The demand for Deep Neural Network (DNN) execution (including both inference and training) on mobile system-ona-chip (SoCs) has surged, driven by factors like the need for real-time latency, privacy, and reducing vendors' costs.Mainstream mobile GPUs (e.g., Qualcomm Adreno GPUs) usually have a 2.5D L1 texture cache that offers throughput superior to that of on-chip memory.However, to date, there is limited understanding of the performance features of such a 2.5D cache, which limits the optimization potential.This paper introduces TMModel, a framework with three components: 1) a set of micro-benchmarks and a novel performance assessment methodology to characterize a non-well-documented architecture with 2D memory, 2) a complete analytical performance model configurable for different data access pattern(s), tiling size(s), and other GPU execution parameters for a given operator (and associated size and shape), and 3) a compilation framework incorporating this model and generating optimized code with low overhead.TMModel is

Recognizing food types through sensor signals for unseen users remains remarkably challenging despite extensive recent studies. The efficacy of prior machine learning techniques is dwarfed by giant variations of data collected from multiple participants, partly because users have varied chewing habits and wear sensor devices in various manners. This work treats the problem as an instance of the domain adaptation problem, where each user represents a domain. We develop the first multi-source domain adaptation (MSDA) method for food-typing recognition, which consists of three major components: stratified normalization, a multi-source domain adaptor, and adaptive ensemble learning. New techniques are developed for each component. Using a real-world dataset comprised of 15 participants, we demonstrate that our method achieves \(1.33\times\) to \(2.13\times\) improvement in accuracy compared with nine state-of-the-art MSDA baselines. Additionally, we perform an in-depth ablation study to examine the behavior of each component and confirm its efficacy.

Finger-vein recognition has attracted extensive attention due to its exceptional level of security and privacy. Recently, deep neural networks (DNNs), such as convolutional neural networks (CNNs) showing robust capacity for feature representation, have been proposed for vein recognition. The architectures of these DNNs, however, have primarily been manually designed based on human prior knowledge, which is both time-consuming and error-prone. To overcome these problems, we propose AG-NAS, an Attention Gated recurrent unit-based Neural Architecture Search to automatically search for the optimal network architecture, thereby improving the recognition performance for different finger-vein recognition tasks. First, we combine the self-attention mechanism and gated recurrent unit (GRU) to propose an attention GRU module employed as a controller to generate the architectural hyperparameters of candidate neural networks automatically. Second, we investigate a parameter-sharing supernet policy to reduce the search space, computation, and time costs. Finally, we conduct rigorous experiments on our finger-vein database and two public finger-vein databases. The experimental results demonstrate that the proposed AG-NAS outperforms the representative approaches and achieves state-of-the-art recognition accuracy.

Over the last years, sensor-based continuous authentication on mobile devices has achieved great success on personal information protection. These proposed mechanisms, however, require both legal and illegal users’ data for authentication model training, which takes time and is impractical. In this paper, we present MAuGANs, a lightweight and practical Memory-Augmented Autoencoder-based continuous Authentication system on smartphones with conditional transformer Generative Adversarial Networks (GANs), where the conditional transformer GANs (CTGANs) are used for data augmentation and the memory-augmented autoencoder (MAu) is utilized to identify users. Specifically, MAuGANs exploits the smartphone built-in accelerometer and gyroscope sensors to implicitly collect users’ behavioral patterns. With the normalized legitimate user's sensor data, MAuGANs uses a CTGAN composed of a conditional transformer-based generator and a conditional transformer-based discriminator to create additional training data for the MAu. Then, the MAu is trained on the augmented legitimate user's data. The trained MAu reconstructs the current user data and then calculates the reconstruction error between the reconstructed data and current user data. To carry out user authentication, MAuGANs compares the reconstruction error with a predefined authentication threshold. We evaluate the performance of MAuGANs on our dataset, where our extensive experiments demonstrate that MAuGANs reaches the best authentication performance, when comparing with the representative state-of-the-art methods, by 0.33% EER and 99.65% accuracy on 10 unseen users.

