Combat Cloud connects aircraft, drones, ships, and ground forces to share reliable tactical information in real time, despite electronic warfare.
In summary
The Combat Cloud is a distributed combat network that connects manned and unmanned platforms, sensors, and weapon systems, from the tactical to the operational level. It aggregates and shares heterogeneous data with minimal latency to accelerate the “detect-decide-act” cycle. In concrete terms, data links (Link 16, TTNT, MADL), multi-orbit satellite relays, and edge computing nodes process information locally, encrypt it, and disseminate it according to interoperability and priority rules. Embedded AI performs data fusion and proposes candidate tracks or shots; humans validate. The benefit is twofold: sustainable information superiority and a “sensor-to-shooter” time compressed to a few seconds. The challenges are severe: electronic warfare, cyberattacks, spectral dependency, congestion, and technological sovereignty. In Europe (SCAF/FCAS, GCAP) and the United States (JADC2/ABMS), architectures aim for network resilience, “zero trust” compartmentalization, and directional links with a low probability of interception.
The concept and its technical building blocks
The combat cloud is a “system of systems.” It connects fighter jets, drones, helicopters, surface vessels, submarines, radars, ground stations, and artillery units. Nodes exchange structured messages (tracks, sensor status, engagement orders, videos) within mesh networks. On the tactical link side, Link 16 (960–1,215 MHz) offers typical data rates of 31.6/57.6/115.2 kbit/s with TDMA framing and frequency hopping in 51 channels; “enhanced throughput” extensions are available. TTNT (Tactical Targeting Network Technology) adds ad hoc high-speed, low-latency capability for time-critical targeting. MADL, in the Ku band, favors narrow beams with low probability of interception/detection for stealth platforms. Beyond the horizon, satellite relay becomes a key component: geostationary satellites have a round-trip latency of around 500–700 ms, while LEO constellations have a latency of around 20–50 ms, which is useful for drone control or tactical video. In European programs (SCAF/FCAS; GCAP), the “Air/Multi-Domain Combat Cloud” aims for a decentralized and cyber-resilient topology where any platform can become an access point, message broker, or relay. These technical choices are publicly described by manufacturers and allied organizations.
Stakeholder participation: who talks to whom?
This network includes ISR sensors (AESA radars, EO/IR, ESM/ELINT), effectors (long-range artillery, anti-ship missiles, loitering munitions), C2 systems, and relay platforms. An F-35 can push a fused track to a ship, while a MALE drone transmits stabilized video to a rocket battery via a ground node. Ground forces publish their PNT position, ships share air/surface tracks, and satellites provide weather, imagery, and missile alerts. Joint centers ensure deconfliction and prioritization through quality of service policies. In terms of standards and methods, NATO federates networks via Federated Mission Networking (FMN), an interoperability and governance framework (Spirals profiles) to ensure “Day-0 interoperability.” The US JADC2/ABMS programs serve as the technical and doctrinal framework for “any sensor–best shooter,” while “Project Convergence” demonstrations link satellites, airborne sensors, and ground effectors to validate the multi-domain firing chain.
Step-by-step operation: from detection to effect
Typical sequence. 1) Acquisition: an AESA radar detects a track at 180 km; an ESM sensor classifies the emission. 2) Timestamping and georeferencing: the information is timestamped and PNT-aligned. 3) Publication: the track is published on a tactical bus (pub/sub) via Link 16/TTNT; high priority if threat. 4) Data fusion at the edge: an edge computing node correlates the track with EO/IR video and maritime AIS, enriched with a quality estimate. 5) Orchestration: a rules engine assigns an available “shooter” based on range, ROE, and risk of fratricide. 6) Authorization: embedded AI proposes, humans approve; the decision is sent back to the shooter. 7) Firing and evaluation: a salvo is fired; BDA (battle damage assessment) by sensor drone; loop closed. In ABMS/Project Convergence experiments, the “sensor-to-shooter” time has been reduced from minutes to windows of around ten seconds, depending on the scenario and connections. The operational benefit is measurable: fewer “escape windows,” better joint coordination, and more rational use of ammunition.
Concrete uses: from the sky to the sea
In the air, “cooperative engagement” allows a fighter to engage outside its own radar bubble, relying on a remote track. On the ground, an MLRS battery can receive an updated impact point from a drone, recalculate a ballistic solution, and fire in 20–40 seconds. At sea, a frigate can relocate a surface-to-air shot at the request of an air patrol vessel and share the illumination. The combat cloud also facilitates anti-drone warfare: RF detection, triangulation, automatic assignment of a jammer, then a kinetic effector if necessary. “Kill web” logic is replacing linear “kill chains”: multiple sensor-effector paths coexist, and the algorithm chooses the best one at a given moment based on network resilience, radio weather, geography, and ROE.
