In the literature, machine learning (ML) has been implemented at the base station (BS) and user equipment (UE) to improve the precision of downlink channel state information (CSI). However, ML implementation at the UE can be infeasible for various reasons, such as UE power consumption. Motivated by this issue, we propose a CSI learning mechanism at BS, called CSILaBS, to avoid ML at UE. To this end, by exploiting channel predictor (CP) at BS, a light-weight predictor function (PF) is considered for feedback evaluation at the UE. CSILaBS reduces over-the-air (OTA) feedback overhead, improves CSI quality, and lowers the computation cost of UE. Besides, in a multiuser environment, we propose various mechanisms to select the feedback by exploiting PF while aiming to improve CSI accuracy. We also address various ML-based CPs, such as NeuralProphet (NP), an ML-inspired statistical algorithm. Furthermore, inspired to use a statistical model and ML together, we propose a novel hybrid framework composed of a recurrent neural network and NP, which yields better prediction accuracy than individual models. The performance of CSILaBS is evaluated through an empirical dataset recorded at Nokia Bell-Labs. The outcomes show that ML elimination at UE can retain performance gains, for example, precoding quality.

Channel state information (CSI) is of pivotal importance as it enables wireless systems to adapt transmission parameters more accurately, thus improving the system's overall performance. However, it becomes challenging to acquire accurate CSI in a highly dynamic environment, mainly due to multi-path fading. Inaccurate CSI can deteriorate the performance, particularly of a massive multiple-input multiple-output system. This paper adapts machine learning for CSI prediction. Specifically, we exploit time-series models of deep learning (DL) such as recurrent neural network (RNN) and bidirectional long-short term memory. Further, we use NeuralProphet (NP), a recently introduced time-series model, composed of statistical components, e.g., autoregressive and Fourier terms, for CSI prediction. Inspired by statistical models, we also develop a novel hybrid framework comprising RNN and NP to achieve better prediction accuracy. The proposed channel predictors performance is evaluated on a real-time dataset recorded at the Nokia Bell-Labs campus in Stuttgart, Germany. Numerical results show that DL brings performance gain when used with statistical models and showcases robustness.

Unmanned aerial vehicles (UAVs) evolution has gained an unabated interest for the use in several applications, such as agriculture, aerial surveillance, goods delivery, disaster recovery, intelligent transportation. The main features of this technology are high coverage, strong line-of-sight (LoS) links, promising throughput, cost-effective and flexible deployment. Currently, the Third Generation Partnership Project (3GPP) is working on the specification of release-17 (R-17) new radio (NR) for non-terrestrial networks (NTN). Therefore, owing to the drastic increase of UAV technology, in this paper, we propose channel state information (CSI) compression and its recovery with the aid of machine learning (ML)-based twin channel predictors. Due to the characteristic of gaining higher LoS communication paths in UAV network, the proposed strategy can bring potential benefits such as over-the-air (OTA)-overhead reduction, minimizing mean-squared-error (MSE) of a channel and maximizing precoding gain. Simulation-based results corroborate the validity of the proposed strategy, which can reap benefits in multiple factors.

In wireless communication, accurate channel state information (CSI) is of pivotal importance. In practice, due to processing and feedback delays, estimated CSI can be outdated, which can severely deteriorate the performance of the communication system. Besides, to feedback estimated CSI, a strong compression of the CSI, evaluated at the user equipment (UE), is performed to reduce the over-the-air (OTA) overhead. Such compression strongly reduces the precision of the estimated CSI, which ultimately impacts the performance of multipleinput multiple-output (MIMO) precoding. Motivated by such issues, we present a novel scalable idea of reporting CSI in wireless networks, which is applicable to both time-division duplex (TDD) and frequency-division duplex (FDD) systems. In particular, the novel approach introduces the use of a channel predictor function, e.g., Kalman filter (KF), at both ends of the communication system to predict CSI. Simulation-based results demonstrate that the novel approach reduces not only the channel mean-squared-error (MSE) but also the OTA overhead to feedback the estimated CSI when there is immense variation in the mobile radio channel. Besides, in the immobile radio channel, feedback can be eliminated, which brings the benefit of further reducing the OTA overhead. Additionally, the proposed method provides a significant signal-to-noise ratio (SNR) gain in both the channel conditions, i.e., highly mobile and immobile.