With the rapid development of the Internet of Things (IoTs) and mobile communications, mobile devices have become indispensable in our daily lives. Given the substantial amount of private information stored on these devices, the security of mobile devices has emerged as a significant concern for users. Different from conventional methods such as PINs, fingerprints, and face IDs, which authenticate users only during the initial login stage, continuous authentication ensures consistent verification while mobile devices are in use. Current continuous authentication methods require extensive data from a series of users for effective training. Nevertheless, it is challenging to collect sufficient amount of data within a limited time. In this paper, we propose CALL, an unsupervised sensor-based Continuous Authentication system with a Low-rank transformer using Learning-to-rank algorithms. The lightweight CALL is capable of providing both spatial and temporal features for end-to-end authentication. Specifically, CALL utilizes time series data from a legitimate user, collected by the accelerometer, gyroscope, and magnetometer sensors on smartphones, to train a pure one-dimensional autoencoder for spatial features and a shuffle low-rank Transformer (SLRT) for temporal features in the training phase. In the authentication phase, the trained pure one-dimensional autoencoder captures spatial features by reconstructing input data to obtain the reconstruction error, and SLRT captures temporal features by predicting a ranking vector that reveals the order of the shuffled feature sequence. The predicted ranking vector is then used to recover the shuffled sequence and the similarity between the frequency spectrum sequences of the recovered sequence and the original time series data is calculated. The reconstruction error and similarity are compared against pre-defined thresholds, and CALL authenticates a user as legitimate only if both values fall below their respective thresholds. Finally, we evaluate the performance of CALL on UCI_HAR, WISDM_HARB, and our dataset, and the extensive experiments illustrate that CALL reaches the best performance with 96.43%, 95.24% and 96.92% accuracy, and 4.28%, 4.76% and 3.86% EERs on the three datasets, outperforming state-of-the-art continuous authentication methods.

Deep neural networks can be easily fooled by the adversarial example, which is a specially crafted example with subtle and intentional perturbations. A plethora of papers have proposed to use filters to effectively defend against adversarial example attacks. However, we demonstrate that the automatic filter-based defenses may not be reliable. In this article, we present URL2AED, Using a Reinforcement Learning scheme TO escape the automatic filter-based Adversarial Example Defenses. Specifically, URL2AED uses a specially crafted policy gradient reinforcement learning (RL) algorithm to generate adversarial examples (AEs) that can escape automatic filter-based AE defenses. In particular, we properly design reward functions in policy-gradient RL for targeted attacks and non-targeted attacks, respectively. Furthermore, we customize training algorithms to reduce the possible action space in policy-gradient RL to accelerate URL2AED training while still ensuring that URL2AED generates successful AEs. To demonstrate the performance of the proposed URL2AED, we conduct extensive experiments on three public datasets in terms of different perturbation degrees of parameter, different filter parameters, transferability, and time consumption. The experimental results show that URL2AED achieves high attack success rates for automatic filter-based defenses and good cross-model transferability.

Graph embedding converts a graph data into a low dimensional space to preserve the original graph information. However, graph data can be reconstructed by malicious adversaries to train machine learning models from graph embeddings. This paper studies to what extent an adversary (without the original graph data) can recover the original graph data from graph embeddings. To quantify the original graph information leakage from graph embeddings, we develop a deep neural network model InferNet that can be used by adversaries to infer the original graph information from an adversary-accessible graph embedding database. More specifically, we propose the data-free reversed knowledge distillation technique to support the InferNet training even if the original graph dataset is absent. To ensure the performance of InferNet, we design two cycle-consistency loss functions to have an interactive training of InferNet over three series of datasets. To further enhance the performance of InferNet, we provide a joint training algorithm that simultaneously trains the pseudo-sample generator and InferNet, which significantly reduces the storage space. We evaluate the performance of InferNet on three datasets, and the intensive experiments demonstrate that InferNet can infer the original graph information from the graph embedding dataset with high accuracy.

