Common Randomness (CR) can be considered as a resource for future communication devices and systems that not only improve the transmission capacity of identification systems but also provide robust information theoretical security. To achieve this, CR must be generated such that the reconciliation cost must be lower than the size of the key that is being generated. This can be achieved by using filters. In this live demonstration, we prepare a CR generation mechanism using Software Defined Radio (SDR) and achieve this tradeoff using multi-bit quantization and Savitzky Golay filter.
Common Randomness (CR) can be considered as a resource for future communication devices and systems that not only improve the transmission capacity of identification systems but also provide robust information theoretical security. To achieve this, CR must be generated such that the reconciliation cost must be lower than the size of the key that is being generated. This can be achieved by using filters. In this live demonstration, we prepare a CR generation mechanism using Software Defined Radio (SDR) and achieve this tradeoff using multi-bit quantization and Savitzky Golay filter.
Common Randomness (CR) as a resource is gaining increasing use cases for addressing futuristic communication demands such as scalability, technological sovereignty, and trustworthiness. For instance, CR can be applied to enhance data encryption. It also enables scalable identification (ID) codes and can support up to double-exponential ID capacity. Moreover, CR strengthens wireless communication systems by providing robustness against jamming attacks. However, from the viewpoint of a communication system, two distinct terminals or devices that wish to generate CR must first observe a correlated random source. In a wireless communication system, this source can be the channel’s stochastic nature, as reflected in the received signal. Traditionally, such an attempt has already been made for physical-layer secret key generation. However, the primary objective of PLSK is to provide information-theoretic security, which does not require generating CR at a very high rate. In novel communication methods such as ID-codes, for a discrete memoryless channel, the ID capacity is lower-bounded by CR capacity. In other words, to realize the benefits of CR assisted ID codes, it is required to generate a vast number of CR bits from very few channel observations. Therefore, there is a need to re-examine the PLSK methodology and modify it so that the resulting generated CR can be adopted into novel communication concepts, such as ID codes. This thesis aims to redesign PLSK so that the resulting information-theoretic CR rate more closely approaches the CR capacity than traditional PLSK schemes. The thesis introduces an initial PLSK implementation on Software-Defined Radios (SDRs) and discusses the trade-offs of attempting to improve the rate of CR generation in the PLSK system. Subsequently, the PLSK method is redesigned to enhance the CR rate. This is achieved through a two-step process: modifying CR generation methods and improving the storage of their components. In the methods section, I combine filters with low cutoff frequencies, multi-bit quantization, and gray-code encoding with a novel functional block called randomization. In the storage section, I propose introducing a 16-bit RSSI instead of an 8-bit RSSI to enhance CR rate through a novel paradigm called engineering entropy. I demonstrate and discuss the resulting improvement in the CR rate achieved with SDRs, compared to the corresponding state-of-the-art PLSK strategy.:1 Introduction 1 1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.4 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2 Physical Layer Secret Key Generation for Post-Quantum Cryptography 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2 Implementation of Physical Layer Secret Key (PLSK) . . . . . . . 9 2.3 Experimental Setup . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.6 Further Development . . . . . . . . . . . . . . . . . . . . . . . . 15 3 Overview of Common Randomness Generation 17 3.1 Theoretical Foundation . . . . . . . . . . . . . . . . . . . . . . . 17 3.2 Illustrative Example for Common Randomness Generation . . . 19 4 ImprovingCommonRandomness Rate Through Functional Blocks 23 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.2 Improving Common Randomness (CR) Rate . . . . . . . . . . . 26 4.3 Review of CR Generation Capacity . . . . . . . . . . . . . . . . 27 4.4 Individual Stages . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.5 Evaluation Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . 38 4.6 The CFO-CR Testbed: The Experimental Setup . . . . . . . . . . 40 4.7 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.8 Conclusion and Future Directions . . . . . . . . . . . . . . . . . 54 5 Improving Common Randomness Rate Through RSSI Storage 59 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 5.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 5.3 General System Model . . . . . . . . . . . . . . . . . . . . . . . 61 5.4 Extraction Stages . . . . . . . . . . . . . . . . . . . . . . . . . . 63 5.5 Testbed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.6 Evaluation Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . 71 5.7 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 5.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Conclusions 91 Bibliography 93
Common Randomness (CR) provides sequences of random variables at two physically separated locations. Ideally, the random variables at the two locations should be identical, i.e., have low probability of discrepancy, and should have high entropy. Previous CR research has focused on CR for physical layer security (mainly physical layer secret key generation), where a low CR generation rate, i.e., a low rate of CR bits per second is sufficient. However, emerging semantic communication paradigms, e.g., identification via channels, require high CR rates. We develop and evaluate a Carrier Frequency Offset (CFO) based methodology for high-rate CR generation from reciprocal observations of a common wireless channel between two distinct wireless terminals. The proposed CFO-CR methodology proceeds in several stages, including channel probing, random parameter extraction, noise reduction, quantization, information reconciliation, and randomization. Our evaluations with single-carrier software-defined radios, for which we make measurement traces publicly available, indicate that high-rate CR generation should observe (probe) the CFO and employ a Savitzky Golay low-pass filter with a low cut-off frequency for noise reduction in conjunction with multi-bit quantization, Gray code encoding, and a shuffling based randomization. We provide insights into the tradeoffs between the reconciliation cost for correcting bit discrepancies and the CR generation parameters. Our proposed CFO-CR methodology can generate 2048 bits of CR at a comparatively low reconciliation cost of 72 bytes while only making 256 channel observations and passing all common randomness tests. For generating 2048 bits of CR, other state-of-the-art approaches either require more channel observations (≥2048) or incur a higher reconciliation cost (≥450bytes).
