spectrum worth, on a secondary basis, to these and perhaps other partners? What potential partners make the most sense? How do commercial operators, utility companies, and others fit into the scheme of things? Other considerations also have to be taken into account. How much spectrum will actually be available for secondary usage, when (if) voice is added to the network, how will that impact the amount of secondary spectrum that will be available, and will voice be available on a secondary basis as well? (Voice over LTE, when and if it happens, has the highest priority level of all network traffic in LTE so the amount of capacity it will consume must be calculated into the spectrum equation from the start.)
Next, while the Public Safety network is being built on the worldwide standard of LTE, it will be different from commercial networks in at least one important regard. The standards bodies have designed LTE to provide more capacity and more speed down from the cell site to the devices and less capacity and speed from the devices. This is in keeping with the Internet model where customers tend to download much more data, stream videos, and download webpages than they upload from their devices. Yes, they send pictures and an occasional video, but by and large the most traffic is in a downward direction, which comes from the cell site to the device. Public Safety, on the other hand, will most likely turn this model around. That is, there will be more data, video, and information sent from the devices back to command centers, Incident Commanders, and others while the amount of data and video sent to the devices will be less, and less often.
To put this in perspective, OpenSignal found that average LTE speeds on a 10X10-MHz network were 13 Mbps in the download direction while OpenSignal clocked Verizon’s download speed at 10 Mbps. While AT&T only states on its website that its LTE speeds are “10X faster than 3G speeds,” Verizon states that its download speeds will average 12-15 Mbps and its upload speeds will be 2-6 Mbps. The tests run by OpenSignal only measured the download speed but other tests have indicated that the upload speeds run about 50-60% of the download speeds. If Public Safety uses more capacity in the upload mode, then the congestion point will be in the upload direction where less capacity is available to begin with. If an incident occurs within a single cell sector (as most incidents do), the number of video streams and other data in the upload direction will be about 50% of the download capability, thus Public Safety will have to do a very good job of managing this capacity and bandwidth.
When it comes to spectrum sharing, the fact that most Public Safety data and video will be in the uplink direction certainly plays to network operators that need more capacity for downloading information to their customers. However, it also means that in major metro areas with multiple incidents during the day, the network uplink capacity will have to be monitored and well managed by Public Safety. It is still not clear to me if the intent is to provide true pre-emptive priority for Public Safety. Pre-emptive means that first responders could, in fact, knock a commercial user off the air in a given area or given cell sector in order to ensure that the first responders have the bandwidth they need available, when they need it. The bottom line here is that FirstNet and Public Safety need to be cautious about how much spectrum sharing will be permitted and where. The Public Safety usage of the network will only increase over time as new applications and new devices are brought online. We cannot afford to assume that simply because there is excess capacity on day one of the network’s operation that this capacity will remain available as the network grows in size and the number of users increases. FirstNet has stated that this network will also be available to federal agencies and we will need to keep in mind their capacity requirements as well. If this network had been up and running in the Big Bear area of California during the recent week-long manhunt, it is very possible that commercial operations in that area would have been disrupted for the entire week because of the number of law enforcement personnel in the area and the need to update them with video and images as the manhunt continued.
Also at issue is the fact that the further a device is from the center of a cell site, the slower the data speed is and the less capacity there is. Our tests reveal that an average cell sector that covers out to three miles from the center could provide the following data speeds:
Distance from Center of Cell Download Upload
1.25 Miles 34 Mbps 16 Mbps
2.00 Miles 25 Mbps 12 Mbps
3.00 Miles (edge of cell) 768 Kbps 257 Kbps
The speeds listed above are based on the total potential capacity of a cell sector. AT&T and Verizon limit per-user capacity to provide for more users within a single cell sector having access to the shared capacity of that sector. As you can see from the above, the upload capacity and speed will have to be carefully managed. According to the Federal GAO report on data capacity and rates, an HD resolution video will require 3.5 Mbps of data capacity, while a typical video sent from an incident will require 1.2 Mbps of the total available capacity. So at the edge of the cell sector, only low-resolution video will be possible. If the system is designed correctly, the number of times that an incident occurs at the edge of a cell within a metro area will be minimized, but in suburban and rural areas, the edge-of-cell limitations will be more prevalent. This is why Public Safety may need to look at the type of Wi-Fi (4.9 GHz) vehicular repeaters mentioned in the first part of this series. A mobile unit in a rural area could make use of a higher-powered device for LTE, thereby increasing the data rate at the cell edge.