Deep learning has achieved significant success on intelligent medical treatments, such as automatic diagnosis and analysis of medical data. To train an automatic diagnosis system with high accuracy and strong robustness in healthcare, sufficient training data are required when using deep learning-based methods. However, given that the data collected by sensors that are embedded in medical or mobile devices are inadequate, it is challenging to train an effective and efficient classification model with state-of-the-art performance. Inspired by generative adversarial networks (GANs), we propose TS-GAN, a Time-series GAN architecture based on long short-term memory (LSTM) networks for sensor-based health data augmentation, thereby improving the performance of deep learning-based classification models. TS-GAN aims to learn a generative model that creates time-series data with the same space and time dependence as the real data. Specifically, we design an LSTM-based generator for creating realistic data and an LSTM-based discriminator for determining how similar the generated data are to real data. In particular, we design a sequential-squeeze-and-excitation module in the LSTM-based discriminator to better understand space dependence of real data, and apply the gradient penalty originated from Wasserstein GANs in the training process to stabilize the optimization. We conduct comparative experiments to evaluate the performance of TS-GAN with TimeGAN, C-RNN-GAN and Conditional Wasserstein GANs through discriminator loss, maximum mean discrepancy, visualization methods and classification accuracy on health datasets of ECG_200, NonInvasiveFatalECG_Thorax1, and mHealth, respectively. The experimental results show that TS-GAN exceeds other state-of-the-art time-series GANs in almost all the evaluation metrics, and the classifier trained on synthetic datasets generated by TS-GAN achieves the highest classification accuracy of 97.50% on ECG_200, 94.12% on NonInvasiveFatalECG_Thorax1, and 98.12% on mHealth, respectively.

Mobile devices are becoming increasingly popular and are playing significant roles in our daily lives. Insufficient security and weak protection mechanisms, however, cause serious privacy leakage of the unattended devices. To fully protect mobile device privacy, we propose ADFFDA, a novel mobile continuous authentication system using an Adaptive Deep Feature Fusion scheme for effective feature representation, and a transformer-based GAN for Data Augmentation, by leveraging smartphone built-in sensors of the accelerometer, gyroscope and magnetometer. Given the normalized sensor data, ADFFDA utilizes the transformer-based GAN consisting of a transformer-based generator and a CNN-based discriminator to augment the training data for CNN training. With the augmented data and the especially-designed CNN based on the ghost module and ghost bottleneck, ADFFDA extracts deep features from the three sensors by the trained CNN, and exploits an adaptive-weighted concatenation method to adaptively fuse the CNN-extracted features. Based on the fused features, ADFFDA authenticates users by using the one-class SVM (OC-SVM) classifier. We evaluate the authentication performance of ADFFDA in terms of the efficiency of the transformer-based GAN, GAN-based data augmentation, CNN architecture, adaptive-weighted feature fusion, OC-SVM classifier, and security analysis. The experimental results show that ADFFDA obtains the best authentication performance w.r.t representative approaches, by achieving a mean equal error rate of 0.01%.

Mobile devices have been playing significant roles in our daily lives, which has made device security and privacy protection extremely important. These mobile devices storing user sensitive and private information, therefore, need rigorous user authentication mechanisms. In this article, we present SearchAuth, a novel continuous authentication system on smartphones exploiting a neural architecture search (NAS) to find an optimal network architecture and an auto augmentation search (AAS) to more effectively train the optimal network along with the best data augmentation policies, by leveraging the accelerometer, gyroscope, and magnetometer on smartphones to capture users’ behavioral patterns. Specifically, SearchAuth consists of three stages, i.e., the offline stage, registration stage, and authentication stage. In the offline stage, we utilize the NAS on sensor data of the accelerometer, gyroscope, and magnetometer to find an optimal network architecture based on the designed search space. With the optimal network architecture, namely, NAS-based model, the AAS automatically optimizes the augmentation of the input data for more effectively training the model that is for feature extraction. In the registration stage, we use the trained NAS-based model to learn and extract deep features from the legitimate user’s data, and train the LOF classifier with 55 features selected by the PCA. In the authentication stage, with the well-trained NAS-based model and LOF classifier, SearchAuth identifies the current user as a legitimate user or an impostor when the user starts operating a smartphone. Based on our dataset, we evaluate the performance of the proposed SearchAuth, and the experimental results demonstrate that SearchAuth surpasses the representative authentication schemes by achieving the best accuracy of 93.95%, F1-score of 94.30%, and EER of 5.30% on the LOF classifier with dataset size of 100.

