The Age of Assured Connectivity Is Over: Command and Control in a World Where the Signal Won’t Hold

On the first of June last year, 117 small drones folded out of the roofs of cargo trucks parked along quiet Russian roads and flew, at close range, into the flightlines of five strategic air bases. Ukrainian operators sitting thousands of kilometers away guided them onto the fuel tanks of Tu-95 and Tu-22M bombers. By the end of the day, Ukraine had destroyed or damaged roughly a third of Russia’s long-range cruise missile fleet. These are aircraft that cannot be replaced, because the Soviet industrial base that built them does not exist anymore. Per-drone cost: about $2,000. Operation Spider Web is now being studied in every serious defense ministry in the world.

It should be, though not for the reason most commentary has fixated on. The attack was a genuine achievement. Commercial FPV airframes smuggled across the largest country on Earth in hidden cargo compartments and piloted by operators aiming at specific weak points on strategic bombers — that’s hard, and the Ukrainian drone program deserves the credit it’s getting. But the thing that ought to keep defense planners up at night is the command and control architecture that made those drones effective: distributed launch, tight local autonomy on every airframe, a kill chain that ran through the operator’s hands on a laptop rather than through any centralized network, and a willingness to build the whole system on the assumption that every external signal would be denied, spoofed or jammed at the moment it mattered most.

That assumption is the single most important change in modern warfighting.

And it applies to both sides of the line. Eight months earlier, 100 miles from the Russian border, a British brigade and an Estonian division got overrun in a NATO exercise called “Hedgehog” by ten Ukrainians with drones and a laptop running Delta, a battlefield management system roughly analogous to what the Pentagon has been trying to build for a decade under the name CJADC2. Seventeen armored vehicles, 30 strikes, half a day. Two battalions were effectively erased. One NATO commander who watched the drill summed it up with a single line.

“We’re f—d.”

What happened at Hedgehog wasn’t that Western forces were outgunned. They were out-architected.

The signal is the target

For most of the last 30 years, American and allied forces have trained and fought with a set of implicit assumptions about the electromagnetic spectrum: that GPS is reliable, that radios work, that our networks are ours. Those assumptions are no longer operative, and the adversaries who understand this best are the ones who can’t afford to pretend otherwise.

You can jam a radio. You can spoof GPS. You can sniff out an operator with a cheap RF detector and a map. Spoofing creates what military analysts have started calling "electronic fog": drones, ships and aircraft that still believe they know where they are and are wrong. In the Gulf on a single day in February of last year, more than 1,100 commercial vessels had their navigation systems disrupted simultaneously. Shipping through the Strait of Hormuz, roughly a fifth of global oil and gas, slowed to a crawl. The jammers involved can be bought online for less than the cost of a decent mountain bike.

For civilian populations in contested areas, this is already daily life. Pilots in the Baltic region report degraded GPS on nearly every sortie. Merchant crews in the eastern Mediterranean plan their voyages assuming positioning will fail. Ukrainian families in the east check their phones and see the blue dot land somewhere in Russia. This is the new baseline. It isn’t a scenario anyone is war-gaming anymore. It’s the weather.

And it is no longer something that happens only abroad. In September, European Commission President Ursula von der Leyen’s plane was jammed on approach to Plovdiv, forcing the pilots to land using paper maps. The Associated Press has tracked nearly 80 such incidents since 2022, most attributed to Russia. Closer to my own backyard, cartels operating along the southern border are now routinely spoofing and jamming GPS to defeat U.S. counter-drone systems, a tactic they have been openly transplanting from lessons absorbed in Ukraine. Commercial airline pilots are logging GPS disruptions with increasing frequency over oceans and near contested airspace. The assumption that contested positioning, navigation and timing (PNT) is something that happens in somebody else’s war is no longer tenable.

Any force operating today should assume GPS degradation as a baseline condition. Forces that continue to depend on GPS alone will find themselves fighting blind.