With rapid advances made in quantum computing, information theoretic security is of paramount importance for Internet of Things (IoT), particularly for low-powered devices like sensors and actuators, where there is generally little to no room for complex security features. Common Randomness (CR) can be considered as a resource, that can be used for adding information theoretic security to the IoT systems. The most desired CR is a lengthy CR with uniform distribution. In this study, we design and implement channel reciprocity based CR using a Software Defined Radio (SDR). Furthermore, we also implement Quantum Entanglement (QE) assisted CR generation scheme and compare it against the classical channel reciprocity based scheme. The evaluation of both CR generation schemes has been carried out on the basis of National Institute for Statistical Tests (NIST) and bit discrepancy results. We show that QE assisted CR generation is more efficient compared to channel reciprocity based CR, in terms of generating uniformly distributed CR, if the latter is observing slow fading channel.
Identification is a new paradigm introduced for goal-oriented communication that verifies the semantics of the intended message rather than just transmitting the message bits, which is traditionally done in classical communication systems. One of the key techniques to improve randomized identification capacity is by introducing Common Randomness (CR). Physical Layer Secret Key (PLSK) is one such CR generation technique, where terminals generate secret keys based on changes in channel reciprocity between the wireless terminals and are mainly used for PHY layer encryption. However, the traditional way of evaluating PLSK techniques through bit discrepancies, bit generation rate, and entropy is not sufficient for evaluating their usefulness for identification. This paper proposes three key novel information theory paradigms that can be used alongside other mentioned parameters to evaluate key PLSK techniques for their usefulness in identification, i.e., CR rate, CR capacity, and cost function, followed by a supporting implementation using Software Defined Radio (SDR)s.
Common randomness (CR) is an important resource for future communication paradigms. Identification is one such future communication paradigm, and it has been proven that CR-assisted identification can improve identification capacity by a double exponential factor. To achieve this, the common source of observation must be perfectly correlated and contain sufficient entropy. Channel reciprocity-based CR generation (alternatively termed Physical Layer Secret Key (PLSK) generation) is one such CR generation scheme. Existing high-rate CR generation is based on the Carrier Frequency Offset (CFO); for practical implementation, CFO-based CR generation would require a re-design of the existing protocol stacks to pass the CFO from the physical layer to the higher protocol layers. In contrast, we develop high-rate CR generation based on the Received Signal Strength Indicator (RSSI), which is highly practical as the RSSI is passed to and widely used by the higher protocol layers in existing communication protocol stacks. For RSSI-based CR generation, we demonstrate that the memory resources dedicated to storing channel information play an important role in achieving a high CR generation rate. Specifically, we demonstrate, through an experimental testbed with Ettus N210 USRPs Software Defined Radios (SDRs) (from which we make datasets publicly available), how allocating higher memory resources for storing accurate RSSI channel information at the terminals (transceivers) increases the achievable CR generation rate.
Common randomness (CR) is a valuable resource for enhancing the trustworthiness of 6G communication systems. This article highlights the role of CR in improving scalability, security, and resilience in distributed 6G architectures. It provides an overview of information-theoretic two-source models for CR generation and summarizes known capacity results, together with their implications for practical systems under different communication settings, including one-way, two-way, helper-assisted, and interactive scenarios. Furthermore, the article explores CR-assisted secure identification over Gaussian channels as an application and highlights selected open research problems in CR generation for next-generation wireless systems. The article targets researchers and practitioners interested in bridging information-theoretic tools with emerging 6G communication systems.