Hybrid analog and digital BeamForming (HBF) is one of the enabling transceiver technologies for millimeter Wave (mmWave) Multiple Input Multiple Output (MIMO) systems. This technology offers highly directional communication, which is able to confront the intrinsic characteristics of mmWave signal propagation. However, the small coherence time in mmWave systems, especially under mobility conditions, renders efficient Beam Management (BM) in standalone mmWave communication a very difficult task. In this paper, we consider HBF transceivers with planar antenna panels and design a multilevel beam codebook for the analog beamformer comprising flat top beams with variable widths. These beams exhibit an almost constant array gain for the whole desired angle width, thereby facilitating efficient hierarchical BM. Focusing on the uplink communication, we present a novel beam training algorithm with dynamic beam ordering, which is suitable for the stringent latency requirements of the latest mmWave standard discussions. Our simulation results showcase the latency performance improvement and received signal-to-noise ratio with different variations of the proposed scheme over the optimum beam training scheme based on exhaustive narrow beam search.

The exchange of confidential messages is an inherent problem in wireless communication due to the broadcast nature of the radio channel. In this paper, we enhance standard cryptography with information-theoretic techniques by exploiting relays to increase the confidentiality of wireless communication in the presence of one or more eavesdroppers with low-noise receivers. To achieve this, we present a protocol which makes use of relays in two ways. First, the relays re-transmit disjoint encrypted chunks of a message. Second, the relays utilize cooperative jamming techniques to generate pseudo-random signals in order to increase the interference level in the propagation domain. Chunks and interference levels are allocated over relays in such a way that the message can only be decoded within a critical area around the intended receiver. Our simulation results show that this area can be minimized under realistic assumptions on propagation environment and channel knowledge.

In this paper, the problem of interference coordination in heterogeneous networks, comprising macro base stations (MBS) and several underlay small cells (SCS) is analyzed. Such problem is normally solved by coordinating and orthogonalizing the transmissions of the MBS and the SCS in order to reduce the global interference level, at the price of underutilized network resources. To solve this problem, an interference coordination scheme, based on the exploitation of device-to-device (D2D) enabled relaying, and allowing full spectrum usage is presented. This scheme is based on the idea of substituting long-range transmission with short-range D2D communications, which are characterized by low power transmissions, thus introducing low interference to the overall system. The proposed algorithm is compared to standard enhanced inter cell interference coordination (eICIC) scheme and its performance gain is assessed through numerical simulations.

In this paper, we present a framework for including geographical context information within the process of relay selection, where both the relay and destination can be user equipment (UE) hence enabling relaying via device-to-device (D2D) communication. Geographical information, acquired through positioning systems and radio maps, are used to infer pathloss, link budgets and relay generated interference. Thus, this information becomes available at the core network without the aid of extra signaling and channel sounding. Tools from image processing can be exploited to analyze the radio map and determine so called isolated regions. Then, these regions are used by the proposed framework to effectively select the relay. Three different relay selection policies that exploit geographical context via the knowledge of the spatial isolated regions are introduced. Each of these policies is analyzed via simulation and performance is compared with respect to baseline schemes. The pertinence of adopting geographical information is confirmed by the gain in terms of useful power and reduced interference level generated.

In this paper, the problem of resource allocation and interference minimization within a device-to-device (D2D) network is analyzed. A clustered structure for the D2D network is proposed along with a two-level distributed scheme operating both at cluster and link levels. By exploiting the peculiarities of the network, the proposed scheme is shown to efficiently configure the transmission parameters. At regime, the algorithm operating at the cluster-level is proved to minimize the average multiple access interference (MAI) arising in the network. A conjecture linking this interference with the algorithm parameters is presented and validated through numerical simulations. The algorithm operating at the link-level is shown to efficiently set the transmit sub-channel and power level satisfying the quality of service (QoS) requirements for the largest possible set of devices using the minimum amount of power. The pertinence of the proposed scheme is highlighted by numerical simulations validating the theoretical findings and assessing the algorithm performance.

