Smaller cells have been the most important contributor to throughput improvement since the birth of cellular networks. They are likely to evolve further in the shift to cell-free massive MIMO (CF mMIMO), where multiple closely placed access points (APs) collaborate to serve users. This scheme is particularly suitable for millimeter wave (mmWave) communication, which enables very high data rates with its large bandwidth, but encounters severe challenges of high path loss and blockage. The CF mMIMO network is a good countermeasure to these two challenges by utilizing overlapping signals from different APs and macro-diversity. In this work, we demonstrate that mmWave CF mMIMO network optimization is largely an AP-user assignment problem. To solve this large-scale, nondifferentiable problem, we propose an unsupervised machine learning (ML) approach, which looks for the optimal solution autonomously without labels. A customized graph neural network architecture tailored to the problem properties is proposed, which enables distributed optimization without a central unit, allows for a varying number of users, and hierarchical permutation-equivariance of APs and users. A teacher-student model is applied to prune the graph, where the teacher model uses a fully connected graph for maximum performance, and the student model uses a pruned graph to reproduce the teacher's behavior with less communication in fronthaul. Moreover, a special training method is designed, which relaxes the combinatorial problem to a continuous one. In this way, we can apply gradient-based neural network training. An entropy-inspired penalty is introduced to make the relaxed problem equivalent to the original one. The analytical augmented Lagrangian method is combined with ML for the constrained optimization. Simulation results show that the proposed approach outperforms baselines in both performance and computation time. In addition, with a properly pruned graph, the proposed approach performs inference in a distributed manner with sparse message passing between APs, realizing a low signaling overhead in fronthaul, and a performance close to the fully connected graph.

We consider a superdense coding based quantum communication system utilizing entangled qubits. These qubits are added to and retrieved from a pool of available entangled qubits. In this work, we examine the reliability of such a system with respect to probability of depletion of the entangled qubit budget. Furthermore, we analyze the latency ahead of resuming transmission. Our model and analysis includes specific effects, such as quantum decoherence. Additionally, we compare different approaches for transmission after exhaustion of the entangled qubit budget. Since reliability and latency are essential metrics for quantum communication systems, joint analysis of these in relation to the system parameters is of significant interest. The results presented in this work will help guide quantum communication system designers in making modifications to meet reliability and latency specifications.

Non-orthogonal multiple access (NOMA) is a promising multiple access technique. Its performance depends strongly on the wireless channel property, which can be enhanced by reconfigurable intelligent surfaces (RISs). In this paper, we jointly optimize base station (BS) precoding and RIS configuration with unsupervised machine learning (ML), which looks for the optimal solution autonomously. In particular, we propose a dedicated neural network (NN) architecture RISnet inspired by domain knowledge in communication. Compared to state-of-the-art, the proposed approach combines analytical optimal BS precoding and ML-enabled RIS, has a high scalability to control more than 1000 RIS elements, has a low requirement for channel state information (CSI) in input, and addresses the mutual coupling between RIS elements. Beyond the considered problem, this work is an early contribution to domain knowledge enabled ML, which exploit the domain expertise of communication systems to design better approaches than general ML methods.

We consider a two-ray channel model in which the distance between transmitter and receiver is only known up to an interval. Due to the unknown distance, destructive interference might occur which significantly reduces the receive power. To mitigate this problem, multiple frequencies can be used in parallel. In this work, we consider a worst-case design approach which allows maximizing the guaranteed quality of service (QoS) despite the uncertainty about the channel. First, we derive the worst-case receive power within the uncertainty region. Next, we compare different approaches to assign frequencies to the user such that the worst-case is maximized. We propose a greedy algorithm, which significantly outperforms standard baseline schemes while also being resource efficient. With this, the communication system can be designed such that a certain performance can always be guaranteed, and ultra-reliability is practically achieved.

