Automated Uncrewed Aerial Vehicle (UAV) landing is crucial for autonomous UAV services such as monitoring, surveying, and package delivery. It involves detecting landing targets, perceiving obstacles, planning collision-free paths, and controlling UAV movements for safe landing. Failures can lead to significant losses, necessitating rigorous simulation-based testing for safety. Traditional offline testing methods, limited to static environments and predefined trajectories, may miss violation cases caused by dynamic objects like people and animals. Conversely, online testing methods require extensive training time, which is impractical with limited budgets. To address these issues, we introduce GARL, a framework combining a genetic algorithm (GA) and reinforcement learning (RL) for efficient generation of diverse and real landing system failures within a practical budget. GARL employs GA for exploring various environment setups offline, reducing the complexity of RL's online testing in simulating challenging landing scenarios. Our approach outperforms existing methods by up to 18.35% in violation rate and 58% in diversity metric. We validate most discovered violation types with real-world UAV tests, pioneering the integration of offline and online testing strategies for autonomous systems. This method opens new research directions for online testing, with our code and supplementary material available at https://github.com/lfeng0722/drone_testig/.

This paper introduces LiSWARM, a low-cost LiDAR system to detect and track individual drones in a large swarm. LiSWARM provides robust and precise localization and recognition of drones in 3D space, which is not possible with state-of-the-art drone tracking systems that rely on radio-frequency (RF), acoustic, or RGB image signatures. It includes (1) an efficient data processing pipeline to process the point clouds, (2) robust priority-aware clustering algorithms to isolate swarm data from the background, (3) a reliable neural network-based algorithm to recognize the drones, and (4) a technique to track the trajectory of every drone in the swarm. We develop the LiSWARM prototype and validate it through both in-lab and field experiments. Notably, we measure its performance during two drone light shows involving 150 and 500 drones and confirm that the system achieves up to 98% accuracy in recognizing drones and reliably tracking drone trajectories. To evaluate the scalability of LiSWARM, we conduct a thorough analysis to benchmark the system's performance with a swarm consisting of 15,000 drones. The results demonstrate the potential to leverage LiSWARM for other applications, such as battlefield operations, errant drone detection, and securing sensitive areas such as airports and prisons.

This paper introduces an autonomous UAV vision system for continuous, real-time tracking of marine animals, specifically sharks, in dynamic marine environments. The system integrates an onboard computer with a stabilised RGB-D camera and a custom-trained OSTrack pipeline, enabling visual identification under challenging lighting, occlusion, and sea-state conditions. A key innovation is the inter-UAV handoff protocol, which enables seamless transfer of tracking responsibilities between drones, extending operational coverage beyond single-drone battery limitations. Performance is evaluated on a curated shark dataset of 5,200 frames, achieving a tracking success rate of 81.9% during real-time flight control at 100 Hz, and robustness to occlusion, illumination variation, and background clutter. We present a seamless UAV handoff framework, where target transfer is attempted via high-confidence feature matching, achieving 82.9% target coverage. These results confirm the viability of coordinated UAV operations for extended marine tracking and lay the groundwork for scalable, autonomous monitoring.

This paper proposes an Emergency Battery Service (EBS) for drones in which an EBS drone flies to a drone in the field with a depleted battery and transfers a fresh battery to the exhausted drone. The authors present a unique battery transfer mechanism and drone localization that uses the Cross Marker Position (CMP) method. The main challenges include a stable and balanced transfer that precisely localizes the receiver drone. The proposed EBS drone mitigates the effects of downwash due to the vertical proximity between the drones by implementing diagonal alignment with the receiver, reducing the distance to 0.5 m between the two drones. CFD analysis shows that diagonal instead of perpendicular alignment minimizes turbulence, and the authors verify the actual system for change in output airflow and thrust measurements. The CMP marker-based localization method enables position lock for the EBS drone with up to 0.9 cm accuracy. The performance of the transfer mechanism is validated experimentally by successful mid-air transfer in 5 seconds, where the EBS drone is within 0.5 m vertical distance from the receiver drone, wherein 4m/s turbulence does not affect the transfer process.

Cooperative UAV transport operations in GPS-denied environments pose significant challenges in localization, coordination, and payload stability. This paper introduces a vision-based Leader-Follower drone system using MAVROS and depth cameras for real-time pose estimation and control. The leader transmits pose and velocity updates to the follower, ensuring synchronized movements. The system maintained a 50 Hz update rate, achieving 28 FPS, 12 ms latency, and 1.2 cm position error on a straight path. The 3-DEE system effectively managed payload-induced attitude variations with low vibration levels and improved speed accuracy. These results confirm the system's robustness for precise localization and stable cooperative UAV transport.

