As unmanned vehicular networks become more prevalent in civilian and defense applications, the need for robust security solutions grows in parallel. While ROS 2 offers a flexible platform for robotic operations, its security model lacks the adaptability required for dynamic trust management and proactive threat mitigation. To address these shortcomings, we propose a novel framework that integrates containerized ROS 2 nodes with Kubernetes-based orchestration, a dynamic trust management subsystem, and integrability with simulators for real-time and protocol-flexible network simulation. By embedding trust management directly within each ROS 2 container and leveraging Kubernetes, we overcome ROS 2’s security limitations by enabling real-time monitoring and machine learning-driven anomaly detection (via an autoencoder trained on custom data), facilitating the isolation or removal of suspicious nodes. Additionally, Kubernetes policies allow seamless scaling and enforcement of trust-based security rules, mitigating the static constraints of the default ROS 2 security stack. This approach delivers a robust, scalable, and adaptive platform for unmanned vehicle fleets operating in contested or untrusted domains.

FABRIC is a national and international scale research infrastructure built to enable cutting-edge research in a wide range of networking and computing areas and domain sciences that depend on advanced networking and computing capabilities. Since its full operation in October 2023, FABRIC has seen rapidly emerging experiments in different research topics. This paper provides a description of FABRIC’s hardware resources, software services, and an analysis of the characteristics of the emerging experiments that are leveraging FABRIC’s unique capabilities. The analysis showed a wide spectrum of topic areas, frequent emphasis on at-scale and high performance experiments that interact with real facilities and real people, and growing complexity and rigor in tools, data and measurement. Researchers are seen sharing well-developed artifacts with others to enable larger experiments and reproducible research.

Modern cellular communication systems are expected to be capable of delivering high data rates with low latency and ultra-reliability. Using frequencies in the FR2 or millimeter wave (mmWave) range for communication can provide large data rates and cover densely populated areas, but only over short distances as they are susceptible to blockages. Such architectures lack the reliability needed to support the stricter requirements of V2X applications, especially in mobility scenarios. In this work, we study how packet duplication can be used to meet the stringent reliability and latency requirements of modern cellular networks under various mobility scenarios. Simultaneously transmitting multiple instances (duplicates) of a packet over different uncorrelated channels, although resource-consuming, could effectively reduce the packet failure probability without increasing the system complexity. This paper therefore proposes and evaluates novel enhancements to the packet duplication process to improve radio resource utilization and meet requirements under mobility and varying network load conditions.

The fifth generation (5G) networks and beyond need paradigm shifts to realize the increasing demands of next-generation services for high throughputs, low latencies, and reliable communication under various mobility scenarios. However, these promising features have critical gaps that need to be filled before they can be fully implemented for mobile applications in complex environments like smart cities. Even though the sub-6 GHz bands can provide reliable and larger coverage, they cannot provide high data rates with low latencies due to a scarcity of spectrum available in these bands. On the other hand, the extremely limited transmission range of high bandwidth millimeter wave (mmWave) frequencies leads to poor reliability, especially for mobility scenarios. 5G networks using mmWave-based ultra-dense networks (UDN) deployments face challenges in terms of frequent handovers, increasing complexity and cost of deployment, etc. To address the challenges in high density base station deployments, we study and evaluate novel deployment strategies using multi-connectivity and compare their performance with UDN. In this work, we propose and evaluate 5G deployments with multi-connectivity, which can be used to ensure that these 5G networks are able to meet the demanding requirements of future services with efficient resource management.

