Nir Goldstein is Founder & CEO of Guardian7.
For the past decade, mesh networking has been sold to militaries as the answer to battlefield communications. The pitch is simple: no single point of failure, self-healing routes and peer-to-peer resilience. It’s a good story. It’s also incomplete, and the gap between the pitch and the battlefield is where soldiers pay the price.
I’ve lived on both sides of that gap. Years spent in a special operations unit taught me what happens when a network plan looks perfect on a whiteboard and falls apart the moment a unit crosses a ridgeline. I later built a company around solving that exact problem.
The Physics Problem
Mesh advocates like to describe their architecture as inherently robust because there’s no central node to take down. That’s true in principle. But it assumes the physical layer underneath the mesh is reliable, and in practice, it often isn’t. The U.S. Army’s own testing office has documented this directly: The frequency spectrum the “Army is committed to using for its soldier radio waveform … has shorter transmission ranges than [legacy] radios and it’s especially susceptible to being blocked by hills and mountains.” A mesh topology can reroute traffic around a failed node all day long, but it can’t reroute around a mountain.
The part of the electromagnetic spectrum the Army is committed to using for its soldier radio waveform, which transports IP-based voice and data to and from individual radios, has shorter transmission ranges than the Army’s legacy SINCGARS radios and it’s especially susceptible to being blocked by hills and mountains.
This is the first failure mode: Mesh resilience is a layer two or three solution to what’s often a layer one problem. Terrain, foliage, urban clutter and adversary jamming all degrade the radio link itself, and no amount of clever routing logic recovers a signal that never arrived.
Complexity Is A Vulnerability, Not A Feature
The second failure mode is operational, not technical. The Pentagon’s Director of Operational Test and Evaluation (DOT&E) has repeatedly flagged that Army tactical networks are too complex for units to operate under stress. In one review, loading initial network plans, updating task organization and running a communications security changeover were described as “lengthy and cumbersome tasks requiring each individual radio to be manually updated”—a process that could take more than 24 hours for a brigade combat team. That’s an unacceptable delay for units in active combat operations.
Twenty-four hours is a lifetime in contact. A network that requires that kind of manual reconfiguration isn’t resilient—it’s fragile in a different way because it depends on time and calm that combat doesn’t provide.
More recent testing tells the same story with different equipment. DOT&E’s assessment of the Army’s modernized tactical network found that “light infantry companies equipped with the Leader Radio and Manpack [weren’t] operationally effective when operating the voice and data network in dense vegetation, the primary area of operations.” That’s not an edge case. That’s the job.
Why ‘No Single Point Of Failure’ Is The Wrong Design Goal
The industry’s answer to these failures has usually been to add more redundancy, more nodes, more alternate paths and more frequency options. Redundancy helps, but it doesn’t address the underlying issue: Nobody is making real-time decisions about which path, which frequency and which configuration to use as conditions change block by block. Mesh gives you options, but it doesn’t give you judgment.
What tactical communications needs is a layer that senses the spectrum environment in real time and makes those routing and configuration decisions automatically, the way a human operator would but at machine speed and without the 24-hour setup tax. That’s a materially different engineering problem than building a better mesh protocol. It requires treating the network less like a static architecture and more like an adaptive system that reads its environment continuously and reconfigures itself without waiting for a soldier to manually reload every radio.
The Path Forward
None of this means mesh networking should be abandoned. It means mesh is a component, not a complete answer. The military has spent years testing this the hard way, and the test data is public. The next generation of tactical communications systems needs to treat spectrum awareness and dynamic reconfiguration as first-class design requirements, not afterthoughts bolted onto a mesh protocol built for a calmer environment than a firefight.
The units I served with didn’t need a perfect network. They needed one that worked when everything else was going wrong. That’s still the bar, and it’s the one the industry should be building toward.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?