As the number of IoT devices has increased rapidly, IoT botnets have exploited the vulnerabilities of IoT devices. However, it is still challenging to detect the initial intrusion on IoT devices prior to massive attacks. Recent studies have utilized power side-channel in-formation to identify this intrusion behavior on IoT devices but still lack accurate models in real-time for ubiquitous botnet detection. We propose the first online intrusion detection system called DeepAuditor for multiple IoT devices via power auditing. To de-velop the real-time system, we propose a lightweight power auditing device called Power Auditor. We also design a distributed CNN classifier for online inference in a laboratory setting. In order to protect data leakage and reduce networking redundancy, we then propose a privacy-preserved inference protocol via Packed Homo-morphic Encryption and a sliding window protocol in our system. The classification accuracy and processing time are measured, and the proposed classifier outperforms a baseline classifier, especially against unseen patterns. We also demonstrate that the distributed CNN design is secure against any distributed components. Over-all, the measurements are shown to the feasibility of our real-time distributed system for intrusion detection on IoT devices.

As the number of IoT devices has increased rapidly, IoT botnets have exploited the vulnerabilities of IoT devices. However, it is still challenging to detect the initial intrusion on IoT devices prior to massive attacks. Thus, a new approach that monitors these ini-tial intrusions is needed. Power side-channel information can be used because it does not require any modification in programming languages or operating systems on diverse IoT devices. We propose a distributed power side-channel auditing system for online IoT intrusion detection. To meet the real-time requirement, we develop a lightweight power auditing device. We then design a distributed CNN classifier for online inference in a laboratory setting. Two distributed protocols are also proposed in order to protect data leakage and reduce networking redundancy. In this work, we demonstrate the feasibility of our real-time distributed system for intrusion detection on IoT devices.

Automatic food type recognition is an essential task of dietary monitoring. It helps medical professionals recognize a user's food contents, estimate the amount of energy intake, and design a personalized intervention model to prevent many chronic diseases, such as obesity and heart disease. Various wearable and mobile devices are utilized as platforms for food type recognition. However, none of them has been widely used in our daily lives and, at the same time, socially acceptable enough for continuous wear. In this paper, we propose a food type recognition method that takes advantage of Airpods Pro, a pair of widely used wireless in-ear headphones designed by Apple, to recognize 20 different types of food. As far as we know, we are the first to use this socially acceptable commercial product to recognize food types. Audio and motion sensor data are collected from Airpods Pro. Then 135 representative features are extracted and selected to construct the recognition model using the lightGBM algorithm. A real-world data collection is conducted to comprehensively evaluate the performance of the proposed method for seven human subjects. The results show that the average f1-score reaches 94.4% for the ten-fold cross-validation test and 96.0% for the self-evaluation test.

Despite many conveniences of using IoT devices, they have suffered from various attacks due to their weak security. Besides well-known botnet attacks, IoT devices are vulnerable to recent covert-channel attacks. However, no study to date has considered these IoT covert-channel attacks. Among these attacks, researchers have demonstrated exfiltrating users' private data by exploiting the smart bulb's capability of infrared emission.

Tremor is a common symptom among Parkinson's disease (PD) patients at all stages. To measure tremor, we utilized IMU sensing data from the wrists while PD patients were drawing. With 30 patients' IMU sensing data obtained from standard tremor rating scale activities, we conducted data analysis for identifying any tremor episodes and extracting tremor amplitude. In this demo, we demonstrate that our preliminary analysis and results show the potential of measuring kinetic tremors effectively using these methods.

With the widespread usage of mobile devices, the authentication mechanisms are urgently needed to identify users for information leakage prevention. In this article, we present CAGANet, a convolutional neural network (CNN)-based continuous authentication on smartphones using a conditional Wasserstein generative adversarial network (CWGAN) for data augmentation, which utilizes smartphone sensors of the accelerometer, gyroscope, and magnetometer to sense phone movements incurred by user operation behaviors. Specifically, based on the preprocessed real data, CAGANet employs CWGAN to generate additional sensor data for data augmentation that are used to train the designed CNN. With the augmented data, CAGANet utilizes the trained CNN to extract deep features and then performs principal component analysis (PCA) to select appropriate representative features for different classifiers. With the CNN-extracted features, CAGANet trains four one-class classifiers of OC-SVM, LOF, isolation forest (IF), and EE in the enrollment phase and authenticates the current user as a legitimate user or an impostor based on the trained classifiers in the authentication phase. To evaluate the performance of CAGANet, we conduct extensive experiments in terms of the efficiency of CWGAN, the effectiveness of CWGAN augmentation and the designed CNN, the accuracy on unseen users, and comparison with traditional augmentation approaches and with representative authentication methods, respectively. The experimental results show that CAGANet with the IF classifier can achieve the lowest equal error rate (EER) of 3.64% on 2-s sampling data.

