In this work, we investigate the problem of an online trajectory design for an Unmanned Aerial Vehicle (UAV) in a Federated Learning (FL) setting where several communities exist, each defined by a unique task to be learned. In this setting, spatially distributed devices belonging to each community collaboratively contribute towards training their community model via wireless links provided by the UAV. Accordingly, the UAV acts as a mobile orchestrator coordinating the transmissions and the learning schedule among the devices in each community, intending to accelerate the learning process of all tasks. We propose a heuristic metric as a proxy for the training performance of the different tasks. Capitalizing on this metric, a surrogate objective is defined which enables us to jointly optimize the UAV trajectory and the scheduling of the devices by employing convex optimization techniques and graph theory. The simulations illustrate the out-performance of our solution when compared to other handpicked static and mobile UAV deployment baselines.

Multi-hop relaying with mobile relay nodes mounted at the top of train cabins is a promising technology to overcome the high propagation loss, blockage and mobility sensitivity in millimeter wave band communication for high speed train. In this context, we aim to maximize the data rate for passenger-oriented applications under multi-hop mobile relay architecture. Adopting the system deployment and antenna configuration in 3GPP, based on the transmitting power configuration and instantaneous train position, we propose an approach to optimize the average transmission data rate by selecting the optimal multi-hop relaying scheme. The analysis and numerical results exhibit that the proposed multi-hop relaying strategies can outperform a single hop 5G link communication under certain transmitting power regimes and train position ranges, which can serve as feasibility analysis and guideline for a possible architecture in next generation train radio deployment.

In this work, we consider the sum rate performance of joint processing coordinated multi-point transmission network (JP-CoMP, a.k.a Network MIMO) in a so-called distributed channel state information (D-CSI) setting. In the D-CSI setting, the various transmitters (TXs) acquire a local, TX-dependent, estimate of the global multi-user channel state matrix obtained via terminal feedback and limited backhauling. The CSI noise across TXs can be independent or correlated, so as to reflect the degree to which TXs can exchange information over the backhaul, hence allowing to model a range of situations bridging fully distributed and fully centralized CSI settings. In this context we aim to study the price of CSI distributiveness in terms of sum rate at finite SNR when compared with conventional centralized scenarios. We consider the family of JP-CoMP precoders known as regularized zero-forcing (RZF). We conduct our study in the large scale antenna regime, as it is currently envisioned to be used in real 5G deployments. It is then possible to obtain accurate approximations on so-called deterministic equivalents of the signal to interference and noise ratios. Guided by the obtained deterministic equivalents, we propose an approach to derive a RZF scheme that is robust to the distributed aspect of the CSI, whereby the key idea lies in the optimization of a TX-dependent power level and regularization factor. Our analysis confirms the improved robustness of the proposed scheme with respect to CSI inconsistency at different TXs, even with moderate number of antennas and receivers (RXs).

This paper analyzes the impact of the imperfect Channel State Information (CSI) on the Gaussian input mutual information between the channel input and output. The existing bounds of the mutual information under imperfect CSI depend on the estimated channel realization, while for physical layer abstraction, the goal is to provide such bounds as a function of the perfect CSI variables without performing a complicated channel estimation procedure to obtain the imperfect channel. Indeed, the main motivation is to provide an efficient physical layer (PHY) abstraction tool in large system level simulations. In particular, performance taking into account CSI inaccuracies can be obtained without relying on time-consuming Monte-Carlo simulations. The PHY abstraction model with imperfect CSI is derived for the Multiple-Input Multiple-Output Orthogonal Frequency Division Multiplexing (MIMO-OFDM) systems and particularly target the 3GPP-5G millimeter wave Massive-MIMO system level simulations. The optimization of pilot density and power boosting of OFDM modulations according to the proposed bounds lead to the same results as the one proposed in the 3GPP standards, which illustrates the applicability of the results presented in this paper.

Hybrid precoding design for a wideband multiuser multiple input multiple output (MIMO) millimeter wave (mmWave) system is more challenging than the narrowband design. This is due to the fact that the radio frequency (RF) precoder and combiner are frequency non-selective and they should be optimized with respect to all scheduled users on all allocated subcarriers. In this work, we adopt the separate stage hybrid precoding strategy and propose two algorithms with good performance and complexity trade-off. Based on the proposed strategies of minimal equivalent channel gain maximization or average wideband sum rate maximization, the non-convex RF beamforming can be solved first by a semi-definite relaxation and codebook projection. In the second stage, conventional sum rate maximization digital precoder/combiner can be obtained based on smaller dimension equivalent channel feedback. Simulations show that the proposed algorithms outperform the state of art algorithms in multiple receive antennas setting.

Precoding for millimeter wave (mmWave) system is different from conventional sub-6GHz system due to the architecture and hardware constraints. In this work, we propose a new two-stage hybrid sum rate maximization precoding algorithm for a downlink mmWave multi-user multiple data streams system with large number of antennas at both the transmitter (TX) and the receiver (RX) side. The proposed algorithm will first select the radio frequency (RF) domain precoder and combiners based on the channel state information (CSI) locally available at each user. Conventional sum rate maximization base band precoding is then performed at TX with CSI feedback of the smaller dimension equivalent user channel. Simulation shows that the proposed algorithm offers higher sum rates compared with the state of the art two-stage hybrid precoding algorithm in the literature. It also approaches the sum rate performance of the conventional full digital sum rate maximization design.

Obtaining accurate global channel state information (CSI) at multiple transmitter devices is critical to the performance of many coordinated transmission schemes. Practical CSI local feedback often leads to noisy and partial CSI estimates at each transmitter. With rate-limited bi-directional backhaul, transmitters have the opportunity to exchange few CSI-related bits to establish global channel state information at transmitter (CSIT). This paper investigates possible strategies toward this goal. We propose a novel decentralized algorithm that produces minimum mean square error-optimal global channel estimates at each device from combining local feedback and information exchanged through backhauls. The method adapts to arbitrary initial information topologies and feedback noise statistics and can do that with a combination of closed-form and convex approaches. Simulations for coordinated multi-point transmission systems with two or three transmitters exhibit the advantage of the proposed algorithm over conventional CSI exchange mechanisms when the coordination backhauls are limited.

This paper considers the problem of transmitter (TX) cooperation with distributed channel state information (CSI), where two or more transmitters seek to jointly precode messages while communicating over a rate-limited coordination link. Specifically we address a so-called master-slave scenario where one master (M-) TX is endowed with perfect CSI while K slave (S-) TXs have zero prior CSI information. We are interested in possible strategies for how the M-TX may efficiently guide the S-TXs over the coordination links so as to maximize the network's figure of merit. Strategies related to communicating quantized CSI or quantized precoding decisions are described and compared. Optimal and sub-optimal low complexity approaches are shown, exhibiting gains over conventional methods.