Defense 5G: securing tactical and sovereign communications networks
Tactical 5G moves the core out of the data centre and onto vehicle-mounted, deployable and coalition-shared hardware. Encryption at rest and in transit does not protect it there. Confidential Computing isolates core functions while they run, ties key release to attested state, and with post-quantum cryptography keeps communications protected if hardware is captured.
Modern military communications depend on the rapid exchange of operational data. Sensors provide situational awareness, uncrewed systems transmit imagery, and mobile units access command and information systems. 5G can support these applications with high bandwidth, low latency and connectivity for large numbers of devices. These capabilities also make 5G relevant to military communications. NATO Allied Command Transformation is therefore exploring the technology for areas such as military communications, logistics and field-based operational applications.
As more operational traffic depends on 5G, the systems controlling that traffic become increasingly important. The 5G core is particularly sensitive. It determines who can access the network, which connections are established and how data is routed. An attacker seeking to compromise the network itself rather than a single connection would therefore target exactly these functions and the information they process. In tactical deployments, the core may also run on hardware far outside a protected data centre.
The security question is therefore straightforward: how does the 5G core remain protected if an attacker gains access to the server, the virtualisation layer or privileged administrative accounts? This is where Confidential Computing adds another layer of protection.
The 5G core as a critical component of tactical networks
The 5G core acts as a central trust component within the network. Its Network Functions perform different tasks, from authentication and session management to forwarding user traffic.
For armed forces, this role is particularly sensitive. When a vehicle, sensor or mobile device connects to the network, the core verifies its identity and authorisation, establishes the required session and applies the relevant access policies. Depending on the Network Function involved, it processes device and subscriber identities, authentication information, session states, certificates and cryptographic key material.
If this information falls into an adversary’s hands, the impact extends beyond the loss of confidential data. It can reveal information about participants and network structures, support the preparation of further attacks or, depending on the compromised core function, enable attacks on authentication and ongoing communications.
Private 5G networks also support mobile deployment models in the defence sector. Nokia Banshee, for example, provides a portable private 4G/5G network for tactical communications, integrating compute capacity and embedded core functions into a ruggedised field unit, whether deployed with a 4G EPC or a 5G core architecture. In this model, the compute infrastructure moves closer to the area of operations – together with the associated risks.
When the underlying infrastructure becomes part of the threat model
In a protected data centre, physical access can be tightly controlled. A mobile unit, vehicle or deployable edge system, however, operates under different conditions. Hardware may have to be left behind or can be lost in contested environments.
Privileged access to the host operating system and virtualisation layer remains another concern, whether through administrators or maintenance personnel. Multinational deployments add a further dimension. Coalition partners need to share communications and compute resources without fully merging their national security domains or exposing sensitive resources to one another.
Encryption is central to protecting sensitive information in all of these scenarios. Yet most systems still leave one important dimension exposed: data while it is being processed.
Closing the data-in-use gap with Confidential Computing
Conventional encryption protects subscriber information while it is stored in a database (data at rest). TLS or IPsec can secure communication between systems and Network Functions while data is transferred (data in transit). But when the core authenticates a device, establishes a session or applies a policy, the required information has to be available in memory for processing (data in use).
This processing stage is the data-in-use gap. If an attacker gains sufficient control over the host or virtualisation layer, they may attempt to read memory contents or manipulate the running workload. In a 5G core, this can expose particularly sensitive identity, session and key data, depending on the affected function. Under these conditions, communications cannot be considered fully protected.
Confidential Computing addresses this gap by isolating workloads inside hardware-based Trusted Execution Environments (TEEs). The CPU protects code and data – including the 5G core and its Network Functions – during execution from other workloads and privileged components outside the protected environment.
Remote Attestation adds another control. It produces cryptographically verifiable information about the protected environment in which a workload is running and its measured state. An organisation can compare that evidence against its security policies and, for example, release keys or secrets only once the required state has been verified.
Security benefits of Confidential 5G for defence organisations
Confidential Computing protects the confidentiality and integrity of sensitive 5G core workloads even when the underlying infrastructure is physically exposed, shared or compromised through privileged access. This provides several concrete benefits:
- Protection if hardware is captured: If a tactical server or mobile 5G unit falls into an adversary’s hands, physical access creates additional opportunities for manipulation, forensic analysis and attacks on running workloads or key material. Hardware-based isolation and attestation-bound key release limit this risk.
- Restricting privileged access: An attacker may target the infrastructure beneath the 5G core rather than the core itself, for example through an administrative account, the virtualisation layer or maintenance access. Confidential Computing establishes an additional security boundary between these privileged layers and the core workload.
- Verifiable integrity: Remote Attestation makes it possible to verify the measured state of a workload and tie key release to that evidence. This prevents a manipulated software component or configuration from automatically gaining access to sensitive resources.
