Modern tactical operations run on data. Positional awareness, command coordination, intelligence distribution, targeting — functions that a generation ago relied on voice radio and physical map overlays now depend on continuous, reliable data flows across distributed formations. The network has become as fundamental to operational effectiveness as the weapons systems it supports.
The network architecture, in many cases, has not kept pace with that shift.
How Tactical Warfighting Changed
The change has been rapid and uneven. Over the past two decades, the operational dependence on connectivity has grown significantly — unmanned systems require data links, joint fires coordination moves through digital networks, and situational awareness is now built from a continuous feed of position and sensor data rather than periodic voice reports.
What has not grown at the same pace is the underlying network architecture that carries all of it. Tactical networks in many units remain a collection of purpose-built, point solutions — radios, satellite terminals, and tactical data links fielded to solve specific problems at specific moments, but never designed to function as a coherent, adaptive system under the conditions modern operations demand.
The result is a structural gap. The operational requirement is for connectivity that is resilient, adaptive, and capable of sustaining data flows under adversarial pressure. The network, in many cases, was built for conditions that no longer reflect the battlefield.
What Breaks When Connectivity Breaks
In a data-dependent operational environment, connectivity is not a convenience — it is a prerequisite for mission functions that commanders should be able to take for granted.
Positional awareness depends on it. Units sharing location data across a distributed formation requires that data moves reliably, with low latency, between nodes that may be moving, in terrain that blocks line-of-sight communications, under conditions where an adversary is actively attempting to disrupt the signal environment.
Command coordination depends on it. Decisions and policies from higher levels reach subordinate units through the network. When that link degrades or disappears, the command structure does not pause — it continues operating on older information, with less certainty, and greater risk.
Imagery and intelligence distribution depend on it. The value of real-time surveillance and reconnaissance data is time-sensitive. A degraded link does not just slow the data — it changes what the data is worth.
These are not hypothetical failure modes. Denied, degraded, intermittent, and limited connectivity — DDIL — is the expected operating condition in contested environments, not an exception case to be engineered around. Electronic warfare is pervasive on the modern battlefield. Adversaries have studied and invested in the ability to disrupt communications precisely because they understand the dependency.
The Architectural Mismatch
The core problem is a mismatch between how tactical networks were built and how they actually need to perform.
Legacy approaches to connectivity in tactical environments relied heavily on single-transport dependencies. A unit might have satellite communications as its primary link and limited alternatives if that link was jammed, degraded by weather, or simply unavailable due to orbital geometry. Failover, when it existed, was often manual — requiring operator intervention at exactly the moment when operator attention was most constrained.
In an enterprise environment, this produces a helpdesk ticket. In a tactical environment, it produces a gap in situational awareness.
The assumption that underpins both of these design choices — that the network will generally be available, and that human intervention can address exceptions — does not hold in contested, mobile, austere environments.
What a Different Architecture Looks Like
Addressing this requires reconsidering the foundational design principles, not just adding capacity or redundancy to an existing approach.
A network architecture built for DDIL environments treats connectivity as inherently unreliable and designs for that reality from the outset. It makes use of multiple transport underlays simultaneously — cellular, satellite, line-of-sight radio, wired — and shifts traffic dynamically based on real-time link conditions, without operator intervention. This is not simple failover. The system continuously measures per-link performance characteristics — latency, jitter, packet loss, and available bandwidth — and applies policy-driven traffic steering to route each application flow across the best available path at that moment. When a line-of-sight path is blocked, traffic moves automatically. When satellite latency spikes beyond a defined threshold, the system deprioritizes that link for specific forwarding classes of traffic and redistributes load across remaining transports. Mission-critical traffic — positional data, command traffic, imagery — can be assigned dedicated SLA policies that ensure it is steered preferentially and protected from link degradation that might be acceptable for lower-priority flows.
Topology also matters at the tactical edge in ways that do not apply in enterprise environments. A formation of vehicles and dismounted personnel cannot assume a hub-and-spoke model where every node has direct connectivity to a command element. Nodes need the ability to relay traffic through intermediate nodes automatically — daisy-chaining across vehicles or backpack-mounted devices to maintain end-to-end connectivity when direct paths are unavailable. The routing fabric needs to build and maintain awareness of these multi-hop paths dynamically, updating as the formation moves and link conditions change.
It also means designing for disconnected operation. In a software-defined wide area network architecture, the management plane, control plane, and data plane are logically separated. In a stable environment, that separation is an operational convenience — it simplifies centralized management. In a DDIL environment, it is a resilience requirement. The data plane must be capable of continuing to forward traffic based on locally held policy and routing state, even when the control plane is unreachable.
And it means applying end-to-end encryption not as a compliance requirement but as an operational necessity. In a contested RF environment, the transport layer is not trusted. Encrypted tunnels between every node in the network — established independently of the underlying transport carrying them — ensure that traffic intercepted at the RF layer yields no intelligence value.
What This Looks Like in Practice
Versa has deployed this architecture at the tactical edge across Department of War environments, including as a core technology provider in DISA’s Thunderdome program. In field deployments, Versa Operating System nodes are deployed across formations — on individual backpacks, mounted on vehicles, and at fixed command elements — running as a unified software-defined wide area network fabric across whatever transports are available at a given moment.
In a representative scenario, a combined formation of aircraft, vehicles, dismounted backpack nodes, and UAVs operate as a single connected fabric using multiple forms of transport. Traffic dynamically routes across available links based on continuously measured performance conditions. When line-of-sight communications are blocked by terrain, the system identifies the next viable path and shifts traffic automatically. When a satellite link becomes the only available transport but latency is high, it is used only for traffic that has no alternative — higher-priority mission flows are held on better-performing paths wherever they exist. No operator intervention is required to make those adjustments.
Nodes that lose connectivity to the control plane continue forwarding traffic based on locally held state. End-to-end encrypted tunnels are maintained between peer nodes regardless of the transport carrying them, so traffic intercepted cannot be exploited.
The units operating at the edge of that network do not have the luxury of assuming the network will be there. The architecture should reflect that reality — and in the deployments where it does, the difference is measurable in operational continuity, not just network uptime.
Rick Macchio is a Senior Systems Engineer at Versa, focused on federal and tactical network modernization.