In recent years, data-intensive applications have been increasingly deployed on cloud systems. Such applications utilize significant compute, memory, and I/O resources to process large volumes of data. Optimizing the performance and cost-efficiency for such applications is a non-trivial problem. The problem becomes even more challenging with the increasing use of containers, which are popular due to their lower operational overheads and faster boot speed at the cost of weaker resource assurances for the hosted applications. In this paper, two containerized data-intensive applications with very different performance objectives and resource needs were studied on cloud servers with Docker containers running on Intel Xeon E5 and AMD EPYC Rome multi-core processors with a range of CPU, memory, and I/O configurations. Primary findings from our experiments include: 1) Allocating multiple cores to a compute-intensive application can improve performance, but only if the cores do not contend for the same caches, and the optimal core counts depend on the specific workload; 2) allocating more memory to a memory-intensive application than its deterministic data workload does not further improve performance; however, 3) having multiple such memory-intensive containers on the same server can lead to cache and memory bus contention leading to significant and volatile performance degradation. The comparative observations on Intel and AMD servers provided insights into trade-offs between larger numbers of distributed chiplets interconnected with higher speed buses (AMD) and larger numbers of centrally integrated cores and caches with lesser speed buses (Intel). For the two types of applications studied, the more distributed caches and faster data buses have benefited the deployment of larger numbers of containers.

The fifth generation (5G) networks and beyond are key to meeting the exponentially increasing demands of next generation services for high throughput and reliable low latency communication under various mobility scenarios. These promising features have critical gaps to be filled before they can be fully implemented for mobile applications in complex environments like smart cities. Millimeter wave (mmWave) communications is a key enabler for a significant increase in the performance of these networks. However, due to the extremely limited transmission range of mmWave frequencies, 5G network deployments are designed to have several small cells operating in the mmWave frequency range using Ultra-dense networking (UDN) techniques to provide continuous coverage. But, such deployments not only face challenges in terms of a higher number of handovers, higher latency, lower reliability, and higher interference levels but also in terms of site acquisition, logistics, unbalanced load distributions, and power requirements. Multi-connectivity can improve the performance of UDNs and provide better deployment strategies as it provides multiple simultaneous links between the User Equipment (UE) and base stations. In such systems, packet duplication can be used to meet the stringent reliability and latency requirements of modern cellular networks as data packets are duplicated and transmitted concurrently over two or more independent links. The downside to packet duplication, however, is the increased usage of resources like spectrum, power, etc. In this work, we explore and analyze different techniques that could aid in reducing radio resource utilization without sacrificing the improvements in reliability and latency observed through packet duplication. To perform this study, we develop a novel 5G deployment with new radio dual connectivity (NR-DC) and packet duplication to improve reliability. We then analyze possible enhancements in the system to improve radio resource utilization when packet duplication is implemented. In this article, we propose and evaluate this novel 5G network deployment with multi-connectivity using Simu5G network simulator for enabling future 5G systems. The proposed 5G architecture is shown to meet the requirements of next generation applications. Our simulations show that the proposed techniques improve the throughput by up to 165.72%, the latency by up to 91.22%, and the packet loss decreases to near zero compared to a single link system.

Vehicles-to-Everything or V2X communications provide attractive advantages in achieving reliable and high-performance connectivity amongst ground and aerial military vehicles. The 5G New Radio (NR) based cellular-V2X (C-V2X) technology, can support wide coverage areas with higher data rates and lower latencies needed for demanding military applications ranging from real-time sensing to navigation of autonomous military ground vehicles. Millimeter wave technology (mmWave) is critical to meet such throughput and latency requirements. However, mmWave links have a low transmission range and are often subject to blockages due to factors like weather, terrain, etc. that make them unreliable. Multi-connectivity with packet duplication can be used to enhance the reliability and latency by transmitting concurrently over independent links between a mobile device and multiple base stations. We propose and evaluate a novel method based on new radio dual connectivity (NR-DC) and packet duplication techniques to achieve reliable communication between military ground vehicles, especially in mobility scenarios. We further propose and analyze a Channel State Information Reference Signal Received Quality (CSI-RSRQ) based duplication strategy to improve the system's radio resource utilization. Channel State Information Reference Signal symbols in downlink transmissions are used to accurately compute the CSI-RSRQ values of the radio channel in real-time. This is critical on the battlefield for real-time awareness and adaptive control in fast-changing environments. Prototyped in the Simu5G network simulator and MATLAB, our results show packet duplication achieved less than 5 milliseconds of latency with zero packet loss under mobility.