It is appealing but challenging to achieve real-time deep neural network (DNN) inference on mobile devices, because even the powerful modern mobile devices are considered as "resource-constrained" when executing large-scale DNNs. It necessitates the sparse model inference via weight pruning, i.e., DNN weight sparsity, and it is desirable to design a new DNN weight sparsity scheme that can facilitate real-time inference on mobile devices while preserving a high sparse model accuracy. This paper designs a novel mobile inference acceleration framework GRIM that is General to both convolutional neural networks (CNNs) and recurrent neural networks (RNNs) and that achieves Real-time execution and high accuracy, leveraging fine-grained structured sparse model Inference and compiler optimizations for Mobiles. We start by proposing a new fine-grained structured sparsity scheme through the Block-based Column-Row (BCR) pruning. Based on this new fine-grained structured sparsity, our GRIM framework consists of two parts: (a) the compiler optimization and code generation for real-time mobile inference; and (b) the BCR pruning optimizations for determining pruning hyperparameters and performing weight pruning. We compare GRIM with Alibaba MNN, TVM, TensorFlow-Lite, a sparse implementation based on CSR, PatDNN, and ESE (a representative FPGA inference acceleration framework for RNNs), and achieve up to 14.08× speedup.

Smartphones have become crucial and important in our daily life, but the security and privacy issues have been major concerns of smartphone users. In this article, we present DeFFusion, a CNN-based continuous authentication system using Deep Feature Fusion for smartphone users by leveraging the accelerometer and gyroscope ubiquitously built into smartphones. With the collected data, DeFFusion first converts the time domain data into frequency domain data using the fast Fourier transform and then inputs both of them into a designed CNN, respectively. With the CNN-extracted features, DeFFusion conducts the feature selection utilizing factor analysis and exploits balanced feature concatenation to fuse these deep features. Based on the one-class SVM classifier, DeFFusion authenticates current users as a legitimate user or an impostor. We evaluate the authentication performance of DeFFusion in terms of impact of training data size and time window size, accuracy comparison on different features over different classifiers and on different classifiers with the same CNN-extracted features, accuracy on unseen users, time efficiency, and comparison with representative authentication methods. The experimental results demonstrate that DeFFusion has the best accuracy by achieving the mean equal error rate of 1.00% in a 5-second time window size.

Joint angle analysis facilitates research into injury prevention, rehabilitation, and activity monitoring. Sensors used in such analysis must be unobtrusive, accurate, and if used for motion, capable of monitoring fast-paced, dynamic motions. To effectively contribute to these applications, we created a body-mounted electromagnet-based sensing system for joint angle analysis called Magneto. Our system is wireless, features a high sampling rate, is not subject to drift, and is unaffected by outside magnetic noise. Magnetic field readings are influenced by noise due to magnetic interference from the Earth's magnetic field, the environment, and nearby ferrous objects. Magneto uses the combination of an electromagnet and magnetometer to remove environmental interference from a magnetic field reading. We evaluated this sensing method to show its performance when removing the interference in three-movement dimensions, in six environments, and with six different cycling rates. Then, we localized the electromagnet with respect to the magnetic field reader in any direction within a 13.8 cm range with a relative error of 2.3% for the distance and an average error of 3.43° for the orientation angle. We applied Magneto in a pilot study: calculating elbow flexion angles. In this study, we calculated elbow flexion angles to the nearest 15° with 93.8% accuracy.

In typical scenarios for indoor localization and tracking, it is essential to accurately track the pedestrians when they are crossing the connections of different spaces. In this paper, we propose a magnet barcode scanning-based solution for indoor pedestrian tracking. We assemble multiple magnet bars into magnet arrays as a unique magnet barcode, and deploy different magnet barcodes at different connections to label them. We embed an inertial measurement unit (IMU) into the pedestrian`s shoes. When the pedestrian crosses these connections, the magnetometer from the IMU scans the magnet barcode and recognize its corresponding ID. In this way, indoor pedestrian tracking can be regarded as a process of continuously scanning different magnet barcodes. By performing correlation analysis on these barcodes, the trace of pedestrian can be effectively depicted in the indoor map. To build a unique magnet barcode based on the magnet bar arrays, we provide an optimized structure for building the magnet barcode. To tackle the diversities of the pedestrian's gait traces in identifying the magnet barcode, we provide a generalized model based on the space axis for magnet barcode identification. As far as we know, this is the first work to use the magnet bar array to construct the magnet barcode for indoor pedestrian tracking. The real experiment results show that our system can achieve an average accuracy of 88.9% in identifying the magnet barcodes and an average accuracy of 93.1 % for indoor pedestrian tracking.

