Federated learning (FL) enables collaborative model training without sharing raw data, making it attractive for privacy-sensitive domains, e.g., healthcare, finance, and IoT. A major obstacle, however, is the potential heterogeneity of tabular data across clients, in practical settings, where schema mismatches and incompatible feature spaces prevent straightforward aggregation. To address this challenge, this paper proposes FedLLM-Align, a federated learning framework that leverages pretrained transformer based language models for feature extraction. Towards this objective, FedLLM-Align serializes tabular records into text and derives semantically aligned embeddings from a pretrained LLM encoder, e.g, DistilBERT, facilitating lightweight local classifier heads that can be trained in a federated manner using standard aggregation schemes, e.g., FedAvg, while keeping all raw data records local. To quantify the merits and trade-offs of FedLLM-Align, we evaluate the proposed framework on binary classification tasks from two different domains: i) Coronary heart disease prediction on partitioned Framingham Heart Study data, and ii) Customer churn prediction on a financial dataset. FedLLM-Align outperforms state-of-the-art baselines by up to 25

Cardiovascular diseases (CVD) cause over 17 million deaths annually worldwide, highlighting the urgent need for privacy-preserving predictive systems. We introduce FedCVD++, an enhanced federated learning (FL) framework that integrates both parametric models (logistic regression, SVM, neural networks) and non-parametric models (Random Forest, XGBoost) for coronary heart disease risk prediction. To address key FL challenges, we propose: (1) tree-subset sampling that reduces Random Forest communication overhead by 70 Evaluated on the Framingham dataset (4,238 records), FedCVD++ achieves state-of-the-art results: federated XGBoost (F1 = 0.80) surpasses its centralized counterpart (F1 = 0.78), and federated Random Forest (F1 = 0.81) matches non-federated performance. Additionally, our communication-efficient strategies reduce bandwidth consumption by 3.2X while preserving 95 Compared to existing FL frameworks, FedCVD++ delivers up to 15

Considering key technology enablers for digital twin applications throughout the Arab world.

The rapid development of the smart grid has intensified research efforts towards automating power transmission system inspections using unmanned intelligent vehicles (UIVs). A major challenge hindering their large-scale deployment is the short operational endurance of these vehicles. This paper addresses this limitation by proposing the use of an energy transmitter (ET) and a reconfigurable intelligent surface for wirelessly charging the vehicles through radio frequency (RF) signals. Additionally, we tackle the nonlinearity of RF-to-direct current (DC) conversion, which shows efficiency saturation at high power levels, by introducing a novel power-splitting (PS) energy harvesting (EH) receiver architecture in the inspection vehicles. The inspection tasks are divided into multiple time slots, and a power transfer efficiency (PTE) maximization problem is formulated to jointly optimize the beamforming design at both the transmitter and the RIS, as well as the transmit power allocation and receiving power splitter settings for each time slot. This approach aims to meet the energy requirements of the unmanned inspection vehicles while adhering to the total available power constraint at the ET. To solve the non-convex PTE maximization problem, we propose an efficient solution utilizing alternating optimization, the Dinkelbach’s method, and successive convex approximation, considering the unique properties of channel structure and finite phase resolution. Numerical results demonstrate that our proposed system and design significantly extend the lifespan of inspection vehicles and achieve high energy efficiency in a low-complexity manner.

This paper presents FedCVD, a federated learning model designed for predicting cardiovascular disease (CVD) by employing logistic regression and Support Vector Machine (SVM) algorithms. FedCVD utilizes the privacy and scalability advantages offered by federated learning to facilitate collaborative model training using decentralized patient data, ensuring confidentiality. To evaluate the effectiveness of the proposed model, experiments were conducted using the 10-year risk of coronary heart disease Kaggle dataset. To address data imbalance challenges, three techniques—Random Over Sampling, Random Under Sampling, and Synthetic Minority Oversampling Technique (SMOTE)—were explored. The study demonstrates promising performance,For the federated logistic regression with SMOTE achieving an AUC value of 0.7048. Comparative analysis with a centralized logistic regression model shows competitive results, with an AUC value of 0.7081 using Random Over Sampling. For the federated SVM model, an AUC value of 0.7340 is achieved using Random Under Sampling. In comparison, a centralized machine learning approach utilizing SVM and Random Over Sampling achieves an AUC value of 0.6962. These findings highlight the effectiveness of the proposed federated learning approach, surpassing the performance of centralized machine learning models for CVD prediction.