The capabilities of drones are increasing every day, as is the ease with which civilians can buy and fly them. Most drones are equipped with a camera that is used by a point-of-view operator and, at the same time, can be used for image capture. The use of drones creates a threat to privacy whereby anyone who can fly a drone can take pictures without permission. This study aims to create a 2-axis tracker system that can recognize a drone and locate the position of the drone camera so that a laser beam can track and dazzle the drone camera. The depth-sensing camera is used to localize the part of the target corresponding to the drone’s camera and is created using the YOLOv5 algorithm as a deep-learning detector model. The drone’s camera range and position are challenging to detect due to its small size. Our adaptive detection method combines drone detection and drone camera detection. The depth-sensing camera provides input in the form of a three-coordinate axis from the target. If only the drone is detected, a predictive algorithm can determine the camera’s position for illumination with the laser. Alternatively, if the drone camera is detected, the laser can follow the target’s movement more quickly. In this study, a green (520 nm) laser module with adjustable power is used to investigate factors that affect the dazzling range. The computer vision detection algorithm can detect and localize the position of the drone camera up to 500 cm with a confidence level of more than 65%. If the target is in the center of the field of view, the accuracy of the target position can reach 98%. The tracker can follow the drone’s movement from 2 m/s to 4 m/s with a maximum error of 1.9 cm from the center point of the drone camera for close range. For long range, the maximum error is 6.2 cm. A laser power of 23.5 mW at 500 cm distance is found to be sufficient to dazzle and track drone cameras.

This paper pushes the limits of automated resource allocation in container environments. Recent works set container CPU and memory limits by automatically scaling containers based on past resource usage. However, these systems are heavy- weight and run on coarse-grained time scales, resulting in poor performance when predictions are incorrect. We propose Escra, a container orchestrator that enables fine-grained, event- based resource allocation for a single container and distributed resource allocation to manage a collection of containers. Escra performs resource allocation on sub-second intervals within and across hosts, allowing operators to cost-effectively scale resources without performance penalty. We evaluate Escra on two types of containerized applications: microservices and serverless functions. In microservice environments, fine-grained and event- based resource allocation can reduce application latency by up to 96.9% and increase throughput by up to 3.2x when compared against the current state-of-the-art. Escra can increase performance while simultaneously reducing 50th and 99th%ile CPU waste by over 10x and 3.2x, respectively. In serverless environments, Escra can reduce CPU reservations by over 2.1x and memory reservations by more than 2x while maintaining similar end-to-end performance.

Social media platforms have been exploited to disseminate misinformation in recent years. The widespread online misinformation has been shown to affect users' beliefs and is connected to social impact such as polarization. In this work, we focus on misinformation's impact on specific user behavior and aim to understand whether general Twitter users changed their behavior after being exposed to misinformation. We compare the before and after behavior of exposed users to determine whether the frequency of the tweets they posted, or the sentiment of their tweets underwent any significant change. Our results indicate that users overall exhibited statistically significant changes in behavior across some of these metrics. Through language distance analysis, we show that exposed users were already different from baseline users before the exposure. We also study the characteristics of two specific user groups, multi-exposure and extreme change groups, which were potentially highly impacted. Finally, we study if the changes in the behavior of the users after exposure to misinformation tweets vary based on the number of their followers or the number of followers of the tweet authors, and find that their behavioral changes are all similar.

Social media platforms have been exploited to conduct election interference in recent years. In particular, the Russian-backed Internet Research Agency (IRA) has been identified as a key source of misinformation spread on Twitter prior to the 2016 U.S. presidential election. The goal of this research is to understand whether general Twitter users changed their behavior in the year following first contact from an IRA account. We compare the before and after behavior of contacted users to determine whether there were differences in their mean tweet count, the sentiment of their tweets, and the frequency and sentiment of tweets mentioning @realDonaldTrump or @HillaryClinton. Our results indicate that users overall exhibited statistically significant changes in behavior across most of these metrics, and that those users that engaged with the IRA generally showed greater changes in behavior.

Internet users are routinely exposed to fake news in their social media feeds. The main goal of this paper is to identify the factors readers consider important in discriminating against fake news from true news when reading an online news article. We design and conduct three surveys using Amazon Mechanical Turk to identify the top factors and rate them under diverse scenarios. Our results suggest that people perceive news Source and Content to be the most important factors, in general, to distinguish fake news from true news, however, their importance reduces in practice when people actually read a news article. Furthermore, the importance of different factors in the credibility determination of a news article varies with people's political leanings. Our work is the first of its kind and offers new insights into how people determine the legitimacy of online news articles.

This paper presents a new approach to memory disaggregation called FluidMem that leverages the userfault mechanism in Linux to achieve full memory disaggregation in software. FluidMem enables dynamic and transparent resizing of an unmodified Virtual Machine’s (VM’s) memory footprint in the cloud. As a result, a VM’s memory footprint can seamlessly scale over multiple machines or even be downsized to a near-zero footprint on a given server. FluidMem’s architecture provides flexibility to cloud operators to manage remote memory without requiring guest intervention, while also supporting paging out the entirety of a VM’s pages within its address space. FluidMem integrates with a remote memory backend in a modular way, easily supporting systems such as RAMCloud to harness remote memory. We demonstrate FluidMem outperforms an existing memory disaggregation approach based on network swap. Microbenchmarks are evaluated to characterize the latency of different components of the FluidMem architecture, and two memory-intensive applications are demonstrated using FluidMem, the Graph500 benchmark, and MongoDB. Additionally, we show FluidMem can flexibly and efficiently grow and shrink the memory footprint of a VM as defined by a cloud provider.