The fifth generation (5G) and beyond networks have the potential to meet the increasing demands of emerging applications and industry verticals for high throughputs and reliable low latency communication under various mobility scenarios. Also, there is a great demand for new user experiences that would require greater uplink performance from these networks. To meet these expectations from 5G networks, critical gaps need to be filled before they can be fully implemented for mobile applications in complex environments like smart cities. Millimeter wave frequencies can significantly increase the performance of these networks, but they also suffer from reliability issues due to limited transmission range. Many uplink enhancement technologies, like multi-connectivity, carrier aggregation, etc., have been proposed to improve the performance, but they all either add to the user equipment (UE) complexity or increase the deployment costs. We develop a novel multilink scheme for the uplink where the UE performs only a single transmission on a common channel, and every base station that receives this signal would accept and process it. Our solution reduces frequent handovers and allows high UE mobility with no additional complexity. We propose and evaluate this novel 5G deployment with multilink using the Simu5G network simulator for enabling future 5G systems. The proposed 5G architecture is shown to meet the requirements of next-generation applications. Our simulations show that the proposed technique improves the throughput by up to 54%, the latency by up to 27.61%, and the reliability by up to 91.36% compared to a single link system.

The fifth generation (5G) networks and beyond enable technologies that are key to meeting the exponentially increasing demands of next generation services for reliable low latency communication under various mobility scenarios. Millimeter wave technology is a key enabler for 5G use cases; however, mmWave links have a lower transmission range and are subject to blockages. Multi-connectivity (MC) enhances the system performance by providing multiple simultaneous links between the user equipment (UE) and the base stations. In such systems, packet duplication can be used to meet the stringent reliability and latency requirements of modern cellular networks as data packets are duplicated and transmitted concurrently over two independent links. We propose and evaluate a novel 5G deployment with new radio dual connectivity (NR-DC) and packet duplication using the Simu5G network simulator for enabling future 5G systems, especially in mobility scenarios. We also analyze possible enhancements in the system to improve the radio resource utilization when packet duplication is implemented. Our simulation results show the improvement in reliability and latency that NR-DC with optimized packet duplication could bring to next generation applications and services like enhanced vehicle-to-everything (V2X) scenarios.

Operational medical environments require reliable hands-free solutions to extract data from audio captured under noisy scenarios during rescue missions and provide timely information. However, approaches using automatic speech recognition (ASR) and natural language processing (NLP) techniques are complex as these conversations have a wide range of noise, involve medical terms from multiple speakers, and occur in high-stress environments, among others. These are further complicated by the lack of large training datasets for operational medical scenarios. To address these issues, we developed a platform that enables resilient hands-free data collection, preserves complete documentation through stages of care, and presents the information in near real-time, critical for the medical operation. Our work uniquely focused on systematic evaluation and improvement of a deep neural network-based ASR system by leveraging realistic testing data obtained from medical simulations of battlefield scenarios, which to our knowledge have not been addressed in any prior work. The system performance is shown to improve significantly using multi-style training, language model adaptation for the medical domain, speech enhancement, and NLP techniques.

Speech enhancement aims to improve the intelligibility and quality of speech that is affected by noise. In this paper, we propose a novel speaker-aware speech enhancement (SASE) method that extracts speaker information using long short-term memory (LSTM) layers, and then uses a convolutional recurrent neural network (CRN) to embed the extracted speaker information. It is shown in a series of comprehensive experiments that only a few seconds of reference audio suffice for the proposed SASE method to perform better than LSTM and CRN baseline systems. The addition of a self-attention mechanism can further boost relevant speech-quality metrics.

Multi-stage learning is an effective technique for invoking multiple deep-learning modules sequentially. This paper applies multi-stage learning to speech enhancement by using a multi-stage structure, where each stage comprises a self-attention (SA) block followed by stacks of temporal convolutional network (TCN) blocks with doubling dilation factors. Each stage generates a prediction that is refined in a subsequent stage. A feature fusion block is inserted at the input of later stages to re-inject original information. The resulting multi-stage speech enhancement system, multi-stage SA-TCN, is compared with state-of-the-art deep-learning speech enhancement methods using the LibriSpeech and VCTK datasets. The multi-stage SA-TCN system's hyperparameters are fine-tuned, and the impact of the SA block, the feature fusion block, and the number of stages are determined. The use of a multi-stage SA-TCN system as a front-end for automatic speech recognition systems is also investigated. It is shown that the multi-stage SA-TCN systems perform well relative to other state-of-the-art systems in terms of speech enhancement and speech recognition scores.