Resilience and power consumption are two important performance metrics for many modern communication systems, and it is therefore important to define, analyze, and optimize them. In this work, we consider a wireless communication system with secret-key generation, in which the secret-key bits are added to and used from a pool of available key bits. We propose novel physical layer resilience metrics for the survivability of such systems. In addition, we propose multiple power allocation schemes and analyze their trade-off between resilience and power consumption. In particular, we investigate and compare constant power allocation, an adaptive analytical algorithm, and a reinforcement learning-based solution. It is shown how the transmit power can be minimized such that a specified resilience is guaranteed. These results can be used directly by designers of such systems to optimize the system parameters for the desired performance in terms of reliability, security, and resilience.

space-division multiple access (SDMA) plays an important role in modern wireless communications. Its performance depends on the channel properties, which can be improved by reconfigurable intelligent surfaces (RISs). In this work, we jointly optimize SDMA precoding at the base station (BS) and RIS configuration. We tackle difficulties of mutual coupling between RIS elements, scalability to more than 1000 RIS elements, and high requirement for channel estimation. We first derive an RIS-assisted channel model considering mutual coupling, then propose an unsupervised machine learning (ML) approach to optimize the RIS with a dedicated neural network (NN) architectureRISnet, which has good scalability, desired permutation-invariance, and a low requirement for channel estimation. Moreover, we leverage existing high-performance analytical precoding scheme to propose a hybrid solution of ML-enabled RIS configuration and analytical precoding at BS. More generally, this work is an early contribution to combine ML technique and domain knowledge in communication for NN architecture design. Compared to generic ML, the problem-specific ML can achieve higher performance, lower complexity and permutation-invariance.

Millimeter wave (mmWave) cell-free massive MIMO (CF mMIMO) is a promising solution for future wireless communications. However, its optimization is non-trivial due to the challenging channel characteristics. We show that mmWave CF mMIMO optimization is largely an assignment problem between access points (APs) and users due to the high path loss of mmWave channels, the limited output power of the amplifier, and the almost orthogonal channels between users given a large number of AP antennas. The combinatorial nature of the assignment problem, the requirement for scalability, and the distributed implementation of CF mMIMO make this problem difficult. In this work, we propose an unsupervised machine learning (ML) enabled solution. In particular, a graph neural network (GNN) customized for scalability and distributed implementation is introduced. Moreover, the customized GNN architecture is hierarchically permutation-equivariant (HPE), i.e., if the APs or users of an AP are permuted, the output assignment is automatically permuted in the same way. To address this combinatorial problem, we relax it to a continuous problem, and introduce an information entropy-inspired penalty term. The training objective is then formulated using the augmented Lagrangian method (ALM). The test results show that the realized sum-rate outperforms that of the generalized serial dictatorship (GSD) algorithm and is very close to an upper bound in a small network scenario, while the upper bound is impossible to obtain in a large network scenario.

We consider a wireless communication system with a passive eavesdropper, in which a transmitter and legitimate receiver generate and use key bits to secure the transmission of their data. These bits are added to and used from a pool of available key bits. In this work, we analyze the reliability of the system in terms of the probability that the budget of available key bits will be exhausted. In addition, we investigate the latency before a transmission can take place. Since security, reliability, and latency are three important metrics for modern communication systems, it is of great interest to jointly analyze them in relation to the system parameters. The results presented in this work will allow system designers to adjust the system parameters in such a way that the requirements of the application in terms of both reliability and latency are met.

Resilience and power consumption are important performance metrics for many modern communication systems, and it is therefore important to define, analyze, and optimize them. In this work, we consider a wireless communication system with secret-key generation, in which the secret-key bits are added to and used from a pool of available key bits. We propose novel resilience metrics for the survivability of such a system and analyze them. In addition, we investigate the problem of minimizing the transmit power such that a specified resilience is guaranteed. These results can be used directly by designers of such systems to optimize the system parameters for the desired performance in terms of reliability and resilience.

We consider a wireless communication system with a passive eavesdropper, in which a transmitter and legitimate receiver generate and use key bits to secure the transmission of their data. These bits are added to and used from a pool of available key bits. In this work, we analyze the reliability of the system in terms of the probability that the budget of available key bits will be exhausted. In addition, we investigate the latency before a transmission can take place. Since security, reliability, and latency are three important metrics for modern communication systems, it is of great interest to jointly analyze them in relation to the system parameters. In particular, we show under what conditions the system may remain in an active state indefinitely, i.e., never run out of available secret-key bits. The results presented in this work will allow system designers to adjust the system parameters in such a way that the requirements of the application in terms of both reliability and latency are met.