Efficient resource management in cloud computing research is a crucial problem because resource over-provisioning increases costs for cloud providers and cloud customers; resource under-provisioning increases the application latency, and it may violate service level agreements, which eventually makes cloud providers lose their customers and income. As a result, researchers have been striving to develop optimal resource management in cloud computing environments in different ways, such as container placement, job scheduling and multi-resource scheduling. Machine learning techniques are extensively used in this area. In this paper, we present a comprehensive survey on the projects that leveraged machine learning techniques for resource management solutions in the cloud computing environment. At the end, we provide a comparison between these projects. Furthermore, we propose some future directions that will guide researchers to advance this field.

Drones have carried weapons, drugs, explosives and illegal packages in the recent past, raising strong concerns from public authorities. While existing drone monitoring systems only focus on detecting drone presence, localizing or fingerprinting the drone, there is a lack of a solution for estimating the additional load carried by a drone. In this paper, we present a novel passive RF system, namely DroneScale, to monitor the wireless signals transmitted by commercial drones and then confirm their models and loads. Our key technical contribution is a proposed technique to passively capture vibration at high resolution (i.e., 1Hz vibration) from afar, which was not possible before. We prototype DroneScale using COTS RF components and illustrate that it can monitor the body vibration of a drone at the targeted resolution. In addition, we develop learning algorithms to extract the physical vibration of the drone from the transmitted signal to infer the model of a drone and the load carried by it. We evaluate the DroneScale system using 5 different drone models, which carry external loads of up to 400g. The experimental results show that the system is able to estimate the external load of a drone with an average accuracy of 96.27%. We also analyze the sensitivity of the system with different load placements with respect to the drone's body, flight modes, and distances up to 200 meters.

In serverless computing, developers define a function to handle an event, and the serverless framework horizontally scales the application as needed. The downside of this function-based abstraction is it limits the type of application supported and places a bound on the function to be within the physical resource limitations of the server the function executes on. In this paper we propose a new abstraction for serverless computing: a developer supplies a process and the serverless framework seamlessly scales out the process's resource usage across the datacenter. This abstraction enables processing to not only be more general purpose, but also allows a process to break out of the limitations of a single server – making serverless computing more serverless. To realize this abstraction, we propose ServerlessOS, comprised of three key components: (i) a new disaggregation model, which leverages disaggregation for abstraction, but enables resources to move fluidly between servers for performance; (ii) a cloud orchestration layer which manages fine-grained resource allocation and placement throughout the application's lifetime via local and global decision making; and (iii) an isolation capability that enforces data and resource isolation across disaggregation, effectively extending Linux cgroup functionality to span servers.

User emotions are important contextual features in building context-aware pervasive applications. In this paper, we explore the question of whether it is possible to predict user emotions from their smartphone activities. To get the ground truth data, we have built an Android app that collects user emotions along with a number of features including their current location, activity they are engaged in, and smartphones apps they are currently running. We deployed this app for over a period of three months and collected a large amount of useful user data. We describe the details of this data in terms of statistics and user behaviors, provide a detailed analysis in terms of correlations between user emotions and other features, and finally build classifiers to predict user emotions. Performance of these classifiers is quite promising with high accuracy. We describe the details of these classifiers along with the results.

Drones are increasingly flying in sensitive airspace where their presence may cause harm, such as near airports, forest fires, large crowded events, secure buildings, and even jails. This problem is likely to expand given the rapid proliferation of drones for commerce, monitoring, recreation, and other applications. A cost-effective detection system is needed to warn of the presence of drones in such cases. In this paper, we explore the feasibility of inexpensive RF-based detection of the presence of drones. We examine whether physical characteristics of the drone, such as body vibration and body shifting, can be detected in the wireless signal transmitted by drones during communication. We consider whether the received drone signals are uniquely differentiated from other mobile wireless phenomena such as cars equipped with Wi- Fi or humans carrying a mobile phone. The sensitivity of detection at distances of hundreds of meters as well as the accuracy of the overall detection system are evaluated using software defined radio (SDR) implementation.

We describe opportunities and challenges with wireless robotic materials. Robotic materials are multi-functional composites that tightly integrate sensing, actuation, computation and communication to create smart composites that can sense their environment and change their physical properties in an arbitrary programmable manner. Computation and communication in such materials are based on miniature, possibly wireless, devices that are scattered in the material and interface with sensors and actuators inside the material. Whereas routing and processing of information within the material build upon results from the field of sensor networks, robotic materials are pushing the limits of sensor networks in both size (down to the order of microns) and numbers of devices (up to the order of millions). In order to solve the algorithmic and systems challenges of such an approach, which will involve not only computer scientists, but also roboticists, chemists and material scientists, the community requires a common platform --- much like the "Mote" that bootstrapped the widespread adoption of the field of sensor networks --- that is small, provides ample of computation, is equipped with basic networking functionalities, and preferably can be powered wirelessly.

