Sameh Yamany, PhD, Chief Technology Officer, VIAVI Solutions
Advances in quantum computers mean we’re fast approaching the point where traditional cryptographic techniques can be broken, and cyber criminals are already capturing encrypted data in anticipation of being able to use it later.
To counter such attacks, the US National Institute of Standards and Technology (NIST) has already announced three standardized algorithms to protect against post-quantum attacks (FIPS 203, 204 and 205) with others in development. And in the US (and many other countries) their adoption is starting to be mandated in law.
FIPS 203 and FIPS 204 both implement lattice-based cryptography techniques and are used, respectively, to establish secure connections by creating secret keys; and for giving digital signatures that prove a message's origin and integrity. FIPS 205 is also a digital signature algorithm, but instead uses hash-based cryptography, which creates larger signature sizes and results in slower verification, making it a backup for FIPS 204 should a weakness be discovered.
The algorithms can be applied in three ways: via software-coded post quantum cryptography (PQC), hardware-based quantum key distribution (QKD), or (for transitional architectures) through a hybrid approach.
However, it is commonly agreed that preparation is behind where it should be. The reason for this is not a lack of mathematical proof on their effectiveness, but instead the challenges hindering the rollout of these algorithms are ones of interoperability and scalability.
Table 1: A framework for testing QKD, PQC and Hybrid Systems, comparing classical and quantum components
Test and Measurement Techniques
Testing and verification of each implementation is essential to guarantee compliance and effectiveness.
For PQC systems, which have much longer encryption keys and algorithms that are operationally much heavier versus traditional ones. The impact on performance and architectural impact therefore needs to be measured. This means validating computational efficiency, encryption speed, key size variations and key generation speed.
It should also implement large-scale stress testing and emulate users and traffic to quantify performance against KPIs that include latency, throughput, and MoS scores under load.
Finally, PQC system algorithms need to comply with standards and have guaranteed interoperability, Testing therefore needs to include failure scenarios, for example mismatched PQC key selections between parties, or one party not supporting PQC at all.
Conversely, QKD system testing needs to examine and validate the quantum channel itself. This requires the measurement of the qubit error rate (QBER) to check for potential eavesdroppers. Furthermore, depending on whether the system is using discrete- or continuous-variable QKD, evaluation should test for precise clock synchronization, or validation of its phase reference alignment.
Like PQC, physical stress testing is vital, with emulated live network used to generate a range of controlled optical stresses, from polarization disturbances and reflection events to interference from DWDM signals. This testing must also extend to the service layer.
Hybrid systems combine classical, PQC, and QKD technologies with testing used to validate the quantum-classical interface and ensure that timing and synchronization are stable. QKD needs to be well integrated with IPsec or TLS, with security testing used to demonstrate resilience against side-channel attacks and confirm PQC will be used in the event of QKD links failing.
Performance testing is also essential, with the simultaneous measurement of not only quantum, but also classical metrics. And it should be confirmed that both quantum and classical traffic can co-exist within the same DWDM fiber.
At this point we should also mention key management system (KMS) interoperability. This bridges cryptographic controls across QKD, PQC, and hybrid systems. A KMS needs to exchange keys between disparate cryptographic domains, interface with multiple standards, and operate across vendors and protocols.
This naturally requires rigorous testing to validate for standard adherence, and for resilience, in the latter by simulating key delivery delays, corruption and loss. Similarly, testing needs to ensure the KMS can handle large-scale key distribution and validate the secure handshake and authentication mechanisms between QKD nodes.
More detailed information
In-lab validation is an essential part of protecting against a harvest now, decrypt later strategy, helping organizations move from planning to implementation.
For further information on this topic, a white paper examining each of the themes above as well as further considerations such as field testing, and how AIOps can deploy machine learning techniques to proactively secure infrastructure and detect threats is available for free from VIAVI. To find out more visit www.viavisolutions.com/qsafe.