Cardiovascular diseases (CVD) have emerged as the leading cause of death globally. High cost Electrocardiogram (ECG) equipment challenges CVD diagnosis in developing countries, especially in rural, underserved areas. This paper presents a novel approach for mapping patient-friendly Ballistocardiogram (BCG) waveforms to ECG waveforms needed for CVD diagnosis, using machine learning. This is inspired by the inherent corre-lation between the two waveforms originating from the same person's heart. More specifically, we propose a Deep Learning (DL) model composed of Convolutional Neural Networks (CNNs) and Long-Short Term Memory (LSTM) for BCG-to-ECG signal mapping. To train and test the proposed model, a bed-based ballistocardiography public dataset is employed. For performance evaluation, we employ the Pearson Correlation Coefficient (PCC) as a widely-accepted metric for quantitatively assessing signals' correlation. Our proposed approach is proven effective as evidenced by achieving a PCC of 0.97, on average, when correlating our deep learning-generated ECG waveform to the real, measured ECG waveform (groundtruth). BCG-to-ECG signal mapping opens an ample room for the integration of ECG's rich diagnostic knowledge and the BCG's continuous monitoring for early diagnosis of cardiovascular diseases, in a cost-effective manner.

Generative Adversarial Networks (GANs) exhibit great potential in many areas. In this paper, we explore their potential in multi-step time series forecasting. To the extent of our knowledge, this task has not been extensively researched yet, possibly due to its unique challenges when trying to model the original temporal behavior of the data. We propose a model for concrete multi-step forecasting where we mix the generative power of the unsupervised GAN loss with the deterministic prediction capabilities of supervised losses. We do this in a rather simple, sequential manner that proves to be helpful for both components of the architecture. The unsupervised component does its job by offering multiple generated predictions that follow the temporal dynamics of the time series, while the supervised component acts as a prediction selector that inspects the provided outputs and creates the most accurate one. We apply this approach in the energy sector, particularly using real industry data on oil well production, as provided to us by Raisa Energy. This learning approach leverages the generative component to provide superior results to those of the supervised counterpart. The approach also stabilizes the overall training, thereby improving the results and providing a more reliable training process.

Due to several applications, including offshore oil field development and monitoring, oceanographic data collection, environmental monitoring, disaster avoidance, and port security, underwater acoustic (UWA) communications have attracted considerable attention in recent years. A communication system must be developed with a thorough understanding of the underlying channel and its characteristics. This paper introduces a novel channel model and simulator for the UWA channel. The channel model is empirically based on statistical analysis for the well-recognized KAM11 experiments. The proposed channel modeling approach offers a comprehensive statistical characterization of all the parameters used to construct the UWA channels.

In this paper we introduce a scalable, privacy-preserving, federated learning framework, coined FLoBC, based on the concept of distributed ledgers underlying blockchains. This is motivated by the rapid growth of data worldwide, especially decentralized data which calls for scalable, decenteralized machine learning models which is capable of preserving the privacy of the data of the participating users. Towards this objective, we first motivate and define the problem scope. We then introduce the proposed FLoBC system architecture hinging on a number of key pillars, namely parallelism, decentralization and node update synchronization. In particular, we examine a number of known node update synchronization policies and examine their performance merits and design trade-offs. Finally, we compare the proposed federated learning system to a centralized learning system baseline to demonstrate its performance merits. Our main finding in this paper is that our proposed decentralized learning framework was able to achieve comparable performance to a classic centralized learning system, while distributing the model training process across multiple nodes without sharing their actual data. This provides a scalable, privacy-preserving solution for training a variety of large machine learning models. Graphical abstract

A Digital Twin is a virtual replica of a physical asset that can be used to simulate, monitor, and control its behavior in real-time. It has been increasingly used to better understand what is produced versus what was designed to bridge the gap between design and execution and to verify complex cyber-physical systems and assess their operation, especially under hypothetical scenarios. In this paper, we propose a novel architecture and implementation of a digital twin for a popular automated guided vehicle (AGV), namely TurtleBot3, which is a robot kit with open-source software. The design and development of the AGV Digital Twin architecture are introduced. This includes simulating the environment using Gazebo, modeling the AGV using a functional mock-up unit (FMU), monitoring the integrated sensors, and actuating the behavior of the virtual AGV and its physical counterpart using a dockerized private cloud with the aid of Robot Operating System (ROS). Along with providing a reference prototype for interested researchers and practitioners, the performance of the proposed system based on dockerized private cloud is contrasted to AWS services in terms of data recording latency as well as the inference time of MobilenetV1 on different setups. The results show that the proposed architecture for Digital Twins can effectively improve the efficiency and flexibility of AGV operation and provide valuable insights for maintenance, optimization, and testing "what-if" scenarios, with an average mean difference percentage of 4.1% and 6.46% for the Digital Twin and Physical Twin speed and position similarity. The proposed architecture is also 9.15 and 3.15 times faster for object detection inference time when compared to private cloud and edge devices, respectively. Additionally, it outperforms the public cloud in data transmission with a 5.5 times speed up.