Cyberbullying is a major problem affecting more than half of all American teens. Prior work has largely focused on detecting cyberbullying after the fact. In this paper, we investigate the prediction of cyberbullying incidents in Instagram, a popular media-based social network. The novelty of this work is building a predictor that can anticipate the occurrence of cyberbullying incidents before they happen. The Instagram media-based social network is well-suited to such prediction since there is an initial posting of an image typically with an associated text caption, followed later by the text comments that form the basis of a specific cyberbullying incident. We extract several important features from the initial posting data for automated cyberbullying prediction, including profanity and linguistic content of the text caption, image content, as well as social graph parameters and temporal content behavior. Evaluations using a real-world Instagram dataset demonstrate that our method achieves high performance in predicting the occurrence of cyberbullying incidents.

The theme of cybersafety is an important emerging research topic on the Internet that manifests itself daily as users navigate the Web and networked applications. Examples of cybersafety issues include cyberbullying, cyberthreats, recruiting minors via Internet services for nefarious purposes, using deceptive means to dupe vulnerable populations, exhibiting misbehaving behaviors such as using profanity or flashing in online video chats, and many others. These issues have a direct negative impact on the social, psychological and in some cases physical well-being of the end users. An important characteristic of these issues is that they fall in a grey legal area, where perpetrators may claim freedom of speech or rights to free expression despite causing harm. The main goal of this inaugural workshop on cybersafety is to bring together the researchers and practitioners from academia, industry, government and research labs working in the area of cybersafety to discuss the unique challenges in addressing various cybersafety issues and to share experiences, solutions, tools, and techniques. The focus is on the detection, prevention and mitigation of various cybersafety issues, as well as education and promoting safe practices.

As online social networks have grown in popularity, teenage users have become increasingly exposed to the threats of cyberbullying. The primary goal of this research paper is to investigate cyberbullying behaviors in Vine, a mobile based video-sharing online social network, and design novel approaches to automatically detect instances of cyberbullying over Vine media sessions. We first collect a set of Vine video sessions and use CrowdFlower, a crowd-sourced website, to label the media sessions for cyberbullying and cyberaggression. We then perform a detailed analysis of cyberbullying behavior in Vine. Based on the labeled data, we design a classifier to detect instances of cyberbullying and evaluate the performance of that classifier.

Usage of social networks has exploded over the past decade or so. Users now routinely share their thought, opinions, feelings as well as their daily activities on various social networks. An interesting consequence of this explosive usage of social networks is that it is possible to glean the current mood and emotion of a user from his or her social network postings. A question that arises in this context is: Can we use any differentiating features exhibited by people on their online social activities to build appropriate classifiers that can identify the positivity or negativity of users with high accuracy and low false positive and negative rates?

Cyberbullying is a major problem affecting more than half of all American teens, and has been attributed to suicidal behavior among teens. Instagram, a media-based mobile social network, is one of the most popular social networks used for cyberbullying. In this paper, we describe the development of classifiers to detect cyberbullying in Instagram. We identify systems issues that need to be considered in the design of a cyberbullying detection system.

We present algorithms for gesture recognition using in-network processing in distributed sensor arrays embedded within systems such as tactile input devices, sensing skins for robotic applications, and smart walls. We describe three distributed gesture-recognition algorithms that are designed to function on sensor arrays with minimal computational power, limited memory, limited bandwidth, and possibly unreliable communication. These constraints cause storage of gesture templates within the system and distributed consensus algorithms for recognizing gestures to be difficult. Building up on a chain vector encoding algorithm commonly used for gesture recognition on a central computer, we approach this problem by dividing the gesture dataset between nodes such that each node has access to the complete dataset via its neighbors. Nodes share gesture information among each other, then each node tries to identify the gesture. In order to distribute the computational load among all nodes, we also investigate an alternative algorithm, in which each node that detects a motion will apply a recognition algorithm to part of the input gesture, then share its data with all other motion nodes. Next, we show that a hybrid algorithm that distributes both computation and template storage can address trade-offs between memory and computational efficiency.

Augmented reality (AR) applications have recently become popular on modern smartphones. We explore the effectiveness of this mobile AR technology in the context of grocery shopping, in particular as a means to assist shoppers in making healthier decisions as they decide which grocery products to buy. We construct an AR-assisted mobile grocery-shopping application that makes real-time, customized recommendations of healthy products to users and also highlights products to avoid for various types of health concerns, such as allergies to milk or nut products, low-sodium or low-fat diets, and general caloric intake. We have implemented a prototype of this AR-assisted mobile grocery shopping application and evaluated its effectiveness in grocery store aisles. Our application's evaluation with typical grocery shoppers demonstrates that AR overlay tagging of products reduces the search time to find healthy food items, and that coloring the tags helps to improve the user's ability to quickly and easily identify recommended products, as well as products to avoid. We have evaluated our application's functionality by analyzing the data we collected from 15 in-person actual grocery-shopping subjects and 104 online application survey participants.

Cyberbullying has emerged as an important and growing social problem, wherein people use online social networks and mobile phones to bully victims with offensive text, images, audio and video on a 24/7 basis. This paper studies negative user behavior in the Ask.fm social network, a popular new site that has led to many cases of cyberbullying, some leading to suicidal behavior.We examine the occurrence of negative words in Ask.fm's question+answer profiles along with the social network of “likes” of questions+answers. We also examine properties of users with “cutting” behavior in this social network.

