Autonomous Unmanned Aerial Vehicle (UAV) swarms are increasingly used as rapidly deployable aerial relays and sensing platforms, yet practical deployments must operate under partial observability and intermittent peer-to-peer connectivity. We present a graph-based multi-agent reinforcement learning framework trained under centralized training with decentralized execution (CTDE): a centralized critic and global state are available only during training, while each UAV executes a shared policy using local observations and messages from nearby neighbors. Under restricted communication, neighbor relations are induced by an SNR-threshold connectivity graph. Our architecture encodes local agent state and nearby entities with an agent-entity attention module and aggregates inter-UAV messages with neighbor self-attention over a signal-quality-limited communication graph defined by a channel model. We evaluate the framework on a cooperative relay-deployment task, DroneConnect. Experimental results show that the proposed method achieves an approximately 12

Smart homes with multiple smart devices often face group decision making situations where multiple users of these devices need to converge on specific actions. For example, a smart thermostat assumes that the occupants of that smart home agree on the temperature settings at various times. While there have been several detailed studies on access control of shared IoT devices in smart homes, there is a paucity of studies that specifically consider group decisions on IoT device settings. Access control involves the decision of a few (e.g., the spouses), and IoT device settings (e.g., thermostat temperature setting) involve input from all occupants of that smart home at any point in time. What complicates matter further is the fact that traditional expected utility theory assumes rational decision-making, which renders it unsuitable to model bias and irrationalities in human decision-making processes. We therefore study the dynamics of IoT-related group decisions in multi-user smart homes through the lens of Cumulative Prospect Theory (CPT). Specifically, we study single point vs. range choice IoT device setting preferences of smart home device users and show with an example that single and range choice could possibly result in entirely different decision outcomes.

Single channel wireless systems are increasingly adopted in ultra-low-power IoT, LPWAN, and low-latency applications due to their hardware simplicity, energy efficiency and application requirements. However, fixed-frequency operation in these communication systems makes them extremely vulnerable to persistent interference, as traditional defenses like frequency hopping become infeasible. This paper presents a single channel reliable communication scheme resilient to both benign and malicious persistent interferences. Our design addresses two core challenges: (i) separating legitimate signals from the interfering ones without a clean reference, and (ii) defending against time-varying interference patterns. To this end, we propose extended-duration bit transmissions that enable amplitude-aware decoding, combined with a predictive interference subtraction technique based on short-term signal stability. In addition, we introduce cryptographically coordinated, pseudo-random bit-timing variations that help with decoding by misaligning interfering signal transition with bit boundaries. We evaluate our interference resilient system through extensive experiments on a prototype testbed implemented with software-defined radio, mostly demonstrating 95-100% accuracy in data recovery whereas interference-agnostic baselines fail completely. While optimized for single channel systems, our approach remains complementary to multi channel defenses in broader wireless settings.

Millimeter-wave (mmWave) radars have demonstrated strong performance under challenging visibility conditions and support high-resolution imaging through techniques such as synthetic aperture radar (SAR). The increasing availability of compact, low-cost mmWave systems has made them particularly appealing for low-power platforms in mission-critical tactical scenarios. However, traditional SAR imaging demands extensive sampling, leading to significant delays and excessive data collection—particularly in environments where only a small subset of the scene contains mission-relevant information.To overcome this inefficiency, we introduce mmPrism, a perception-aware attention scheduling framework for mmWave radar that intelligently concentrates sensing and computation on Target Areas of Interest (TAIs). By combining rapid coarse scanning with spotlight-mode SAR under the control of a perception-aware TAI scheduler, mmPrism reduces unnecessary overhead while preserving task-relevant detail. Experimental results demonstrate substantial improvements in recognition throughput and latency, with minimal compromise in imaging resolution.