This work presents a machine learning approach to optimize the energy efficiency (EE) in an interference network. This optimization problem is non-convex and it is difficult to find its global optimum. We propose an unsupervised machine learning framework to approach the global optimum. While the training of the neural network (NN) takes moderate time, applying the trained model requires very low computational complexity. In particular, we introduce a novel objective function based on the reparameterization trick, which makes it possible to prune poor local optima to converge to the global optimum. Furthermore, we design a dedicated NN architecture SINRnet for signal-to-interference-noise ratio (SINR)-related optimization problems in interference networks, which is permutation-equivariant and classifies channels according to their positions in the SINR expression. In this way, we encode our domain knowledge into the NN design. Training and testing results show that the proposed method outperforms the successive convex approximation (SCA) algorithm, achieving an EE close to the global optimum found by the branch-and-bound algorithm and with reasonable computational effort. Thus, the proposed approach finds a balance between computational complexity and performance.

The reconfigurable intelligent surface (RIS) is a promising technology that enables wireless communication systems to achieve improved performance by intelligently manipulating wireless channels. In this paper, we consider the sum-rate maximization problem in a downlink multi-user multi-input-single-output (MISO) channel via space-division multiple access (SDMA). Two major challenges of this problem are the high dimensionality due to the large number of RIS elements and the difficulty to obtain the full channel state information (CSI), which is assumed known in many algorithms proposed in the literature. Instead, we propose a hybrid machine learning approach using the weighted minimum mean squared error (WMMSE) precoder at the base station (BS) and a dedicated neural network (NN) architecture, RISnet, for RIS configuration. The RISnet has a good scalability to optimize 1296 RIS elements and requires partial CSI of only 16 RIS elements as input. We show it achieves a high performance with low requirement for channel estimation for geometric channel models obtained with ray-tracing simulation. The unsupervised learning lets the RISnet find an optimized RIS configuration by itself. Numerical results show that a trained model configures the RIS with low computational effort, considerably outperforms the baselines, and can work with discrete phase shifts.

Ensuring a reliable and simultaneously secure transmission of data is one of the major challenges for wireless communication systems. This is especially difficult when no perfect channel state information (CSI) at the transmitter is available. In this work, we consider a two-ray ground reflection scenario with a passive eavesdropper. At the transmitter, there only exists limited knowledge about the channels to both the legitimate receiver and the eavesdropper. We propose a simple frequency diversity scheme which maximizes the worst-case secrecy capacity for the considered scenario. In particular, we show how to optimally adjust the frequency spacing between the used frequencies. Thereby, we can guarantee a certain secrecy rate at which data can be transmitted both reliably and securely for all locations of the receivers.

This work presents a machine learning approach to optimize the energy efficiency (EE) in a multi-cell wireless network. This optimization problem is non-convex and its global optimum is difficult to find. In the literature, either simple but suboptimal approaches or optimal methods with high complexity are proposed. In contrast, we propose an unsupervised machine learning framework to approach the global optimum. While the neural network (NN) training takes moderate time, application with the trained model requires very low computational complexity. In particular, we introduce a novel objective function based on stochastic actions to solve the non-convex optimization problem. Besides, we design a dedicated NN architecture SINRnet for the power allocation problems in the interference channel that is permutation-equivariant. We encode our domain knowledge into the NN design and shed light into the black box of machine learning. Training and testing results show that the proposed method without supervision and with reasonable computational effort achieves an EE close to the global optimum found by the branch-and-bound algorithm and outperform the successive convex approximation (SCA) algorithm. Hence, the proposed approach balances between computational complexity and performance.