In recent years, Channel State Information (CSI) measured by WiFi is widely used for human activity recognition. In this article, we propose a deep learning design for location- and person-independent activity recognition with WiFi. The proposed design consists of three Deep Neural Networks (DNNs): a 2D Convolutional Neural Network (CNN) as the recognition algorithm, a 1D CNN as the state machine, and a reinforcement learning agent for neural architecture search. The recognition algorithm learns location- and person-independent features from different perspectives of CSI data. The state machine learns temporal dependency information from history classification results. The reinforcement learning agent optimizes the neural architecture of the recognition algorithm using a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM). The proposed design is evaluated in a lab environment with different WiFi device locations, antenna orientations, sitting/standing/walking locations/orientations, and multiple persons. The proposed design has 97% average accuracy when testing devices and persons are not seen during training. The proposed design is also evaluated by two public datasets with accuracy of 80% and 83%. The proposed design needs very little human efforts for ground truth labeling, feature engineering, signal processing, and tuning of learning parameters and hyperparameters.

The quality of sleep is an important factor in maintaining a healthy life style. A great deal of work has been done for designing sleep monitoring systems. However, most of existing solutions bring invasion to users more or less due to the exploration of the accelerometer sensor inside the device. This article presents iSleep—a practical system to monitor people’s sleep quality using off-the-shelf smartphone. iSleep uses the built-in microphone of the smartphone to detect the events that are closely related to sleep quality, and infers quantitative measures of sleep quality. iSleep adopts a lightweight decision-tree-based algorithm to classify various events. For two-user scenario, iSleep differentiates the events of two users either when two phones can collaborate with each other or when two phones cannot communicate with each other. The experimental results show that iSleep achieves consistently above 90% accuracy for event classification in a variety of different settings in one-user scenario and above 92% accuracy for distinguishing users in two-user scenario. By providing a fine-grained sleep profile that depicts details of sleep-related events, iSleep allows the user to track the sleep efficiency over time and relate irregular sleep patterns to possible causes.

Continuous authentication monitors the security of a system throughout the login session on mobile devices. In this article, we present SCANet, a two-stream convolutional neural network--based continuous authentication system that leverages the accelerometer and gyroscope on smartphones to monitor users’ behavioral patterns. We are among the first to use two streams of data—frequency domain data and temporal difference domain data—from the two sensors as the inputs of the convolutional neural network (CNN). SCANet utilizes the two-stream CNN to learn and extract representative features and then performs the principal component analysis to select the top 25 features with high discriminability. With the CNN-extracted features, SCANet exploits the one-class support vector machine to train the classifier in the enrollment phase. Based on the trained CNN and classifier, SCANet identifies the current user as a legitimate user or an impostor in the continuous authentication phase. We evaluate the effectiveness of the two-stream CNN and the performance of SCANet on our dataset and BrainRun dataset, and the experimental results demonstrate that CNN achieves 90.04% accuracy, and SCANet reaches an average of 5.14% equal error rate on two datasets and takes approximately 3 s for user authentication.

Aggressive tennis shots with high ball speed are the key factor in winning a tennis match. Today's tennis players are increasingly focused on improving ball speed. As a result, in recent tennis tournaments, records of tennis shot speeds are broken again and again. The traditional method for calculating the tennis ball speed uses multiple high-speed cameras and computer vision technology. This method is very expensive and hard to set up. Another way to calculate the tennis ball speed is to use motion sensors, which are lower cost and easier to set up. In this paper, we propose an approach for tennis ball speed estimation based on a racket-mounted motion sensor. We divide the tennis strokes into three categories: serve, groundstroke, and volley. For a serve, a regression model is proposed to estimate the ball speed. For a groundstroke or volley, two models are proposed: a regression model and a physical model. We use the physical model to estimate the ball speed for advanced players and the regression model for beginner players. Under the leave-one-subject-out cross-validation test, evaluation results show that TennisEye is 10.8% more accurate than the state-of-the-art work.