In monitoring urban areas with dense infrastructures, drone swarms emerge as an efficient means for locating and collecting data from urban targets. In this task, the drone swarms are required to visit a set of valuable sites. However, many challenges exist in the data collection process of drone swarms: complex streets and obstacles require drones to use computer vision and pathfinding algorithms to perceive the environment, avoid collisions, and manage power in real time with their limited onboard computation and battery. In extreme situations where battery limitations prevent the surveillance of all intended sites, drones are desired to collect information from as many valuable targets as possible. It calls for extra smart path planning. Meanwhile, dynamically changing zone priorities and environments can suddenly change target values or render planned sites dangerous, necessitating immediate path recomputation to ensure efficiencies of urban target surveillance. In this paper, we address the following question: Can we enhance the performance, including safety and efficiency, of drone swarms to conduct urban target monitoring, despite constraints like battery life, limited computational resources, and evolving environments? We introduce our solution, D-planner, a system that performs efficient drone swarm path planning with intelligent navigation algorithms, carries out collision-free navigation using a computer vision and pathfinding module, and integrates incremental dynamic path computation that leverages geometries and greedy strategies to figure out the priority of data sites and plan safe paths with a limited deviation from the original paths. To evaluate the system, we constructed a simulation environment of a large city with meter-level precision using Google Maps. Experiments on our system show that D-planner can improve the planning speed of target-rich paths by up to 6.5x and the total value by up to 25.60% compared with baseline solutions.

During military operations, Unmanned Aerial Vehicles (UAVs) can act as aerial base stations to provide temporary communication services for ground users. Modern warfare often takes place in urban environments, where obstacles like trees and buildings can impede signal propagation, leading to reduced communication quality. This paper introduces an innovative approach using UAVs’ observations of surrounding obstacles to make informed decisions about their movements for improved user coverage. We use Deep Reinforcement Learning (DRL) in a multi-agent setting to integrate real-time obstacle observations and cooperative strategies among UAVs for enhanced Line-of-Sight (LoS) connections essential for effective communication, establishing reliable communication networks in urban environments. The DRL-based decision-making framework enables UAVs to dynamically adjust their positions to achieve near-optimal user coverage, representing an advancement in urban warfare communication technology. Simulation results show that UAVs with the DRL-based decision-making framework achieve significantly improved coverage and shorter travel distances for enhanced operational efficiency in dense urban scenarios compared with baselines without considering obstacle observations.

An increasing number of students are becoming interested in learning about the Internet of Things (IoT) space. However, today, we lack scalable and efficient ways to bring hands-on IoT learning to many due to hardware accessibility, system complexity, and deployment environment constraints. This paper presents ThingVisor, an IoT learning platform that enables hands-on IoT development in an immersive virtual space. Specifically, it allows users to design, test, and deploy IoT devices virtually in a simulated IoT world with static and dynamic software verification as a complementary tool to IoT education. ThingVisor consists of (1) a Device Design Stack to configure virtual IoT devices, (2) an Immersive Runtime Stack to interact with devices and environment, and (3) a Device Emulator, which is a runtime environment used to execute virtual devices to get their behaviors. Our experiments confirm the learning effectiveness and user satisfaction of our platform. Additionally, we have demonstrated the scalability and usability of the system through load testing and application of the System Usability Scale. Our results indicate that students can achieve up to a 32% improvement in their scores after engaging with ThingVisor for two weeks, irrespective of their prior experience.

Data plane verification has grown into a powerful tool to ensure network correctness. However, existing methods with monolithic models have memory requirements tied to network sizes, and the existing method of scaling out is too limited in expressiveness to capture practical network features. In this paper, we describe Scylla, a general data plane verifier that provides fine-grained scale-out without the need for a monolithic network model. Scylla creates models for what we call intent-based slices, each of which is constructed at the rule-level granularity with only enough to verify a given set of intents. The sliced models are retained and incrementally updated in memory across a distributed compute cluster in response to network updates. Our experiments show that Scylla makes the scaling problem more granular – tied to the size of the intent-based slices rather than that of the overall network. This enables Scylla to verify large, complex networks in minimum units of work that are significantly smaller (in both memory and time) than past techniques, enabling fast scale-out verification with minimal resource requirement.