This paper investigates the development of accurate and efficient classifiers to identify misbehaving users (i.e., “flashers”) in a mobile video chat application. Our analysis is based on video session data collected from a mobile client that we built that connects to a popular random video chat ser

Online video chat services such as Chatroulette and Omegle randomly match users in video chat sessions and have become increasingly popular, with tens of thousands of users online at anytime during a day. Our interest is in examining user behavior in the growing domain of mobile video, and in particular how users behave in such video chat services as they are extended onto mobile clients. To date, over four thousand people have downloaded and used our Android-based mobile client, which was developed to be compatible with an existing video chat service. The paper provides a first-ever detailed large scale study of mobile user behavior in a random video chat service over a three week period. This study identifies major characteristics such as mobile user session durations, time of use, demographic distribution and the large number of brief sessions that users click through to find good matches. Through content analysis of video and audio, as well as analysis of texting and clicking behavior, we discover key correlations among these characteristics, e.g., normal mobile users are highly correlated with using the front camera and with the presence of a face, whereas misbehaving mobile users have a high negative correlation with the presence of a face.

Online video chat services such as Chatroulette, Omegle, and vChatter that randomly match pairs of users in video chat sessions are quickly becoming very popular, with over a million users per month in the case of Chatroulette. A key problem encountered in such systems is the presence of flashers and obscene content. This problem is especially acute given the presence of underage minors in such systems. This article presents SafeVchat, a novel solution to the problem of flasher detection that employs an array of image detection algorithms. A key contribution of the article concerns how the results of the individual detectors are fused together into an overall decision classifying a user as misbehaving or not, based on Dempster-Shafer theory. The article introduces a novel, motion-based skin detection method that achieves significantly higher recall and better precision. The proposed methods have been evaluated over real-world data and image traces obtained from Chatroulette.com. SafeVchat has been deployed in Chatroulette. A combination of SafeVchat with human moderation has resulted in banning as many as 50,000 inappropriate users per day on Chatoulette. Furthermore, offensive content on Chatoulette has dropped significantly from 33.08% (before SafeVchat installation) to 3.49% (after SafeVchat installation).

The need for highly scalable and accurate detection and filtering of misbehaving users and obscene content in online video chat services has grown as the popularity of these services has exploded in popularity. This is a challenging problem because processing large amounts of video is compute intensive, decisions about whether a user is misbehaving or not must be made online and quickly, and moreover these video chats are characterized by low quality video, poorly lit scenes, diversity of users and their behaviors, diversity of the content, and typically short sessions. This paper presents EMeralD, a highly scalable system for accurately detecting and filtering misbehaving users in online video chat applications. EMeralD substantially improves upon the state-of-the-art filtering mechanisms by achieving much lower computational cost and higher accuracy. We demonstrate EMeralD's improvement via experimental evaluations on real-world data sets obtained from Chatroulette.com.

Online video chat services such as Chatroulette [1] and Omegle [2] that randomly match pairs of users in video chat sessions have become increasingly popular, with over twenty thousand online users at anytime during a day. A key problem encountered in such systems is the presence of misbehaving users ("flashers") and obscene content. Our previous works [3] [4] prove that using some image recognition methods (skin-detection, dense SIFT) and machine learning algorithms could achieve significantly higher recall and better precision for flasher detection. Nowadays, with the rapid development of advanced mobile phones with both front and back cameras, we expect mobile video chat to become a popular extension of online video chat services. However, because of the computation-intensive features used by our previous solutions and mobile phones' hardware limitations such as memory size and CPU capacity, it is difficult to directly apply our previous works to mobile platforms. As smartphones are increasingly equipped with diverse sensing capabilities, we plan to utilize this multi-dimensional sensor information to extend flasher detection on mobile platform. This project explores how we can mine accelerometer and other mobile sensor data to infer some clues to optimize flasher detection accuracy while reducing the computation demands of flasher detection on the mobile device.

Online video chat services, such as Chatroulette, Omegle, and vChatter are becoming increasingly popular and have attracted millions of users. One critical problem encountered in such applications is the presence of misbehaving users ("flashers") and obscene content. Automatically filtering out obscene content from these systems in an efficient manner poses a difficult challenge. This paper presents a novel Fine-Grained Cascaded (FGC) classification solution that significantly speeds up the compute-intensive process of classifying misbehaving users by dividing image feature extraction into multiple stages and filtering out easily classified images in earlier stages, thus saving unnecessary computation costs of feature extraction in later stages. Our work is further enhanced by integrating new webcam-related contextual information (illumination and color) into the classification process, and a 2-stage soft margin SVM algorithm for combining multiple features. Evaluation results using real-world data set obtained from Chatroulette show that the proposed FGC based classification solution significantly outperforms state-of-the-art techniques.

Emergency applications have recently become widely available on modern smart phones. Nearly all of these commercial applications have focused on providing simple accident information in outdoor settings. AR in indoor environments poses unique challenges, due to the unavailability of GPS indoors and WiFi-based positioning limitations. In this paper, we propose the use of Rescue Me, a novel system based on indoor mobile AR applications using personalized pedometry and one that recommends the most optimal, uncrowded exit path to users. We have developed the Rescue Me application for use within large scale buildings, with complex paths. We show how Rescue Meleverages the sensors on a smart phone, in conjunction with emergency information and daily-based user behavior, to deliver evacuation information in emergency situations.