As personal computing platforms, smartphones are commonly used to store private, sensitive, and security information, such as photographs, emails, and Android Pay. To protect such information from adversaries, continuous authentication on smartphone users becomes more and more important. In this paper, we present a novel authentication system, SensorAuth, for continuous authentication of users based on their behavioral patterns, by leveraging the accelerometer and gyroscope ubiquitously built into smartphones. We are among the first to exploit five data augmentation approaches including permutation, sampling, scaling, cropping, and jittering to create additional data by applying them on training data. With the augmented data, SensorAuth extracts sensor-based features in both time and frequency domains within a time window, then utilizes the one-class support vector machine to train the classifier, and finally authenticates users. We evaluate the authentication performance of SensorAuth in terms of the impact of window size, accuracy on each of and combinations of data augmentation approaches, time efficiency, energy consumption, and comparisons with the representative classifiers and with the existing approaches, respectively. The experimental results show that SensorAuth performs highly accurate and time-efficient continuous authentication, by reaching the lowest median equal error rate of 4.66%, and consuming a short authentication time of approximately 5 s.

Garbled circuits are general tools that allow two parties to compute any function without disclosing their respective inputs. Applications of this technique vary from distributed privacy-preserving machine learning tasks to secure outsourced authentication. Unfortunately, the energy cost of garbled circuit evaluation protocols is substantial. This limits the applicability of garbled circuits in scenarios that involve battery-operated devices, such as Internet-of-Things (IoT) devices and smartphones. In this paper, we propose MEG, a Memory- and Energy-efficient Garbled circuit evaluation mechanism. MEG utilizes batch data transmission and multi-threading to reduce memory and energy consumption. We implement MEG on an Android smartphone and compare its performance and energy consumption with state-of-the-art techniques using two garbled circuits of widely different sizes (AES-128 and 256-bit edit distance). Our results show that, compared with “plain” garbled circuit evaluation, MEG decreases memory consumption by more than 90%. When compared with current pipelined garbled circuit evaluation techniques, MEG's energy usage was 42% lower for AES-128 and 23% lower for EDT-256. Furthermore, our multi-thread implementation of MEG decreased circuit evaluation time by up to 56.7% for AES-128, and by up to 13.5% for EDT-256, compared with state-of-the-art pipelining techniques.

We present a two-stream convolutional neural network based authentication system, CNNAuth, for continuously monitoring users' behavioral patterns, by leveraging the accelerometer and gyroscope on smartphones. We are among the first to exploit two streams of the time-domain data and frequency-domain data from raw sensor data for learning and extracting universal effective and efficient feature representations as the inputs of the convolutional neural network (CNN), and the extracted features are further selected by the principal component analysis (PCA). With these features, we use the one-class support vector machine (SVM) to train the classifier in the enrollment phase, and with the trained classifier and testing features, CNNAuth classifies the current user as a legitimate user or an impostor in the continuous authentication phase. We evaluate the performance of the two-stream CNN and CNNAuth, respectively, and the experimental results show that the two-stream CNN achieves an accuracy of 87.14%, and CNNAuth reaches the lowest authentication EER of 2.3% and consumes approximately 3 seconds for authentication.

People watching smartphones while walking causes a significant impact to their safety. Pedestrians staring at smartphone screens while walking along the sidewalk are generally more at risk than other pedestrians not engaged in smartphone usage. In this study, the authors propose Safe Walking , an Android smartphone‐based system that detects the walking behaviour of pedestrians by leveraging the sensors and front camera on smartphones, improving the safety of pedestrians staring at smartphone screens. More specifically, Safe Walking first exploits a pedestrian speed calculation algorithm by sampling acceleration data via the accelerometer and calculating gravity components via the gravity sensor. Then, this system utilises a greyscale image detection algorithm to detect the face and eye movement modes based on OpenCV4Android to determine if pedestrians are staring at the screens. Finally, Safe Walking generates a vibration by a vibrator on smartphones to alert pedestrians to pay attention to road conditions. The authors implemented Safe Walking on an Android smartphone and evaluated pedestrian walking speed, the accuracy of eye movement, and system performance. The results show that Safe Walking can prevent the potential danger for pedestrians staring at smartphone screens with a true positive rate of .