That line, from an analysis in War on the Rocks last February, captures the shift cleanly. The electromagnetic spectrum is now contested, fragile, and increasingly hostile, and the signal will not hold. Planning around the assumption that it will is malpractice at this point. What commanders, and the people building the systems commanders depend on, actually need to be asking is what the system is supposed to do in the moments, or hours, or days that the signal is gone.

Two answers to the same problem

The U.S. Department of Defense has been working on the answer to this problem, formally, for the better part of a decade. The program has had several names; its current one is CJADC2, for Combined Joint All-Domain Command and Control. The vision is ambitious: a networked architecture that fuses sensors, shooters, platforms and services across every domain and every ally into a single coherent operational picture. It’s the right idea, and as anyone inside the program will tell you candidly that it has been a slog.

CJADC2 is built top-down. It starts from the question of how to integrate every existing service’s data standards, comms systems and legacy platforms into a coherent whole, and works outward from there. Every good-faith effort to build it has run into the same structural headwind: The systems it is trying to integrate were built in different decades by different contractors to different specifications, and the institutional machinery required to harmonize them moves on a different timescale than the threat.

Ukraine built its answer the other way. Delta, which NATO officers are now studying and one NATO member has moved to license, started in 2016 as a volunteer project by a group of drone enthusiasts called Aerorozvidka, Ukrainian for aerial reconnaissance. It was, at first, a single application: a digital map of the battlefield. Then a data-fusion layer on top of the map. Then drone feeds. Then satellite imagery. Then artillery coordination. Then secure messaging. Then Link 16 interoperability with F-16s. Today it ingests reporting from reconnaissance units, civilian officials, foreign intelligence partners and vetted bystanders; detects around 12,000 targets a day; and compresses the kill chain from detection to strike into minutes. It runs in a browser. It runs on a phone.

The tempting conclusion is that the West simply needs to move faster — that if we can just ship our version of Delta, we win. That misses the point. The deeper lesson is architectural. Delta was designed from the first line of code on the assumption that no part of its operating environment could be trusted to stay up. Ukrainian systems work in contested environments because they were built to be locally autonomous. They keep functioning when the network is down, the GPS is off, and the command post is out of reach, so that the smallest unit that can still fight has everything it needs, even if everyone above it goes silent.

Local autonomy is a design principle, not a feature

Local autonomy is one of the more important ideas to come out of the war, and it deserves more weight in American defense discourse than it’s gotten. One observer of the Ukrainian drone program put it this way.

“Ukrainian drone developers design systems with local autonomy so they can still operate if jammed. Observers now argue that future armies must adjust to networks that can hop frequencies, integrate civilian spectrum management and use AI to sense and counter jamming in real time.”

This is less a feature list than a philosophy. Local autonomy means that every node in the network, whether a drone, a vehicle, or a handheld, is built to complete its mission without the network. Connectivity is treated as a bonus, not a prerequisite. When the link comes back, the node reconciles with the broader picture. When it doesn’t, the node keeps going.

The alternative, the model most Western platforms were designed around, treats connectivity as ambient: GPS is always there, the datalink is always up, the command network is always present. We have spent 30 years fielding systems that are brittle by design.

This is why the Hedgehog result was so ugly. It wasn’t that the British brigade was poorly trained. It’s that its entire way of operating, moving in column, camping without concealment, relying on comms that assumed air superiority and spectrum access was built for a battlefield that no longer exists. The Ukrainian team was operating in a different category of warfare.

If you take local autonomy seriously as a design principle, several things follow. Networks must be able to hop frequencies and integrate civilian spectrum. AI has to move from the cloud to the edge, so that sensing, target recognition and counter-jamming can happen on-platform in real time. Kill chains have to be drawn up and drawn down rapidly, from below, by the operators closest to the fight. And this is the part most people skip: Every platform needs a way to know where it is, and where everything else is, that does not rely on any external signal at all.

Positioning as the foundation everything else assumes

Positioning is the quiet assumption underneath every C2 function. Targeting depends on it. So does deconfliction, coordination between units, and a shared coordinate system anyone can use. Timing for comms encryption, for networked weapons, for synchronized operations runs on GPS clocks whether operators realize it or not. Take GPS away, and you don’t just lose navigation. You lose the coordinate system the entire force is speaking in.