Digital Twins are evolving as a key component in modern systems with diverse applications like remote prognostics, optimizing run-time operation, anomaly detection, and more. The essential elements of a digital twin are a virtual representation, a physical asset, and the transfer of data/information between the two. IoT deployments are generally characterized by resource constraints, making synchronization of digital twins with IoT devices more challenging. There is a pressing need to optimize the bandwidth of the data transferred between the system and the twin, while ensuring that the twin is able to capture selected key aspects of the current operational state accurately. In this paper, we present TwinSync, a framework that can be utilized to construct flexible real-time representations of deployed IoT systems and efficiently synchronize relevant system states with the twin, over a communication bottleneck, within a configurable application-specific notion of error (henceforth referred to as approximate synchronization). Our approach is optimized to achieve data transfers utilizing less bandwidth without compromising the ability of the twin to replicate real-time system states within the specified approximate synchronization semantics. We evaluate the efficacy of TwinSync's synchronization by conducting both a synthetic analysis and a case study based on a real-life application prototype. Our evaluation indicates that using TwinSync can provide the same or greater accuracy (in many cases) while sending significantly fewer bytes than a bandwidth-insensitive synchronization approach. The result is attributed to a more judicial selection of data to transmit over bottlenecks, compared to bandwidth-insensitive approaches.

Modern battlefields are subject to spoofing of GPS signals. While large aircraft platforms can counter the effects of GPS spoofing via redundant and dissimilar sensors, the disadvantaged nodes with smaller platforms such as Air Launched Effects (ALE), with more limited capabilities, can be vulnerable. That said, other sensors on the platform may give clues to the drone about where it is located. In this paper, we investigate the ability of sensor fusion to remediate spoofing of GPS signals for ALE platforms. We first conduct performance comparison among several complementary techniques, including the use of inertial measurement units (IMUs), communication with nearby ALEs (to compare GPS readings), and received signal strength from networking connections (to estimate distance to neighboring ALEs, etc.) We then propose a novel architecture that performs sensor fusion to intelligently combine observations across multiple sensors so as to maximize the ability to detect GPS spoofing as well as to reconstruct coordinates with confidence levels. From a simulation study based on real-world mobility and sensor traces, we find that our approach can improve location estimates accuracy by multiple orders of magnitude as compared to simple baseline techniques, supplementing the ability for ALEs to navigate and execute missions in GPS-denied environments.

Extant research on Battery Electric Vehicles (BEVs) stress the significance of appropriate charging facilities that are available within reach when needed. For a BEV rental agency that provides free-floating rental vehicles, keeping the vehicles in charged state relies on the availability of convenient charging facilities across a wide area (e.g., city) and rental customers to properly connect vehicle to charger upon returning the vehicle. While a BEV that has zero charge is as good as its absence from the rental customer's perspective, one that does not have enough charge for the short trip envisioned by the renter may suffer a similar fate. We study the dynamics associated with charging BEVs in a free-floating rental platform. Specifically, we model and determine the optimal charged state of BEV rental vehicles in a free-floating platform.

Digital twins have been designed and implemented for diverse applications like smart manufacturing, healthcare, supply chain and retail management. They provide monitoring, remote prognostics and health management capabilities for the various physical assets used in these domains. Many of these capabilities would be beneficial to microservice architectures as well, given the need for lightweight monitoring solutions in multitenant environments. In particular, twins can provide operators with real-time resource usage metrics which help with operational objectives such as resource planning, anomaly detection, rewind and replay and so on. In this work, we propose a design for building digital twins for microservice architectures. As a proof of concept, we focus on modelling the resource utilization as that is a key requirement for monitoring system reliability and security. In general, digital twins require a real world counterpart, a virtual model and a mechanism for consistently keeping both synchronized. We focus on the two latter aspects of the digital twin. Our approach involves converting a formal model of a microservice architecture into a digital twin that can capture and execute an actual cluster's state. We present an extensible architecture connecting the various components of the system and the twin and evaluate the twin's ability to capture the real-time state of a real Kubernetes cluster. We also discuss future extensions which can enhance the system's security by detecting a broad range of attacks.

Ad hoc computing needs at remote locations or locations with insufficient resources due to reasons such as natural disasters require flexible solutions that are readily deployed on demand. We consider one such scenario where unmanned aerial vehicles (UAVs) or drones can be used to provide necessary coverage in terms of computational support. Specifically, we consider an environment where ground-based computational demand is satisfied by aerial drones that share the computational load to provide seamless service. We study transfer and location policies where transfer policy determines whether a job is locally processed by the drone that receives the order and location policy determines where a job is processed if it is sent to another drone. Our results indicate that the mean queue length of jobs waiting to be processed at the drones decreases with sharing the job processing load among the drones in the modeled system. Our results also highlight the beneficial aspects of the two step transfer and location policies.

