These towers were designed to keep our military connected in case of nuclear attack. You can still see the towers on hilltops and prairies across the U.S., a monument to resolve in the face of the Soviet threat and naiveté about the real effects of nuclear war.

Just as Cold War preparedness could not make this network hard enough to survive nuclear war, no amount of money can harden today’s existing carrier networks to public-safety standards. Ma Bell’s network concept—great at the time—became obsolete with the invention of the resilient Internet, which originally was a creation of the Defense Advanced Research Projects Agency, or DARPA, a unit of the U.S. Department of Defense. Likewise, carriers’ macro-cell networks are great for the consumer market, but FirstNet should not depend on them to create a national network.

The best way to deliver an always available and resilient network uses existing towers and mobile base stations close to first responders. In land-mobile radio (LMR), this was done with in-vehicle repeaters for several reasons, including:

• Transmission close to the first responder offers the best signal propagation to penetrate school basements, elevator shafts, tunnels or industrial locations.
• The use of vehicle base stations for backhaul lessens the number of towers needed and subsequently reduces the network’s cost.

This approach can be achieved with Long-Term Evolution (LTE) by establishing more base stations (nodes), each with a smaller footprint (similar to how a BDA operates), and by making base stations low cost but non-interfering. These nodes and their antennas would be mounted on existing public-safety assets, like towers, but also on the roof of a fire station or school, or on a telephone pole.

Low-power base stations can use solar and battery backup to operate without utility power for far longer than today’s LMR nodes. And a network of small base stations can be more resilient than the current macro-cell LMR networks.

Harnessing leading-edge network gear would let FirstNet build its nationwide broadband network for first responders within its budget. If unencumbered by existing carrier networks, and by using mass-market consumer and off-the-shelf components, this network can be built cost effectively.

But here’s the most important point: FirstNet can’t build a network of towers, wired backhaul and a centralized core that is tough enough for the job. It needs to build a network that is resilient and decentralized—and that network does not look anything like the current mobile networks.

FirstNet can’t raise tower wind-loading standards high enough to survive the next Superstorm Sandy. It can’t specify enough days of backup fuel for the generators, nor can it harden backhaul conduits enough.

Instead, FirstNet needs to rethink network design. The new network will require more base stations and antennas than existing networks—but this is affordable when small-cell nodes cost under $3,000.

FirstNet has 1,500 detailed requirements. The resilient network we envision has eight top-level requirements. It must:

1. Be designed in the U.S. by U.S.-based vendors.
2. Be self-healing, to survive when isolated.
3. Use low-cost, small-cell base stations to fill in coverage holes with signals that are closer to users and can penetrate buildings.
4. Leverage carrier networks where advantageous, but it should not be based on carrier architecture
5. Use low-power nodes (<60 W) to enable extended operation during outages.
6. Be low cost, i.e., buildable within the current $7 billion FirstNet budget.
7. Be standards-based to allow hardware interoperability and applications from any vendor.
8. Be controlled by public-safety agencies, not carriers.

Let’s now examine each of these requirements.