Online video chat services such as Chatroulette, Omegle, and vChatter that randomly match pairs of users in video chat sessions are fast becoming very popular, with over a million users per month in the case of Chatroulette. A key problem encountered in such systems is the presence of flashers and obscene content. This problem is especially acute given the presence of underage minors in such systems. This paper presents SafeVchat, a novel solution to the problem of flasher detection that employs an array of image detection algorithms. A key contribution of the paper concerns how the results of the individual detectors are fused together into an overall decision classifying the user as misbehaving or not, based on Dempster-Shafer Theory. The paper introduces a novel, motion-based skin detection method that achieves significantly higher recall and better precision. The proposed methods have been evaluated over real-world data and image traces obtained from Chatroulette.com.

Group recommendation, which makes recommendations to a group of users instead of individuals, has become increasingly important in both the workspace and people’s social activities, such as brainstorming sessions for coworkers and social TV for family members or friends. Group recommendation is a challenging problem due to the dynamics of group memberships and diversity of group members. Previous work focused mainly on the content interests of group members and ignored the social characteristics within a group, resulting in suboptimal group recommendation performance. In this work, we propose a group recommendation method that utilizes both social and content interests of group members. We study the key characteristics of groups and propose (1) a group consensus function that captures the social, expertise, and interest dissimilarity among multiple group members; and (2) a generic framework that automatically analyzes group characteristics and constructs the corresponding group consensus function. Detailed user studies of diverse groups demonstrate the effectiveness of the proposed techniques, and the importance of incorporating both social and content interests in group recommender systems.

Software failures in wireless sensor systems are notoriously difficult to debug. Resource constraints in wireless deployments substantially restrict visibility into the root causes of node-level system and application faults. At the same time, the high cost of deployment ofwireless sensor systems often far exceeds the cumulative cost of allother sensor hardware, so that software failures that completely disable a node are prohibitively expensive to repair in real worldapplications, e.g. by on-site visits to replace or reset nodes. We describe NodeMD, a deployment management system that successfully implements lightweight run-time detection, logging, and notificationof software faults on wireless mote-class devices. NodeMD introduces a debug mode that catches a failure before it completely disables a node and drops the node into a stable state that enables further diagnosis and correction, thus avoiding on-site redeployment. We analyze the performance of NodeMD on a real world application ofwireless sensor systems.

An airborne wireless sensor network (WSN) composed of bird-sized micro aerial vehicles (MAVs) enables low cost high granularity atmospheric sensing of toxic plume behavior and storm dynamics, and provides a unique three-dimensional vantage for monitoring wildlife and ecological systems. This paper describes a complete implementation of our SensorFlock airborne WSN, spanning the development of our MAV airplane, its avionics, semi-autonomous flight control software, launch system, flock control algorithm, and wireless communication networking between MAVs. We present experimental results from flight tests of flocks of MAVs, and a characterization of wireless RF behavior in air-to-air communication as well as air-to-ground communication.

Remote reprogramming of in situ wireless sensor networks (WSNs) via the wireless link is an important capability. Securing the process of reprogramming allows each sensor node to authenticate each received code image. Due to the resource constraints of WSNs, public key schemes must be used sparingly. This paper introduces a mechanism for secure and efficient code distribution that employs public key cryptography only to sign the root of a combined structure consisting of both hash chains and hash trees. The chain based scheme works best when packets are received in the order they are sent with very few losses. Our hash tree based scheme allows nodes to authenticate packets and verify their integrity quickly, even when the packets may arrive out of order, but can result in too many public key operations. Integrating hash chains and hash trees produces a mechanism that is both resilient to losses and lightweight in terms of reducing memory consumption and the number of public key operations that a node has to perform. Simulation shows that the proposed secure reprogramming schemes add only a modest amount of overhead to a conventional non-secure reprogramming scheme, namely Deluge, and are therefore feasible and practical in a wireless sensor network.

In this paper we present FireWxNet, a multi-tiered portable wireless system for monitoring weather conditions in rugged wildland fire environments. FireWxNet provides the fire fighting community the ability to safely and easily measure and view fire and weather conditions over a wide range of locations and elevations within forest fires. This previously unattainable information allows fire behavior analysts to better predict fire behavior, heightening safety considerations. Our system uses a tiered structure beginning with directional radios to stretch deployment capabilities into the wilderness far beyond current infrastructures. At the end point of our system we designed and integrated a multi-hop sensor network to provide environmental data. We also integrated web-enabled surveillance cameras to provide visual data. This paper describes a week long full system deployment utilizing 3 sensor networks and 2 web-cams in the Selway-Salmon Complex Fires of 2005. We perform an analysis of system performance and present observations and lessons gained from our deployment.