This paper studies the Multi-Input-Multi-Output (MIMO) interference networks with arbitrary number of transmitters and receivers, where both the transmitters and receivers are equipped with caches. The main goal is to design content placement and delivery schemes that minimize the worst case normalized delivery time (NDT). First, we propose a delivery scheme for the cache-aided Single-Input-Multiple-Output (SIMO) interference networks. Then, we obtain the achievable NDT of the cache-aided MIMO interference networks using the decomposition property of splitting each multi-antenna transmitter into multiple single antenna transmitters. Furthermore, we derive an information-theoretic bound on the optimal NDT of the cache-aided MIMO interference network. Analytical results show that the proposed scheme is within a multiplicative gap of 2 from the derived lower bound independent of all system parameters for any uncoded cache placement scheme. We also derive a novel delivery scheme for the cache-aided Multi-Input-Single-Output (MISO) interference network outperforming our proposed scheme for the cache-aided MIMO interference network. The numerical results show the superiority of our proposed scheme over the state-of-the-art schemes in the literature. Our results show that the coded caching gain has a more significant contribution in reducing the transmission latency than the spatial multiplexing gain. Our results indicate that the receive-antennas become more effective in reducing the NDT than the transmit-antennas in the presence of caches at the receiver-side. In addition, we show that increasing the number of transmit-antennas has a higher gain in reducing the NDT than adding more transmitters in the cache-aided MISO interference network.

In this paper, we propose an architecture, design and build a prototype of a novel IoT system with intelligence, distributed at multiple tiers including the network edge. Our proposed architecture hosts a modular, three-tier IoT system including the edge, gateway (fog) and cloud tiers. The proposed system relies on data acquired by edge devices to realize a distributed machine learning model and achieve timely response at the edge using a lightweight machine learning model. In addition, it employs more sophisticated machine learning models at the higher fog and cloud tiers for wider-scope, long-term decision making. One of the prime objectives of the proposed system is reducing the volume of data transferred across tiers. This is attained through intelligent data filtering at the edge/gateway tiers to distill key events that avail the most relevant data points to higher-tier machine learning models at the gateway and cloud. This, in turn, reduces the outliers and the redundant data that may impact the gateway and cloud models and reduces the inter-tier communications overhead. To demonstrate the merits of our proposed system, we build a proof-of-concept prototype hosting the three tiers, using COTS components and supporting networking technologies. We demonstrate through extensive experiments the merits of the proposed system. A major finding is that our system is capable of achieving prediction performance comparable to the centralized machine learning baseline model, while reducing the inter-tier communications overhead by up to 80%.

The rapid expansion of data worldwide invites the need for more distributed solutions in order to apply machine learning on a much wider scale. The resultant distributed learning systems can have various degrees of centralization. In this work, we demonstrate our solution FLoBC for building a generic decentralized federated learning system using blockchain technology, accommodating any machine learning model that is compatible with gradient descent optimization. We present our system design comprising the two decentralized actors: trainer and validator, alongside our methodology for ensuring reliable and efficient operation of said system. Finally, we utilize FLoBC as an experimental sandbox to compare and contrast the effects of trainer-to-validator ratio, reward-penalty policy, and model synchronization schemes on the overall system performance, ultimately showing by example that a decentralized federated learning system is indeed a feasible alternative to more centralized architectures.

Proactive content caching has been proposed as a promising solution to cope with the challenges caused by the rapid surge in content access using wireless and mobile devices and to prevent significant revenue loss for content providers. In this paper, we propose an end-to-end Deep Learning framework for proactive content caching that models the dynamic interaction between users and content items, particularly their features. The proposed model performs the caching task by building a probability distribution across different content items, per user, via a Deep Neural Network model and supports, both, centralized and distributed caching schemes. In addition, the paper addresses the key question: Do we need an explicit user-item pairs-based recommendation system in content caching? i.e., do we need to develop a recommendation system while tackling the content caching problem? To this end, an end-to-end Deep Learning framework is introduced. Finally, we validate our approach through extensive experiments on a real-world, public data set, coined MovieLens. Our experiments show consistent performance gains against its counterparts, where our proposed Deep Learning Caching module, dubbed as DLC, significantly outperforms state-of-the-art content caching schemes, serving as a baseline. Our code is available here: https://github.com/heshameraqi/ProactiveContent-Caching-with-Deep-Learning.

In this paper, we propose a novel design for a high-speed underwater acoustic (UWA) receiver chain. Towards this objective, we introduce a novel closed-loop, adaptive, single-rate Doppler scale estimation and compensation algorithm and a novel dual-domain channel estimation and equalization technique. We validate our results and evaluate the proposed schemes using extensive simulations based on channel models generated from the analysis of the Kauai Acomms MURI 2011 (KAM11) UWA communications experiment's recording. The extensive simulations show that the proposed algorithms exhibit superior performance compared to the state-of-the-art.