In this paper, a novel game-theoretic model of the complex interactions between network service providers (NSPs) and users in heterogeneous small-cell networks is investigated. In this game, the NSPs selfishly aim at maximizing their profit while, simultaneously, the users seek to optimize their chosen service's quality-price tradeoff. A Stackelberg formulation in which the NSPs act as leaders and the users as followers is proposed. The users' interactions are modeled as a general nonatomic game. The existence of a Wardrop equilibrium (WE) in the users' game is proven, and its expression as a solution of a fixed-point equation is provided (irrespective of the number of NSPs, services offered, pricing policies, and QoS functions). Moreover, a set of sufficient conditions that ensure the uniqueness of the WE is provided. Notably, the uniqueness of the equilibrium for the particular case of congestion games is shown. An algorithm approximating these equilibria is provided and its convergence to an ε-WE is proven. The existence of Nash equilibria for the leaders' game is shown and illustrated via numerical simulations.

In this paper, the problem of channel selection and power control is jointly analyzed in the context of multiple-channel clustered ad-hoc networks, i.e., decentralized networks in which radio devices are arranged into groups (clusters) and each cluster is managed by a central controller (CC). This problem is modeled by game in normal form in which the corresponding utility functions are designed for making some of the Nash equilibria (NE) to coincide with the solutions to a global network optimization problem. In order to ensure that the network operates in the equilibria that are globally optimal, a learning algorithm based on the paradigm of trial and error learning is proposed. These results are presented in the most general form and therefore, they can also be seen as a framework for designing both games and learning algorithms with which decentralized networks can operate at global optimal points using only their available local knowledge. The pertinence of the game design and the learning algorithm are highlighted using specific scenarios in decentralized clustered ad hoc networks. Numerical results confirm the relevance of using appropriate utility functions and trial and error learning for enhancing the performance of decentralized networks.

In this paper, a decentralized iterative algorithm, namely the optimal dynamic learning (ODL) algorithm, is analysed. The ability of this algorithm of achieving a Pareto optimal working point exploiting only a minimal amount of information is shown. The algorithm performance is analysed in a clustered ad hoc network, where radio devices are assumed to operate above a minimal signal to interference plus noise ratio (SINR) threshold while minimizing the global power consumption. Sufficient analytical conditions for ODL to converge to the desired working point are provided, moreover through numerical simulations the ability of the algorithm to configure an interference limited network is shown. The performances of ODL and of a Nash equilibrium reaching algorithm are numerically compared, and their performance as a function of available resources is studied. The gain of ODL is shown to be larger when the amount of available radio resources is scarce.

In this paper, we address the problem of global transmit power minimization in a self-configuring network where radio devices are subject to operate at a minimum signal to interference plus noise ratio (SINR) level. We model the network as a parallel Gaussian interference channel and we introduce a fully decentralized algorithm (based on trial and error) able to statistically achieve a configuration where the performance demands are met. Contrary to existing solutions, our algorithm requires only local information and can learn stable and efficient working points by using only one bit feedback. We model the network under two different game theoretical frameworks: normal form and satisfaction form. We show that the converging points correspond to equilibrium points, namely Nash and satisfaction equilibrium. Similarly, we provide sufficient conditions for the algorithm to converge in both formulations. Moreover, we provide analytical results to estimate the algorithm's performance, as a function of the network parameters. Finally, numerical results are provided to validate our theoretical conclusions.

Since J. Mitola's seminal work in the 90's, Software Defined Radios (SDRs) have been a hot topic in wireless communication research. Though many notable achievements were reported in the field, the scarcity of computational power on general purpose CPUs has always been a limiting factor. If conveniently applied within an SDR context, classical concepts known in computer science as space/time trade-offs can prove helpful when trying to mitigate this problem. This paper presents a novel criterion to design signal- processing software in an SDR terminal that we call Memory Acceleration (MA). The key feature of MA is making extensive use of memory resources of the SDR platform in order to accelerate the most critical signal processing functions. MA provides substantial acceleration factors when applied to conventional SDRs without reducing their peculiar flexibility and appears particularly suited to the implementation of a fully-SW SDR on a general- purpose processor (GPP). As a case study for the application of MA, results about the implementation of the ETSI DVB-T Viterbi decoder are presented. In such a case, MA provides an acceleration factor of 10.4x with respect to a standard, purely-computational implementation while having no impact on error correction performance of the decoder and by making no use of any other typical performance enhancement technique (e.g. low level programming or parallel computation).