Deployments of wireless LANs consisting of hundreds of 802.11 access points with a large number of users have been reported in enterprises as well as college campuses. However, due to the unreliable nature of wireless links, users frequently encounter degraded performance and lack of coverage. This problem is even worse in unplanned networks, such as the numerous access points deployed by homeowners. Existing approaches that aim to diagnose these problems are inefficient because they troubleshoot at too high a level, and are unable to distinguish among the root causes of degradation. This paper designs, implements, and tests fine-grained detection algorithms that are capable of distinguishing between root causes of wireless anomalies at the depth of the physical layer. An important property that emerges from our system is that diagnostic observations are combined from multiple sources over multiple time instances for improved accuracy and efficiency.

In this paper we present X-MAC, a low power MAC protocol for wireless sensor networks (WSNs). Standard MAC protocols developed for duty-cycled WSNs such as BMAC, which is the default MAC protocol for TinyOS, employ an extended preamble and preamble sampling. While this "low power listening" approach is simple, asynchronous, and energy-efficient, the long preamble introduces excess latency at each hop, is suboptimal in terms of energy consumption, and suffers from excess energy consumption at nontarget receivers. X-MAC proposes solutions to each of these problems by employing a shortened preamble approach that retains the advantages of low power listening, namely low power communication, simplicity and a decoupling of transmitter and receiver sleep schedules. We demonstrate through implementation and evaluation in a wireless sensor testbed that X-MAC's shortened preamble approach significantly reduces energy usage at both the transmitter and receiver, reduces per-hop latency, and offers additional advantages such as flexible adaptation to both bursty and periodic sensor data sources.

Global warming is an increasing concern worldwide. Assessing the contribution of CO2 to this phenomenon is an important issue. This project's goal is to improve understanding of CO2 and H2O transport in a mountainous terrain that confound current efforts to resolve CO2 budgets at regional and global scales.

The irregularities of a low cost wireless communication interface, changing environmental conditions, in-situ deployment and scarce resources make management, monitoring and troubleshooting performance of a sensor network a challenging task. In this paper we present the design of a decentralized fault diagnosis system for a wireless sensor network. Our system distinguishes between multiple root causes of degraded performance and provides efficient feedback into the network to troubleshoot the fault

Establishing secure links between pairs of directly connected sensor nodes is an important primitive for building secure wireless sensor networks. This paper systematically identifies two important security requirements of pairwise key setup in wireless sensor networks, namely opaqueness and inoculation. Transitory master key schemes, such as the LEAP protocol, can satisfy both requirements if the master key has not been compromised. However, if the master key is compromised, every key in the network is exposed to an adversary. To prevent the master key from becoming a single point failure of the whole system, we propose a new opaque transitory master key (OTMK) scheme for pairwise key setup in sensor networks. In OTMK, even if the master key is compromised, an adversary can only exploit a small number of keys nearby the compromised node, while other keys in the network remain safe. To further investigate key establishment schemes, we experimented with a way to compromise a sensor node, and tested our key establishment time in a real sensor network environment.

Wireless sensor networks are highly vulnerable to the failure of base stations. An adversary can render a wireless sensor network useless by launching remote, softwarebased attacks or physical attacks on the base stations. This paper addresses the problem of defending a base station against physical attacks by concealing the geographic location of a base station. Typical packet traffic in a sensor network reveals pronounced patterns that allow an adversary analyzing packet traffic to deduce the location of a base station. The paper investigates several countermeasures against traffic analysis techniques aimed at disguising the location of a base station. First, a degree of randomness is introduced in the multi-hop path a packet takes from a sensor node to a base station. Second, random fake paths are introduced to confuse an adversary from tracking a packet as it moves towards a base station. Finally, multiple, random areas of high communication activity are created to deceive an adversary as to the true location of the base station. The paper evaluates these techniques analytically and via simulation using three evaluation criteria: total entropy of the network, total energy consumed, and the ability to guard against heuristic-based techniques to locate a base station.

Group communication to and from sets of sensor nodes is an important paradigm in wireless sensor networks (WSNs). Securing this group communication is a difficult challenge given the energy-efficiency constraints posed by WSNs. In this paper, we introduce the protocol SLIMCAST, i.e. Secure Level key Infrastructure for MultiCAST and group communication, which uses level keys to provide an infrastructure that dramatically lowers the cost of nodes joining and leaving sensor groups. This level key infrastructure is shown to achieve energyefficient key updates that are localized for group multicast 1→N communication, and can be further leveraged to achieve secure group aggregation N→1 communication. Simulation results comparing the performance of SLIMCAST to traditional secure group communication protocols are presented to demonstrate SLIMCAST’s energy efficiency and flexibility.

Wireless sensor networks face acute security concerns in applications such as battlefield monitoring. A central point of failure in a sensor network is the base station, which acts as a collection point of sensor data. In this paper, we investigate two attacks that can lead to isolation or failure of the base station. In one set of attacks, the base station is isolated by blocking communication between sensor nodes and the base station, e.g. by DOS attacks. In the second attack, the location of the base station is deduced by analyzing data traffic towards the base station, which can lead to jamming and/or discovery and destruction of the base station. To defend against these attacks, two secure strategies are proposed. First, secure multi-path routing to multiple destination base stations is designed to provide intrusion tolerance against isolation of a base station. Second, anti-traffic analysis strategies are proposed to help disguise the location of the base station from eavesdroppers. A performance evaluation is provided for a simulated sensor network, as well as measurements of cryptographic overhead on real sensor nodes.

