TIA 5 - Post-Quantum Security by Design

From both an economic and a societal point of view, it is essential to consider security and privacy as integral parts of network design from the outset. Additionally, the envisioned security approach must be flexible and sustainable enough to meet the challenges of technological developments which can serve as a basis for future attacks. Therefore, the emerging threat of attacks by quantum computers and quantum hardware must be systematically investigated and defense mechanisms based on post-quantum cryptography and (quantum) physical layer security must be developed. A further focus is on the defense against intrusions as well as on the detection of intrinsic malfunctions and functional conflicts in the network. This calls for a holistic security architecture for 6G networks that follows the paradigm of security-by-design.

System-level security for 6G

One of the most important components for fully securing a system is the use of standardized mechanisms to ensure both safety and security. For example, a clear definition of a rights and role concept in relation to the communication of interfaces and services is urgently required. Furthermore, separation concepts, firewall-friendly designs, minimization of the effects of denial of service and the implementation of a zero-trust model must be considered and solutions integrated. One of the main enhancements in 6G will be Open Radio Access Networks (RANs). The measures mentioned can be applied holistically to the Open RAN system. The problem is that these criteria are often not at the focus of initial development and subsequent implementation work regarding security issues often poses major challenges.

Therefore, we are focusing on:

Provision of solutions for the seamless cooperation of effective security metrics and solutions
Analysis of existing system implementations for security problems

One-pagers

Post-quantum cryptography for 6G networks

Once they exist with sufficient power, universal quantum computers will break the currently used public-key cryptography. Since it is to be feared that such powerful quantum computers will exist during the lifespan of the 6G network, from the outset of 6G only cryptography that withstands quantum attackers should be used. While symmetric cryptography can still be used in the era of quantum computers, the current public-key schemes must be replaced by so-called post-quantum cryptography, which cannot be broken even by attackers with access to quantum computers. Compared to classical public-key cryptography, the security of post-quantum schemes relies on more involved hardness assumptions using the mathematics of error-correcting codes, isogenies, lattices, multivariate equations, or constructions using symmetric primitives such as hash functions.

The research topics include:

Analysis of the specific requirements in 6G regarding latency, bandwidth, storage and energy efficiency, and adapting post-quantum schemes for usage in 6G networks.
Optimizations of post-quantum schemes regarding their performance or memory requirements.
Analysis of physical attacks against implementations of suitable post-quantum schemes, including side-channel and fault attacks.
Proof-of-concept demonstration of the usage of post-quantum cryptography in use cases connected to 6G networks.

One-pagers

Privacy and physical layer security aspects

With the emergence of new networking architectures and services, there has been an increased interest in the potential of information-theoretic approaches to provide communication security by exploiting the physical characteristics of propagation channels. This is the case in MTC for IoT where simple and inexpensive devices are to be deployed, and therefore resource-efficient procedures of low complexity are required. Physical layer security is a promising mechanism to achieve confidentiality by exploiting the inherent randomness of wireless channels at the physical layer, which makes it particularly suitable for implementation in IoT systems, where small payloads, limited computational capabilities and low-latency constraints, make it hard to employ current cryptographic methods. As per requirements on physical layer security, no limitations are imposed on the eavesdroppers in terms of their computational capabilities.

The research topics include:

Fundamental limits for confidentiality and reliability of communication under semantic security metric including possibility of attackers having access to quantum hardware.
Construction of secrecy maps for indoor and outdoor spatial regions providing statistical security level guarantees with respect to semantic security metric depending on the positions of legitimate communication parties and the unknown position of a potential eavesdropper.
Update mechanisms for secrecy maps based on real measurements or ray-tracing simulation (cross topic TIA 3).
Quantitative assessment of security outlook improvement based on deployment of reflecting intelligent surfaces (RIS) in several communication scenarios (cross topic TIA 2).

The convergence of sensing and communication as well as envisioned functionalities like in-network computing and machine learning pose serious challenges to privacy, as they heavily rely on distributed processing of private or proprietary data. From a privacy perspective, it is essential to collect only user-related data that is necessary for the operation of the network and, additionally, include proactively privacy-preserving mechanisms in the design of communication protocols. In addition to supplementing existing protocol modules, such as channel charting, with privacy aspects by incorporating differential privacy and/or homomorphic encryption, new methods for analog distributed computation based on methods from lattice coding are also being developed.

One-pagers

Integration of physical layer security into the protocol stack

The key ingredients/enablers developed in previously mentioned research topics of TIA 5 for integration of physical layer security in communication systems can be summarized as follows:

Definition of system-relevant security metrics: Semantic security level quantifies operationally (i.e. for a class of well-defined attacks) the amount of information eavesdroppers can gain about the message that was transmitted. To account for the inherently uncertain wireless environment, the practical application of physical-layer concepts also necessitates a statistical formulation of the semantic secrecy guarantees.
Characterization of the wireless radio environment in the sense of physical layer security: A critical aspect of physical layer security is the level of knowledge of the inherently uncertain wireless environment, required to guarantee a certain security performance. In a first step, secrecy maps can be used to quantify and visualize achievable statistical security levels depending on the locations of the communicating parties. This visualization also makes it possible to identify spatial regions with security deficits.
Enhancement of physical layer security by customization of the wireless environment: Based on the contextual information provided by secrecy maps, reconfigurable intelligent surfaces can be deployed and programmed in such a way that a fraction of the reflecting elements are used to improve the quality of the incoming signal from the legitimate source to the destination, whereas the remaining reflecting elements are used to inject noise, and therefore deteriorate the link, to the spatial region which is rewarding for potential eavesdropper(s).

The statistical secrecy characterization, which provides the users‘ secrecy performance as a function of their position in the wireless environment, can be used to quantify the spatial availability of the security related QoS. In practice, the set of security related QoS levels could, for example, correspond to a prescribed set of modulation and coding schemes (MCS), thus providing relation to system aspects such as resource allocation and link adaptation. This opens the door to treating security as a service, where users can negotiate the security level based on their requirements and the cost. Based on this, the network allocates resources, or configures additional resources (e.g., access points or reconfigurable intelligent surfaces) to support the required service levels. Our goal is the development of novel communication protocols that extend current link adaptation and resource allocation mechanisms to incorporate physical layer security and security related QoS requirements.