I am sure FirstNet is aware of all of this and the PSCR and NIST are reportedly running coverage models and capacity studies that will help determine the number of cell sites that will be required to provide true nationwide coverage. When the FCC released its white paper on costs for this network, it calculated that 41,600 cell sites would be required, yet AT&T and Verizon have many more sites up and running and they don’t cover many of the rural areas that need to be covered. Rightfully, FirstNet has stated that satellite communications will be needed to provide coverage in some of the truly rural areas where even with partners, it will be too expensive to build out Public Safety LTE coverage, but there are many areas where the addition of partners will make it more practical to build out the LTE network.
As mentioned before, I am somewhat leery of taking on spectrum sharing partners in the top 50 markets until we know exactly what the Public Safety demand will be. If I were a commercial network operator, I would be cautious about how much access I would really have to the Public Safety spectrum in these top markets. However, it is obvious that in most rural areas and in suburban areas as well, partners will have access to a lot of the Public Safety spectrum, on a secondary basis. While I was working with the Public Safety Alliance helping to convince Congress to award Public Safety the entire 700-MHz spectrum that it will really need, one of the projects I worked on was to come up with a way to identify potential partners in rural America. What I set out to do was to target specific congressional districts where many of those living in the district had no access to any Internet connectivity or very slow Internet connectivity.
I should say at this point that when I created these business models, the commercial operators had not yet started building out their LTE networks except in major cities. Further, the rural networks had not been able to make deals with the major operators to share spectrum and the Rural Communications Association (now the CCA) was opposing Public Safety because it wanted the spectrum to be auctioned. Therefore, I partnered with the National Rural Telecommunications Cooperative (NRTC). This organization provides services to co-op power companies and some rural telecommunications companies. Since that time, obviously, even more potential partners have been added.
As an example, we did an analysis of the State of Kentucky. At the time of the report, Kentucky had about 4,315,000 residents (based on a 2009 estimate), which included 1,749,000 households that an NTIA report broke down as follows:
Total Households: 1,749,000
Total with Internet access: 960,000 54.86%
Total with dial-up access: 253,000 14.44%
Total with broadband access: 700,000 40.02%
Total with anywhere access: 1,166,000 66.67% (meaning, home, work, or coffee shops)
Based on this information, we determined that 253,000 households only had slow-speed dial-up access to the Internet available while 1,049,000 homes had no access to broadband services. Next we contacted the NRTC and learned that the following rural power companies and telephone companies have facilities within the State of Kentucky:
Big Rivers Electric Corporation
Blue Grass Energy Cooperative Corporation
Clark Energy Cooperative, Inc.
Cumberland Valley Electric
East Kentucky Power Cooperative
Farmers Rural Electric Cooperative
Fleming Mason Energy
Foothills Rural Telephone Cooperative Corporation
Fox Creek R.E.C.C.
Gearheart Communications / dba Inter Mountain Cable, Inc.
Grayson Rural Electric Cooperative Corp
Green River Electric Corporation
Harrison County R.E.C.C.
Henderson Union RECC
Hickman-Fulton Counties Rural Electric Cooperative Corporation
Inter-County Energy Cooperative
Jackson Energy Cooperative
Jackson Purchase Energy
Kentucky Association of Electric Cooperatives
Licking Valley Rural Electric Cooperative Corporation
Logan Telephone Cooperative
Meade County Rural Electric Cooperative
Mountain Rural Telephone Cooperative Corporation
Nolin Rural Electric Cooperative Corporation
Owen Electric Cooperative, Inc.
Pennyrile Rural Electric Cooperative Corporation
Peoples Rural Telephone Cooperative
Salt River Electric Cooperative Corporation
Shelby Rural Electric Cooperative Corporation
South Central Rural Telephone Co.
South Kentucky Rural Electric Cooperative Corporation
SouthEast Telephone, Inc.