In this paper, we propose a Reinforcement Learning-based MAC layer protocol for cognitive radio networks, based on exploiting the feedback of the Primary User (PU). Our proposed model relies on two pillars, namely an infinite-state Partially Observable Markov Decision Process (POMDP) to model the system dynamics besides a queuing-theoretic model for the PU queue, where the states represent whether a packet is delivered or not from the PU's queue and the PU channel state. Based on the stability constraint for the primary user queue, the quality of service (QoS) for the PU is guaranteed. Towards the paper's objectives, three Reinforcement Learning approaches are studied, namely Q-Learning, Deep Q-Network (DQN), and Deep Deterministic Policy Gradient (DDPG). Our ultimate objective is to enhance the channel access techniques in the MAC protocols by solving the POMDP without any prior knowledge of the environment.

In this paper, we propose a reinforcement learning (RL) approach to design an access scheme for secondary users (SUs) in a cognitive radio (CR) network. In the proposed scheme, we introduce a deep Q-network to enable SUs to access the primary user (PU) channel based on their past experience and the history of the PU network's automatic repeat request (ARQ) feedback. In essence, SUs cooperate to avoid collisions with other SUs and, more importantly, with the PU network. Since SUs cannot observe the state of the PUs queues, they partially observe the system's state by listening to the PUs' ARQ packets. To model this system, a Partially Observable Markov Decision Process (POMDP) is adopted, and an RL deep Q-network is employed for the SUs to learn the best actions. A comparative study between the proposed scheme with baseline schemes from the literature is presented. We also compare the proposed scheme with the perfect sensing system (which constitutes an upper bound on the performance) and the system exploiting only the last ARQ feedback. Our results show that the proposed RL based access scheme yields comparable performance to the baseline ARQ-based access schemes, yet, with minimal knowledge about the environment compared to the baseline which assumes perfect knowledge of key system parameters, e.g., PUs arrival rates. On the contrary, our proposed scheme autonomously learns these parameters and, hence, dynamically adapts to their variation.

research-article Open Access Share on Networking research for the Arab world: from regional initiatives to potential global impact Authors: Basem Shihada King Abdullah University of Science and Technology, Thuwal, Saudi Arabia King Abdullah University of Science and Technology, Thuwal, Saudi ArabiaView Profile , Tamer Elbatt The American University in Cairo, New Cairo, Egypt The American University in Cairo, New Cairo, EgyptView Profile , Ahmed Eltawil King Abdullah University of Science and Technology, Thuwal, Saudi Arabia King Abdullah University of Science and Technology, Thuwal, Saudi ArabiaView Profile , Mohammad Mansour American University of Beirut, Lebanon American University of Beirut, LebanonView Profile , Essaid Sabir Hassan II University of Casablanca, Casablanca, Morocco Hassan II University of Casablanca, Casablanca, MoroccoView Profile , Slim Rekhis University of Carthage, Tunis, Tunisia University of Carthage, Tunis, TunisiaView Profile , Sanaa Sharafeddine Lebanese American University, Beirut, Lebanon Lebanese American University, Beirut, LebanonView Profile Authors Info & Claims Communications of the ACMVolume 64Issue 4April 2021 pp 114–119https://doi.org/10.1145/3447748Online:22 March 2021Publication History 1citation1,861DownloadsMetricsTotal Citations1Total Downloads1,861Last 12 Months346Last 6 weeks20 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my Alerts New Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteView all FormatsPDF

In this paper, we study a wireless powered cellular network (WPCN) supported with network coding capability. In particular, we consider a network consisting of k cellular users (CUs) served by a hybrid access point (HAP) that takes over energy transfer to the users on top of information transmission over both the uplink (UL) and downlink (DL). Each CU has k+1 states representing its communication behavior, and collectively are referred to as the user demand profile. Opportunistically, when the CUs have information to be exchanged through the HAP, it broadcasts this information in coded format to the exchanging pairs, resulting in saving time slots over the DL. These saved slots are then utilized by the HAP to prolong the network lifetime and enhance the network throughput. We quantify, analytically, the performance gain of our network-coded WPCN over the conventional one, that does not employ network coding, in terms of network lifetime and throughput. We consider the two extreme cases of using all the saved slots either for energy boosting or throughput enhancement. In addition, a lifetime/throughput optimization is carried out by the HAP for balancing the saved slots assignment in an optimized fashion, where the problem is formulated as a mixed-integer linear programming optimization problem. Numerical results exhibit the network performance gains from the lifetime and throughput perspectives, for a uniform user demand profile across all CUs. Moreover, the effect of biasing the user demand profile of some CUs in the network reveals considerable improvement in the network performance gains.