Today's architecture for interconnecting wireless sensor networks (WSNs) and the Internet is based on treating a WSN as a separate entity from the Internet. Our approach to unifying sensor networks with the Internet is instead to decouple the relationship between the API-enforced database view and the gateway, making the gateway more general by introducing application-level overlay networking into the gateway. In our approach, sensor network packets, rather than being stopped at the gateway, are instead encapsulated into IP packets, and then directed from the gateway to any number of interested applications and services residing remotely on the Internet.

An efficient and reliable file storage system is important to micro sensor nodes so that data can be logged for later asynchronous delivery across a multi-hop wireless sensor network. Designing and implementing such a file system for a sensor node faces various challenges. Sensor nodes are highly resource constrained in terms of limited runtime memory, limited persistent storage, and finite energy. Also, the flash storage medium on sensor nodes differs in a variety of ways from the traditional hard disk, e.g. in terms of the limited number of writes for a flash memory unit. We present the design and implementation of ELF, an efficient log-structured flash-based file system tailored for sensor nodes. ELF is adapted to achieve memory efficiency, low power operation, and tailored support for common types of sensor file operations such as appending data to a file. ELF's log-structured approach achieves wear levelling across flash memory pages with limited write lifetimes. ELF also uniquely provides garbage collection capability as well as reliability for micro sensor nodes. A performance evaluation of an implementation of ELF based on TinyOS and MICA2 sensor motes is presented.

By spreading the workload across a sensor network, load balancing reduces hot spots in the sensor network and increases the energy lifetime of the sensor network. In this paper, we design a node-centric algorithm that constructs a load-balanced tree in sensor networks of asymmetric architecture. We utilize a Chebyshev Sum metric to evaluate via simulation the balance of the routing trees produced by our algorithm. We find that our algorithm achieves routing trees that are more effectively balanced than the routing based on breadth-first search (BFS) and shortest-path obtained by Dijkstra's algorithm.

Students often turn to their peers for help in order to learn a new concept or lesson introduced by a teacher in class. This establishes roles of tutor and learner between students, which can also reverse depending on the subject, with the teacher as a third-party mediator. We discuss our design of a distributed software system architecture that seeks to harness the tutor-learner relationship between peers into a collaborative learning system. Our goal is to provide a verifiable, portable, and inexpensive system of coordinated wireless handhelds that both promotes learning of lesson plans by the students and enhances the tutoring skills of students.

The MANTIS MultimodAl system for NeTworks of In-situ wireless Sensors provides a new multithreaded embedded operating system integrated with a general-purpose single-board hardware platform to enable flexible and rapid prototyping of wireless sensor networks.

Wireless sensor networks require lightweight routing tailored for sensor devices with severe memory, power, and cost constraints. Such lightweight protocols must also support mobility and fault tolerance. The very Lightweight Mobile Multicast (VLM/sup 2/) system addresses these concerns, introducing multicast support into wireless sensor networks. In simulation and in a true implementation on hardware Motes, VLM/sup 2/ achieves multicast with a lightweight footprint of no more than 17 kb per node and also responds with agility to a wide range of mobility.

Due to the limited processing capability, memory constraints, and the power budget of mobile clients, multimedia coders and/or decoders are often difficult to implement on wireless handheld PDAs. In this Universal Tuner project, we designed and implemented a wireless video streaming system that transcodes MPEG-1/2 videos or live TV broadcasting videos to the BW or indexed color Palm OS devices. In our system, the complexity of multimedia compression and decompression algorithms is adaptively partitioned between the encoder and decoder. A mobile client would selectively disable or reenable stages of the algorithm to adapt to the device's effective processing capability. Our variable-complexity strategy of selective disabling of modules supports graceful degradation of the complexity of multimedia coding and decoding into a mobile client's low-power mode, i.e. the clock frequency of its next-generation low power CPU has been scaled down to conserve power. We modified the structure of the standard motion-compensated DCT video codecs to implement a simplified the encoder on a PC server and the decoder on a complexity-constrained PDA viewing client.

WebSplitter symbolizes the union of pervasive multi-device computing and collaborative multi-user computing. WebSplitter provides a unified XML framework that enables multi-device and multi-user Web browsing. WebSplitter splits a requested Web page and delivers the appropriate partial view of each page to each user, or more accurately to each user's set of devices. Multiple users can participate in the same browsing session, as in traditional conferencing groupware. Depending on the access privileges of the user to the different components of content on each page, WebSplitter generates a personalized partial view. WebSplitter further splits the partial view among the devices available to each user, e.g. laptop, wireless PDA, projection display, stereo speakers, orchestrating a composite presentation across the devices. A wireless PDA can browse while remotely controlling the multimedia capabilities of nearby devices. The architecture consists of an XML metadata policy file defining access privileges to XML tags on a Web page, a middleware proxy that splits XML Web content to create partial views, and a client-side component, e.g. applet, enabling user login and reception of pushed browsing data. Service discovery finds and registers proxies, browsing sessions, and device capabilities. We demonstrate the feasibility of splitting the different tags in an XML Web page to different end users browsers, and of pushing updates from the browsing session to heterogeneous devices, including a laptop and a PDA.