With the increasing application of internet of things (IoT), the number of wirelessly transmitting devices is on a rise. It is important that the energy efficiency (EE) is maximized to reduce interference and save energy. This work explores the possibility of power control for maximum EE in wireless interference networks using reinforcement learning (RL) techniques. We apply the soft actor-critic (SAC) algorithm based on entropy regularization that allows to escape local optima and foster exploration. This enables us to solve the energy efficient power control problem with reduced complexity. We demonstrate that the obtained solutions are close to the global optimum. In contrast to supervised machine learning (ML) techniques, we do not need any kind of labeled data in the training phase. The model free approach and the unsupervised nature of RL therefore reduce the required computational effort and has a better scalability as a consequence.

We consider a multi-user two-ray ground reflection scenario with unknown distances between transmitter and receivers. By using two frequencies per user in parallel, we can mitigate possible destructive interference and ensure ultra-reliability with only very limited knowledge at the transmitter. In this work, we consider the problem of assigning two frequencies to each receiver in a multi-user communication system such that the average minimum receive power is maximized. In order to solve this problem, we introduce a generalization of the quadratic multiple knapsack problem to include heterogeneous profits and develop an algorithm to solve it. Compared to random frequency assignment, we report a gain of around 6dB in numerical simulations.

We introduce a phase hopping scheme for reconfigurable intelligent surfaces (RISs) in which the phases of the individual RIS elements are randomly varied with each transmitted symbol. This effectively converts slow fading into fast fading. We show how this can be leveraged to significantly improve the outage performance especially for small outage probabilities without channel state information (CSI) at the transmitter and RIS. Furthermore, the same result can be accomplished even if only two possible phase values are available. Since we do not require perfect CSI at the transmitter or RIS, the proposed scheme has no additional communication overhead for adjusting the phases. This enables robust ultra-reliable communications with a reduced effort for channel estimation.

Modern applications of wireless communication systems often have strict performance requirements. Due to its random nature, channel fading is one of the most limiting factors to provide such guarantees. Even if knowledge about the statistics of the individual links to each antenna is available, there usually are additional uncertainties, e.g., imperfect channel-state information (CSI) or dependency uncertainty between multiple fading links. In this work, we consider the latter and show that a rearrangement algorithm can be applied to calculate the minimum and maximum ergodic capacity for fast fading channels when only the marginal fading distributions are known but the joint distribution is unknown. The algorithm can be used for any number of channels and supports arbitrary marginal distributions. The results are useful for communication system designers, e.g., using the worst-case ergodic capacity for robust system design.

The amount of sensitive data, which is transmitted wirelessly will increase with future technologies. This raises many questions about secure data transmission. Besides cryptography, information-theoretic security gained increasing attention over the recent years. Among others, it deals with the problem of secure data transmission on the physical layer to a legitimate receiver (Bob) in the presence of an eavesdropper (Eve). In this work, we investigate upper and lower bounds on the secrecy outage probability for slowly-fading wiretap channels with an arbitrary dependency structure between the fading channels to Bob and Eve. Both cases of absence of channel-state information at the transmitter (CSI-T) and availability of CSI-T of only the main channel to the legitimate receiver are considered. Furthermore, we derive explicit expressions for the upper and lower bounds for Rayleigh fading and compare them to the case of independent channels. The joint distribution of the legitimate and eavesdropper channels has a tremendous impact on the achievable secrecy outage probability. The bounds enable developing guaranteed secrecy schemes by only measuring the marginal channel distributions.

With emerging wireless technologies like 6G, many new applications like autonomous systems evolve which have strict demands on the reliability and latency of data communications. In the scenario of the commonly investigated independent slow fading links, the zero-outage capacity (ZOC) is zero and retransmissions are therefore inevitable. In this work, we show that a positive ZOC can be achieved under the same setting of slow fading with constant transmit power and without perfect channel state information at the transmitter, if the joint distribution of the channel gains follows certain structures. This allows reliable reception without any outages, thus not requiring retransmissions. Based on a systematic copula approach, we show that there exists a set of dependency structures for which positive ZOCs can be achieved for both maximum ratio combining (MRC) and selection combining (SC). We characterize the maximum ZOC within a finite number of bits. The results are evaluated explicitly for the special cases of Rayleigh fading and Nakagami-$m$fading in order to quantify the ZOCs for common fading models.