The next generation of cellular networks needs paradigm shifts to realize the exponentially increasing demands for high throughputs, low latencies, and reliable communication. These networks will be required to address the requirements of a number of applications and services as well as industry verticals such as intelligent transportation, industrial Internet of Things (IoT), e-Health, augmented reality/virtual reality (AR/VR), and other future networks-based services. Millimeter wave (mmWave) communication is a promising technology for 5G systems due to its potential for multi-gigabit throughput. However, mmWave links suffer from high path loss and blockage. We propose an approach using New Radio Dual Connectivity (NR-DC) to maximize the throughput while ensuring ultra-reliable low latency communication. The performance evaluation conducted using Simu5G, which is an OMNeT++ based network simulator, demonstrates that this multi-connectivity technique can improve the performance of next generation 5G networks. Simulation results show that NR-DC improves the performance significantly in terms of throughput, latency, and reliability by up to 14%, 33%, and 11%, respectively, compared to a single connectivity system.

Background Prehospitalization documentation is a challenging task and prone to loss of information, as paramedics operate under disruptive environments requiring their constant attention to the patients. Objective The aim of this study is to develop a mobile platform for hands-free prehospitalization documentation to assist first responders in operational medical environments by aggregating all existing solutions for noise resiliency and domain adaptation. Methods The platform was built to extract meaningful medical information from the real-time audio streaming at the point of injury and transmit complete documentation to a field hospital prior to patient arrival. To this end, the state-of-the-art automatic speech recognition (ASR) solutions with the following modular improvements were thoroughly explored: noise-resilient ASR, multi-style training, customized lexicon, and speech enhancement. The development of the platform was strictly guided by qualitative research and simulation-based evaluation to address the relevant challenges through progressive improvements at every process step of the end-to-end solution. The primary performance metrics included medical word error rate (WER) in machine-transcribed text output and an F1 score calculated by comparing the autogenerated documentation to manual documentation by physicians. Results The total number of 15,139 individual words necessary for completing the documentation were identified from all conversations that occurred during the physician-supervised simulation drills. The baseline model presented a suboptimal performance with a WER of 69.85% and an F1 score of 0.611. The noise-resilient ASR, multi-style training, and customized lexicon improved the overall performance; the finalized platform achieved a medical WER of 33.3% and an F1 score of 0.81 when compared to manual documentation. The speech enhancement degraded performance with medical WER increased from 33.3% to 46.33% and the corresponding F1 score decreased from 0.81 to 0.78. All changes in performance were statistically significant (P<.001). Conclusions This study presented a fully functional mobile platform for hands-free prehospitalization documentation in operational medical environments and lessons learned from its implementation.

Multi-stage learning is an effective technique to invoke multiple deep-learning modules sequentially. This paper applies multi-stage learning to speech enhancement by using a multi-stage structure, where each stage comprises a self-attention (SA) block followed by stacks of temporal convolutional network (TCN) blocks with doubling dilation factors. Each stage generates a prediction that is refined in a subsequent stage. A fusion block is inserted at the input of later stages to re-inject original information. The resulting multi-stage speech enhancement system, in short, multi-stage SA-TCN, is compared with state-of-the-art deep-learning speech enhancement methods using the LibriSpeech and VCTK data sets. The multi-stage SA-TCN system's hyper-parameters are fine-tuned, and the impact of the SA block, the fusion block and the number of stages are determined. The use of a multi-stage SA-TCN system as a front-end for automatic speech recognition systems is investigated as well. It is shown that the multi-stage SA-TCN systems perform well relative to other state-of-the-art systems in terms of speech enhancement and speech recognition scores.