This is why a $130 jammer on Amazon is a strategic problem. It isn’t jamming one thing; it’s jamming the reference frame.

The good news, if there is any in this story, is that Earth has other reference frames. The planet’s magnetic field is one of them: complex, locally distinctive and present in every cubic inch of airspace and ocean and tunnel on the planet. Magnetometers capable of reading it are already embedded in almost every smartphone, most vehicles, and a large and growing share of autonomous platforms. They have been there for years, quietly acting as compasses and steadying attitude reference for drones. What nobody has done until now is turn them into a precise positioning system.

At AstraNav, this is the problem we’ve been working on: turning the geomagnetic field into a software-only positioning layer that runs on magnetometers already embedded in the systems militaries and commercial operators are operating today. No new satellites. No new infrastructure. No new hardware on the platform. The product, which we call M-GPS, delivers sub-meter positioning that cannot be jammed, cannot be spoofed, does not emit an RF signature for the adversary to sniff, and does not care whether the satellite constellation is operational. It’s the kind of thing local autonomy demands: a way for every node to know where it is, without asking anyone’s permission and without announcing its presence to the spectrum.

The point worth landing here is narrower, and I think more important: Any serious C2 architecture for a contested environment has to be built on a positioning layer that survives the contest. Everything downstream fails when location does.

What the next decade of C2 actually requires

If I had to boil the last few years down to something short enough for a whiteboard, it would be a handful of design principles that should reshape how the DOD and its allies buy, field, and fight with C2 systems.

1. Contested is the baseline.

Any C2 architecture designed around uncontested assumptions, GPS available, datalinks up, spectrum clean, is designing for a conflict we are not going to have. The Ukrainians learned this in months because they had to. We should not need to lose a war to learn it.

2. Architecture has to be built bottom-up.

CJADC2-style integration is still worth pursuing, but the operational value has to land at the level of the small unit, the individual platform, the single operator, and it has to land whether or not the broader network is alive. American programs that are useful only once the full stack is integrated three years from now are not going to survive contact with the enemy.

3. Software-defined is an unfair competitive advantage.

The adversary is iterating in weeks. Hardware-centric procurement cycles measured in years cannot keep pace. Every layer of the C2 stack that can be delivered as software (including, as it turns out, positioning) should be. That’s the only way to close the gap between threat evolution and force adaptation.

4. Resilience is a stack, not a single product.

No single technology solves contested C2. What you want is a layered architecture in which every critical function (comms, positioning, timing, sensing, decision) has a primary, a secondary, and often a tertiary means, and in which the platform can degrade gracefully across them without losing mission effectiveness. GPS alternatives, alternative-PNT more broadly, secure comms, local-edge AI, distributed kill chains — these are not competing technologies. They are components of the same stack.

5. The systems that will win the next fight are the ones that already assume the signal won’t hold.

I’d most like to see this one land with the people making procurement decisions. Every program office should be asking the same question of every platform it funds: What happens to this thing when GPS is denied, when comms are down, when the operator is on their own? If the answer is anything other than ‘keeps fighting,’ the program isn’t ready.

The C2 bar is set

Operation Spider Web, Hedgehog, the Gulf, the Baltic: these are not separate stories. They are the same story told from different angles. The electromagnetic environment that American and allied forces spent 30 years operating in does not exist anymore. The forces that are adapting fastest are the ones that never had the luxury of taking it for granted in the first place. They have built their C2 systems, their weapons, their drones and their command architectures on the assumption that nothing external is reliable and that the smallest unit in the fight must be able to function alone.

That’s the bar now. The signal will not hold. That is a given. The question is how your architecture will react when the signal goes out.

‍

Anton Toutov is the co-founder and CEO of AstraNav, a Texas-based dual-use technology company and the creator of M-GPS®, a software-only platform that converts Earth’s geomagnetic field into precise positioning and navigation data.

‍