The design of reliable and secure codes with finite block length is an important requirement for industrial machine type communications. In this work, we develop an autoencoder for the multi-mode fiber wiretap channel taking into account the error performance at the legitimate receiver and the information leakage at potential eavesdroppers. The estimate of the mutual information leakage includes AWGN and fading channels. The code design is tailored to the specific channel setup where the eavesdropper experiences a mode dependent loss. Numerical simulations illustrate the performance and show a Pareto improvement of the proposed scheme compared to the state-of-the-art polar wiretap codes.

Unreliable fading wireless channels are the main challenge for strict performance guarantees in mobile communications. Diversity schemes including massive number of antennas, huge spectrum bands and multi-connectivity links are applied to improve the outage performance. The success of these approaches relies heavily on the joint distribution of the underlying fading channels. In this work, we consider the ε-outage capacity of slowly fading wireless diversity channels and provide lower and upper bounds for fixed marginal distributions of the individual channels. This answers the question about the best and worst case outage probability achievable over n fading channels with a given distribution, e.g., Rayleigh fading, but not necessarily statistically independent. Interestingly, the best-case joint distribution enables achieving a zero-outage capacity greater than zero without channel state information at the transmitter for n ≥ 2. All results are specialized to Rayleigh fading and compared to the standard assumption of independent and identically distributed fading component channels. The results show a significant impact of the joint distribution and the gap between worstand best-case can be arbitrarily large.

In industrial machine type communications, an increasing number of wireless devices communicate under reliability, latency, and confidentiality constraints, simultaneously. From information theory, it is known that wiretap codes can asymptotically achieve reliability (vanishing block error rate (BLER) at the legitimate receiver Bob) while also achieving secrecy (vanishing information leakage (IL) to an eavesdropper Eve). However, under finite block length, there exists a tradeoff between the BLER at Bob and the IL at Eve. In this work, we propose a flexible wiretap code design for degraded Gaussian wiretap channels under finite block length, which can change the operating point on the Pareto boundary of the tradeoff between BLER and IL given specific code parameters. To attain this goal, we formulate a multi-objective programming problem, which takes the BLER at Bob and the IL at Eve into account. During training, we approximate the BLER by the mean square error and the IL by schemes based on Jensen's inequality and the Taylor expansion and then solve the optimization problem by neural network autoencoders. Simulation results show that the proposed scheme can find codes outperforming polar wiretap codes (PWC) with respect to both BLER and IL simultaneously. We show that the codes found by the autoencoders could be implemented with real modulation schemes with only small losses in performance.

Unreliable fading wireless channels are the main challenge for strict performance guarantees in mobile communications. Diversity schemes including massive number of antennas, huge spectrum bands and multi-connectivity links are applied to improve the outage performance. The success of these approaches relies heavily on the joint distribution of the underlying fading channels. In this work, we consider the ε -outage capacity of slowly fading wireless diversity channels and provide lower and upper bounds for fixed marginal distributions of the individual channels. This answers the question about the best and worst case outage probability achievable over n fading channels with a given distribution, e.g., Rayleigh fading, but not necessarily statistically independent. Interestingly, the best-case joint distribution enables achieving a zero-outage capacity greater than zero without channel state information at the transmitter for n ≥ 2 . Furthermore, the results are applied to characterize the worst- and best-case joint distribution for zero-outage capacity with perfect channel state information everywhere. All results are specialized to Rayleigh fading and compared to the standard assumption of independent and identically distributed fading component channels. The results show a significant impact of the joint distribution and the gap between worst- and best-case can be arbitrarily large.

With an increasing number of wireless devices, the risk of being eavesdropped increases as well. From information theory, it is well known that wiretap codes can asymptotically achieve vanishing decoding error probability at the legitimate receiver while also achieving vanishing leakage to eavesdroppers. However, under finite blocklength, there exists a tradeoff among different parameters of the transmission. In this work, we propose a flexible wiretap code design for Gaussian wiretap channels under finite blocklength by neural network autoencoders. We show that the proposed scheme has higher flexibility in terms of the error rate and leakage tradeoff, compared to the traditional codes.