Operational advantages: speed, mass, and endurance
Advantage 1: Speed of decision-making. Compressed “sense–decide–act” cycles mean less exposure and more targets processed per hour. Advantage 2: Distributed mass. By multiplying inexpensive nodes (drones, passive sensors) and sharing a reliable situation, coverage is extended without concentrating vulnerability. Advantage 3: endurance. The marginal cost of data is low; missile stocks are preserved by choosing the proportionate effector. Advantage 4: surprise. Directional links (MADL), active antennas, and frequency hopping complicate interception. Advantage 5: information superiority. A consistent common operational picture (COP) reduces fratricide and duplication. Feedback from joint exercises shows that data volumes exchanged are “unprecedented” and sensor-to-shooter times have been reduced to tens of seconds, changing interdiction tactics and point air defense.
Risks, constraints, and blind spots
Risk 1: electronic warfare. GNSS jamming degrades geolocation; link jamming alters useful throughput; spoofing can inject false trails. Response: frequency diversity, multi-orbit (LEO/MEO/GEO), degraded inertial mode, and “anti-jam” using narrow beams. Risk 2: cyber. Without “zero trust,” a compromised node propagates the attack. Hence segmentation, strong authentication, micro-segmentation, signed updates, and whitelists. Risk 3: Spectrum congestion. Link 16 TDMA quickly saturates in dense bubbles; TTNT and adaptive OFDM schemes provide relief but require strict spectrum planning. Risk 4: Dependence on non-sovereign software stacks. Outsourcing orchestration and analytics to commercial clouds exposes you to “switch governance.” Risk 5: Ethics and human control. The more embedded AI pre-qualifies, the more machine decisions need to be tracked and audited to avoid bias or runaway behavior.
Robustness and security: how to hold up in a degraded environment
Holding up in electromagnetic noise requires complementary building blocks. 1) Transport diversity: protected fiber, microwave, resilient HF, multi-orbit SATCOM; automatic failover in case of loss (“self-healing”). 2) Robust synchronization: hybrid PNT (GNSS/M-code when available, LEO-PNT backup, local clocks) to keep TDMA aligned; controlled drift. 3) Adaptive protocols: redundant coding, selective ARQ, intelligent metadata reduction. 4) Security by design: approved encryption, dual-factor key management, attribute-governed “need-to-share,” sandbox for third-party payloads. 5) Governance: common ontologies and stable APIs to avoid “talking silos” that are unable to understand each other. 6) Observation: telemetry, network health score, near real-time cyber/EW correlation to trigger reconfiguration.
The role of JADC2/ABMS, SCAF, and GCAP programs
In the United States, JADC2 (and its ABMS pillar for the USAF) is pushing a “digital backbone” to connect sensors, C2, and “any sensor–best shooter” effectors. Demonstrations show firing chains reduced to windows of around ten seconds on specific scenarios, with dozens of simultaneous streams. In Europe, the SCAF/FCAS “Air/Multi-Domain Combat Cloud” and the GCAP “combat cloud” are converging towards a decentralized, cyber-resilient, NATO-interoperable architecture, open to the integration of “remote carriers” and the use of embedded edge computing. These programs have roadmaps in which encrypted connectivity, multi-sensor data fusion, and network resilience are milestones as critical as aircraft cells.
Useful metrics: latency, throughput, availability
Three figures guide the design. 1) End-to-end latency: aim for < 50 ms for fine control and sharing of critical sensors; accept ~500–700 ms in GEO for non-real time. 2) Useful throughput: a text track weighs little, a compressed HD video exceeds 2–6 Mbit/s; hence the use of video proxies and summaries when spectrum is lacking. 3) Availability: aim for “five nines” on critical functions, with multi-path failover and local storage of rules of engagement. These metrics are not theoretical: they determine relay planning, buffer size, and message scheduling.
Combat Cloud is not a magic wand. It requires infrastructure, shared standards, and a rarely popular configuration discipline. Armies that prioritize short-term effects without investing in the software backbone risk building a house of cards. Conversely, an overly centralized approach becomes an easy-to-attack “breaking point.” The line in the sand is clear: distributed architecture, simple rules, human control over engagement, and the ability to operate in “degraded mode” for hours. Those who master these fundamentals will dictate the tempo, even when facing a numerically superior adversary.
Live a unique fighter jet experience