- Coalition Trust: Multinational forces can share infrastructure while maintaining logical separation between their respective trust domains. Confidential Computing protects each partner’s core workloads from other privileged areas of the shared infrastructure. Beyond the 5G core, the same technology can also enable confidential data collaboration between partners, although this is a separate use case.
Typical deployment scenarios for Confidential 5G
The security mechanisms described above are relevant across a number of defence deployment models:
- Tactical 5G networks and 5G-in-a-Box: The core runs on portable or vehicle-mounted hardware, increasing the risk of physical access, manipulation or capture. Confidential Computing protects core workloads and limits access to sensitive key material even in these exposed systems.
- Sovereign military 5G infrastructure: Compromised administrative accounts, insiders or manipulated infrastructure components remain possible attack paths even inside an organisation’s own data centre. Confidential Computing complements existing access controls and reduces the number of underlying layers that a sensitive core workload needs to trust.
- Multinational networks: Coalition deployments require national 5G networks and shared communications infrastructure to interoperate while each nation retains control of its own security domains and secrets. Confidential Computing allows core workloads to run protected in these shared environments and provides participating partners with verifiable evidence of their state.
- Deployable edge: Core functions run close to the area of operations so that communications and local services remain available even when connectivity to central infrastructure is limited or interrupted. Confidential Computing protects these locally executed 5G core workloads from privileged access to the edge systems.
- Dual use for disaster response and civil protection: Mobile networks deployed in crisis areas or during large-scale police operations share many of the same technical characteristics. Cross-organisational use and resilient operation are important requirements here as well.
How enclaive implements Confidential 5G
enclaive brings together the core components required for Confidential 5G in an integrated Kubernetes architecture. Protected execution, Attestation and key management work together as part of the same security architecture.
The technical foundation is provided by Confidential VMs with enclaive Buckypaper, which create the protected runtime environment. Within these Confidential VMs, enclaive Dyneemes deploys and orchestrates dedicated Kubernetes pods. Sensitive 5G core functions can therefore run in a flexible, protected environment isolated from the host, hypervisor and infrastructure operator.
Protected execution is combined with identity verification and key management. Within enclaive’s virtual HSM (vHSM), Nitride handles Workload Identity & Access Management (WIAM) and Attestation, while Vault manages keys and secrets. Vault releases the required key material only after Nitride has verified that the expected workload is running in the intended protected state.
Using these technologies, enclaive is currently developing a prototype for post-quantum-secure 5G infrastructure as part of the NATO DIANA Accelerator Programme, with the aim of supporting secure and resilient communications for the long term.
An additional layer of protection: Post-Quantum Cryptography
Confidential Computing and Post-Quantum Cryptography (PQC) form part of the same trust chain. To keep confidential execution environments protected against future quantum threats, PQC secures the cryptographic key establishment and digital signatures they rely on. enclaive uses the NIST-standardised algorithms ML-KEM for key establishment and ML-DSA for digital signatures.
This scenario is particularly relevant to the defence sector. NIST has stated that a sufficiently capable quantum computer would most likely first be developed by a nation state and has also highlighted the risks if that state were a hostile actor. Even if the so-called Q-Day is still some years away, government bodies and a growing number of companies are already preparing their systems for post-quantum cryptography.
Another driver is the growing threat of “harvest now, decrypt later” attacks, in which encrypted data is collected today with the aim of decrypting it once sufficiently powerful quantum computers become available. For sensitive information – particularly data that must remain confidential for many years – post-quantum protection is therefore already relevant today.
Together with Confidential Computing, PQC adds a strong cryptographic layer of protection for 5G networks.
Requirements for a robust Confidential 5G implementation
Confidential Computing protects sensitive workloads against risks associated with untrusted infrastructure. It does not, however, remediate code vulnerabilities in a core application. Patch management, access controls and network segmentation remain necessary, and Attestation policies need to be clearly defined.
The VM-based approach described here also requires TEE-capable hardware supporting Confidential Computing technologies such as Intel TDX or AMD SEV-SNP. Firmware and the virtualisation environment must support the respective technology.
Integration effort also depends on the specific 5G core, particularly on how its Network Functions run on Kubernetes and how certificates, keys and other secrets are provisioned.
Conclusion
Tactical 5G networks move parts of highly sensitive communications infrastructure out of controlled data centres and into mobile, distributed and, in some cases, shared environments. Conventional hardening remains necessary, but it is not sufficient on its own.
Confidential Computing places the security and trust boundary closer to the workload itself. Core functions can be isolated while they are running, their state can be attested, and access to keys can be tied to that evidence.
For tactical 5G networks, this adds a protection layer precisely where security cannot rely on complete control of the underlying infrastructure. Confidential Computing therefore helps sustain mission-critical communications even in exposed and contested environments.
“Confidential Computing places the security and trust boundary closer to the workload itself.”
See the sealed version on your own stack, with the evidence your auditor gets to check.
TALK TO AN EXPERT →.png)