Security and privacy of vehicles, occupants of such vehicles, roadside infrastructure, and other entities that are part of Vehicular Ad hoc NETwork (VANET) cannot be overstated. Cryptography is commonly used to authenticate and to secure communication among VANET entities. As vehicles are mobile, it is essential for authentication protocols to be lightweight, quick, and with minimal number of passed messages. It is also necessary to ensure that these protocols are secure against attacks. However, extant authentication protocols are not necessarily lightweight and almost all of them are vulnerable to relay attacks. We propose secure and truly lightweight authentication protocols for the VANET environment.

Unmanned aerial vehicles (UAVs) or drones have the potential to supplant helicopters in real-time highway surveillance applications due to cost, form factor, and other considerations. The wireless medium used for communication has the potential to expose a fleet of drones en route to surveillance to link failures and attacks on passed messages. Given the sparse topology, messages among UAVs, vehicles, and trusted authority could be transmitted through collaboration among UAVs and vehicles. Since vehicles and drones in a highway environment are mobile, related ad hoc network is continuously updated to account for reachability of transmitted signals. It is also necessary to authenticate the drones and vehicles to ensure that the transmitted messages are uncorrupted and trusted. To accommodate processing power and mobility constraints, we develop lightweight authentication protocols that facilitate secure message transfer. We also evaluate the security properties of these protocols.

Wireless communication among vehicular ad hoc network (VANET) entities is secured through cryptography, which is used for authentication as well as to ensure the overall security of messages in this environment. Authentication protocols play a significant role and are therefore required to be free of vulnerabilities that allow entity impersonation, unauthorized entry, and general misuse of the system. A resourceful adversary can inflict serious damage to VANET systems through such vulnerabilities. We consider several VANET authentication protocols in the literature and identify vulnerabilities. In addition to the commonly considered vulnerabilities in VANETs, we observe that the often-overlooked relay attack is possible in almost all VANET authentication protocols. Relay attacks have the potential to cause damage in VANETs through misrepresentation of vehicle identity, telematic data, traffic-related warnings, and information related to overall safety in such networks. We discuss possible countermeasures to address identified vulnerabilities. We then develop an authentication protocol that uses ambient conditions to secure against relay attacks and other considered vulnerabilities. We include security proof for the proposed protocol.

As Battery Electric Vehicles (BEV) become more popular because of mandates or through organic growth, bottlenecks include the availability of charging facilities and the sheer time it takes to fully charge these vehicles. BEV rental agencies face the additional burden to decide on the number of vehicles to fully charge and ready to go for the next day. We develop a stochastic model to help facilitate this process. Specifically, we develop expressions for the number of additional vehicles to charge based on Uniform demand, accessible charging facilities, and the available number of partially charged vehicles.

Recent concepts in defense herald an increasing degree of automation of future military systems, with an emphasis on accelerating sensing-to-decision loops at the tactical edge, reducing their network communication footprint, and improving the inference quality of intelligent components in the loop. These requirements pose resource management challenges, calling for operating-system-like constructs that optimize the use of limited computational resources at the tactical edge. This paper describes these challenges and presents IoBT-OS, an operating system for the Internet of Battlefield Things that aims to optimize decision latency, improve decision accuracy, and reduce corresponding resource demands on computational and network components. A simple case-study with initial evaluation results is shown from a target tracking application scenario.

In the aftermath of a disaster, the ability to reliably communicate and coordinate emergency response could make a meaningful difference in the number of lives saved or lost. However, post-disaster areas tend to have limited functioning communication network infrastructure while emergency response teams are carrying increasingly more devices, such as sensors and video transmitting equipment, which can be low-powered with limited transmission ranges. In such scenarios, unmanned aerial vehicles (UAVs) can be used as relays to connect these devices with each other. Since first responders are likely to be constantly mobile, the problem of where these UAVs are placed and how they move in response to the changing environment could have a large effect on the number of connections this UAV relay network is able to maintain. In this work, we propose DroneDR, a reinforcement learning framework for UAV positioning that uses information about connectivity requirements and user node positions to decide how to move each UAV in the network while maintaining connectivity between UAVs. The proposed approach is shown to outperform other greedy heuristics across a broad range of scenarios and demonstrates the potential in using reinforcement learning techniques to aid communication during disaster relief operations.