On July 14, the Center for Strategic and International Studies in Washington, DC held a one-day conference premised on a specific threat: What if, in the future, war comes to the United States via cruise missile? Pointing to new developments in cruise missile technology, and the limitations of existing early warning systems that are focused on the high arcing trajectories of ballistic missiles, the CSIS conference and accompanying report suggests that to defend the continental United States from such a threat, the military should adapt and deploy the kind of cruise missile defenses presently used as regional weapons.

Unlike ballistic missiles, which arc up into space before traveling back down towards earth, cruise missiles fly close to the ground, making it hard for radar on the ground that’s pointed up at space to see them.

The perceived threat from new cruise missiles is driven by tech developments occurring across the globe, as new materials, better aerodynamics, and sophisticated sensors and guidance systems make possible the fielding of weapons, like hypersonic missiles, that had mostly been just theoretical decades ago.

For the United States, the development of long-range bombers in the 1940s, followed by the development of intercontinental ballistic missiles, shattered the notion that the enormous distances of the Atlantic and Pacific oceans were enough to protect the continental US from direct attack. (During World War II, US territories in the Pacific came under direct attack, but the only long-range assault on the 48 states came in the form of incendiary-carrying balloons launched by Japan into the jet stream and carried over to the US.)

With atomic and then thermonuclear payloads, bombers and long-range missiles threatened devastation on an unprecedented scale, and the United States built an elaborate system of early warning sensors focused on detecting early signs of launch, and expanded its first-in-the-world nuclear arsenal to deter attack. North American Aerospace Defense Command (NORAD) is run by both Canada and the United States, and maintains a series of radars and other sensors designed to detect early attacks across the Arctic or elsewhere. (Every December, NORAD highlights its existence by tracking Santa Claus, turning a system designed to detect oblivion into a kid-friendly Christmas tradition.)

At the conference held by CSIS, the threat from cruise missiles was discussed as a way that other countries could attack the United States that is hard to detect by employing existing, ICBM-focused measures. It is also considered hard to deter through threat of nuclear retaliation, operating on the assumption that if a cruise missile with a conventional warhead destroyed a building or killed people in the United States, the President would not immediately respond with a nuclear strike.

“You know, our adversaries are building diverse, expansive ranges of modern offensive missile systems, and we see them – we see them in the news every day,” Stan Stafira, Chief Architect of the Pentagon’s Missile Defense Agency, told the panel. “They’re capable of maneuvering in the midcourse and the terminal phases of their flight, like maneuvering reentry vehicles, multiple independent reentry vehicles, hypersonic glide vehicles, and cruise missiles.”

Part of the broader appeal of hypersonic weapons to nations like Russia, China, and the United States is that the speed and trajectories of the missiles make them harder to detect than ICBMs. The ballistic arc of ICBMs means the launch is visible to radar while it is still ascending, once it clears the horizon line. Meanwhile, both hypersonic glide vehicles and hypersonic cruise missiles, which travel at Mach 5 or above, are designed to fly below that radar horizon, with the cruise missile keeping a close trajectory to earth and the glide vehicle flying in the high atmosphere.

“I want to state that we absolutely believe that nuclear deterrence is the foundation of homeland defense,” said Lieutenant General AC Roper, deputy commander of Northern Command, the part of the US military responsible for North America. “However, we also must have credible deterrence options below the nuclear thresholds, options which allow for a balanced approach of deterrence by denial and deterrence by punishment or cost imposition.”

Deterrence, at its most straightforward, is a strategy of making a big threat on a condition: One country publicly declares it will launch nukes at another if it launches nukes at it, with the intended effect that neither country launches nukes. But because the payload of a cruise missile—it could be nuclear or conventional, unlike ICBMs, which are always nuclear—is unlikely to be known until impact, generals like Roper would prefer to have a range of weapons with which to respond.

Missile defense is one of those options, and the US already employs a few forms. Part of any missile defense system is the sensors, like specially focused radar, that can detect incoming attacks, and then track those weapons as they travel. These radars then send that tracking information to interceptors, which are missiles launched to fly and destroy the incoming attacking missile. Shooting missiles at other missiles is a hard problem because an incoming threat arrives at great speed, and because the cost calculus can favor an attacker. Interceptors, like shorter-ranged Patriot missiles or longer-ranged ballistic interceptors, are often more expensive than the missiles they are intercepting. And unlike interceptors, which have to hit precisely to work, missiles launched in attack can deploy decoys or countermeasures to redirect interceptors away, or can instead be fired in a greater volume, overwhelming interceptors through sheer numerical advantage.

“The resulting 20-year cost to provide even a light defense of a vast area ranged from $77 billion to $466 billion,” reads the CSIS report, citing an analysis from the Congressional Budget Office studying a range of cruise missile defense options. “The considerable cost variation is due to alternative combinations of sensors and interceptors and varying desired warning times of 5 or 15 minutes.”