Next Generation 5G networks are of prime importance to meet the increasing demands of emerging IoT applications and industry verticals for high throughputs and ultra-reliable low latency communication. Future IoT services also require high scalability and Internet connectivity for a wide range of applications, including various mobility scenarios. Communication systems developed so far have not been able to fully address the requirements of IoT applications. However, 5G has the capability to satisfy these needs and provides key enabling technologies for ubiquitous deployment of the IoT technology. We propose and evaluate a novel 5G-IoT architecture using Simu5G network simulator for enabling future IoT systems to support next generation applications. The proposed 5G-IoT architecture is shown to achieve high throughputs of around 1 Gbps with sub-millisecond latency and ultra-high reliability for scalable next generation smart systems.

Introduction The purpose of this study was to characterize the at-risk diabetes and prediabetes patient population visiting emergency department (ED) and urgent care (UC) centers in upstate South Carolina. Methods We conducted this retrospective study at the largest non-profit healthcare system in South Carolina, using electronic health record (EHR) data of patients who had an ED or UC visit between February 2, 2016–July 31, 2018. Key variables including International Classification of Diseases, 10th Revision codes, laboratory test results, family history, medication, and demographic characteristics were used to classify the patients as healthy, having prediabetes, having diabetes, being at-risk for prediabetes, or being at-risk for diabetes. Patients who were known to have diabetes were classified further as having controlled diabetes, management challenged, or uncontrolled diabetes. Population analysis was stratified by the patient’s annual number of ED/UC visits. Results The risk stratification revealed 4.58% unique patients with unrecognized diabetes and 10.34% of the known patients with diabetes considered to be suboptimally controlled. Patients identified as diabetes management challenged had more ED/UC visits. Of note, 33.95% of the patients had unrecognized prediabetes/diabetes risk factors identified during their ED/UC with 87.95% having some form of healthcare insurance. Conclusion This study supports the idea that a single ED/UC unscheduled visit can identify individuals with unrecognized diabetes and an at-risk prediabetes population using EHR data. A patient’s ED/UC visit, regardless of their primary reason for seeking care, may be an opportunity to provide early identification and diabetes disease management enrollment to augment the medical care of our community.

Current Internet Protocol routing provides minimal privacy, which enables multiple exploits. The main issue is that the source and destination addresses of all packets appear in plain text. This enables numerous attacks, including surveillance, man-in-the-middle (MITM), and denial of service (DoS). The talk explains how these attacks work in the current network. Endpoints often believe that use of Network Address Translation (NAT), and Dynamic Host Configuration Protocol (DHCP) can minimize the loss of privacy.We will explain how the regularity of human behavior can be used to overcome these countermeasures. Once packets leave the local autonomous system (AS), they are routed through the network by the Border Gateway Protocol (BGP). The talk will discuss the unreliability of BGP and current attacks on the routing protocol. This will include an introduction to BGP injects and the PEERING testbed for BGP experimentation. One experiment we have performed uses statistical methods (CUSUM and F-test) to detect BGP injection events. We describe work we performed that applies BGP injects to Internet Protocol (IP) address randomization to replace fixed IP addresses in headers with randomized addresses. We explain the similarities and differences of this approach with virtual private networks (VPNs). Analysis of this work shows that BGP reliance on autonomous system (AS) numbers removes privacy from the concept, even though it would disable the current generation of MITM and DoS attacks. We end by presenting a compromise approach that creates software-defined data exchanges (SDX), which mix traffic randomization with VPN concepts. We contrast this approach with the Tor overlay network and provide some performance data.

Speech enhancement is an essential component in robust automatic speech recognition (ASR) systems. Most speech enhancement methods are nowadays based on neural networks that use feature-mapping or mask-learning. This paper proposes a novel speech enhancement method that integrates timedomain feature mapping and mask learning into a unified framework using a Generative Adversarial Network (GAN). The proposed framework processes the received waveform and decouples speech and noise signals, which are fed into two shorttime Fourier transform (STFT) convolution 1-D layers that map the waveforms to spectrograms in the complex domain. These speech and noise spectrograms are then used to compute the speech mask loss. The proposed method is evaluated using the TIMIT data set for seen and unseen signal-to-noise ratio conditions. It is shown that the proposed method outperforms the speech enhancement methods that use Deep Neural Network (DNN) based speech enhancement or a Speech Enhancement Generative Adversarial Network (SEGAN).