Taylor County Rural Electric Cooperative Corporation
Three Oaks Marketing and Development
Warren Rural Electric Cooperative Corporation
West Kentucky RECC
West Kentucky and Tennessee Telecommunications Cooperative
Mapping the areas of Kentucky covered by these companies, we determined that carefully constructed public/private partnerships would provide LTE broadband services to more than 90% of the state’s population and would provide Public Safety with broadband coverage to 95% of the state. The business model we used for this exercise was as follows:
Contributions from the rural power companies and telcos:
- Existing right-of-ways, high-tension towers for cell sites, backhaul where available
In exchange, each of these organizations would have secondary access to the Public Safety LTE spectrum and could:
- Use it for their own smart grid requirements
- Use it for meter reading and other command and control functions
- Resell it to their rural customers
These types of public/private partnerships are viable, these organizations, for the most part, want and need access to broadband, and they have both in-kind and monetary assets to bring to the program. They will have broadband and their customers will have broadband. Further, since all of these companies have service vehicles, they could easily install inexpensive LTE modems in homes and offices throughout their territories as well as providing repair services.
Today, of course, there are many more partners that might be willing to assist with the build-out in rural America as well as the suburban areas of the United States. I have talked with several larger electric companies serving counties surrounding major cities and they are also interested in this type of public/private partnership, and I am sure that many of the commercial operators would be as well. One benefit for the commercial operators might be the additional clout they would have with planning commissions, city councils, and county boards of supervisors. It still takes a long time for network operators to navigate through the planning process for new sites, but if the new site was also to include the Public Safety broadband network, that fact alone might shave some time off of the usually long and tedious process of new cell site placement.
Urban Bandwidth Sharing
Sharing spectrum in dense urban areas will need to be done with care. As mentioned, over time the demand for Public Safety data and video usage will increase and the number of Public Safety devices will grow so shared usage in the major metro areas may, in fact, be possible when the network is first put into operation. However, it is also possible, especially if voice is added to the network, that the amount of excess capacity will shrink to the point where a public sector company finds that its investment in the network is no longer viable. If the public partners are commercial operators, FirstNet must be cognizant of the fact that the traffic usage on these networks spikes in the same area as the incident, and during the same period of time. If the utility companies become our public partners, their usage will also spike during these times. All we have to do is look back to Sandy and other major storms where there were power outages, gas line issues, water, and sewage issues to understand that we could end up with the Public Safety network being overloaded. If Public Safety has pre-emptive priority, then these other vital services might be interrupted as well. Further, as FEMA and other federal agencies arrive on the scene, they too will need access to the network.
Obviously, careful planning and extensive real-time capacity monitoring of the network will be required. If the incident in question is localized to a few cell sectors or a few cell sites, then sharing can continue in other portions of the city. However, if the incident is widespread as it was with Sandy and other types of storms, then the demands for service across a wide portion of the urban network may well exceed the capabilities of the network. FirstNet is well aware that there is a balancing act here. It needs the partners in order to establish the network, but it is risky to bring on partners in areas where the amount of capacity available to them might end up not being their best long-term solution. I believe that FirstNet is looking at other funding models as well, as it should be.
During the last FirstNet meeting, a comment was made by one of the members that this network must be mission-critical, and I fully agree. However, the last part of the statement was that if a first responder were in trouble, he/she would need to be able to rely on this network. I take issue with that today. Since there is no mission-critical voice on the LTE network, no responders in trouble or in need of emergency assistance would pick up a smartphone. They would reach for their push-to-talk microphone, push the button, and ask for help—a one-handed operation—and someone would hear them. Many of them also have a red button on top of their radio that they can push to summon assistance. Yes, the LTE network must be mission-critical in nature, but the reality is that voice over LMR systems will still be widely used. During a life-endangering emergency, LMR voice will be the critical point of contact. The LTE network for voice and data will augment first responders’ capabilities in the field, but not replace LMR any time soon.
We must build the most robust and secure LTE network ever built and we must build it by being smart, partnering where practical, finding new sources of funding, and being creative, BUT we MUST continue to invest in our the Land Mobile Radio Systems that serve Public Safety today and will be needed well into the future. Today and into the foreseeable future, FirstNet’s broadband network is about augmenting the resources of those in the field and in the operations and 9-1-1 centers, but not about replacing the existing voice systems.
Andrew M. Seybold