MIMO beamforming provides high throughput for WiFi networks, but it also leads to high computation and communication overhead due to Channel State Information (CSI) feedback. Explicit CSI feedback provides high beamforming gains, but it introduces extremely high overhead. Implicit CSI feedback has low overhead, but it provides very low beamforming gains. We propose EliMO to completely Eliminate CSI feedback from MIMO without sacrificing beamforming gains. EliMO uses two-way channel estimation to allow WiFi Access Points (AP) to accurately estimate downlink CSI without explicit CSI feedback. To measure downlink CSI at the WiFi AP, the WiFi station (STA) puts the received signal of downlink training symbols into Feedback Training Field (FTF) and sends it back to the AP. The AP estimates the two-way channel using the received signal of FTF. Analysis and experiment results show that EliMO is able to provide as high beamforming gains as explicit CSI feedback and as low overhead as implicit CSI feedback. EliMO significantly reduces computation and communication costs of measuring and sending CSI feedback for smart devices, like smartphones, smartwatches, and wireless drones. We evaluate the throughput and energy consumption of EliMO by experiment measurements in both static and mobile scenarios. Evaluation results show that EliMO provides 5× and 4× throughput as implicit and explicit CSI feedback, respectively. Energy consumption of EliMO is only 85%/30% of that of implicit/explicit CSI feedback.

Multiple-input multiple-output (MIMO) provides high throughput for WiFi networks, but it also leads to high overhead due to channel state information (CSI) feedback. Based on experiment measurements, this paper shows that MIMO has different feedback requirements when the receiver is rotating compared with when the receiver is in other mobility scenarios. Experiments of four popular Android games show that device rotation accounts for around 50% of the running time for these games, which implies that rotation-awareness could improve WiFi efficiency significantly for these games. We propose rotation-aware WiFi (RoFi) channel feedback to eliminate unnecessary CSI feedback while maintaining high throughput. We show the failure of existing mobility-aware methods, including CSI similarity, time-of-flight (ToF), and compression noise, in distinguishing the mobility status of rotation and mobile. RoFi calculates power delay profile (PDP) similarity for rotation detection and performs feedback compression and rate selection accordingly. To deal with false rotation detection and status transition between rotation and static, RoFi uses the power of the strongest path, which is calculated from PDP, to further refine CSI feedback when necessary. The RoFi design is compatible with legacy 802.11 protocols and is easy to be deployed on existing WiFi systems. Evaluation results show that RoFi reduces 25%-40% overhead with negligible signal-to-noise ratio decrease in rotation scenarios. RoFi also consumes 29%-69% less energy compared with state-of-the-art feedback compression and rate selection algorithms.

Wearable glasses are on the rising edge of development with great user popularity. However, user data stored on these devices bring privacy risks to the owner. To better protect the owner's privacy, a continuous authentication system is needed. In this paper, we propose a continuous and noninvasive authentication system for wearable glasses, named GlassGuard. GlassGuard discriminates the owner and an impostor with behavioral biometrics from six types of touch gestures (single-tap, swipe forward, swipe backward, swipe down, two-finger swipe forward, and two-finger swipe backward) and voice commands, which are all available during normal user interactions. With data collected from 32 users on Google Glass, we show that GlassGuard achieves 99% detection rate and 0.5% false alarm rate after 3.5 user events on average when all types of user events are available with equal probability. Under five typical usage scenarios, the system has a detection rate above 93% and a false alarm rate below 3% after less than five user events.

In this paper, we show that public USB charging stations pose a significant privacy risk to smartphone users even when no data communication is possible between the station and the user's mobile device. We present a side-channel attack that allows a charging station to identify which Webpages are loaded while the smartphone is charging. To evaluate this side-channel, we collected power traces of Alexa top 50 Websites on multiple smartphones under several conditions, including battery charging level, browser cache enabled/disabled, taps on the screen, Wi-Fi/LTE, TLS encryption enabled/disabled, time elapsed between collection of training and testing data, and location of the Website. The results of our evaluation show that the attack is highly successful: in many settings, we were able to achieve over 90% Webpage identification accuracy. On the other hand, our experiments also show that this side-channel is sensitive to some of the aforementioned conditions. For instance, when training and testing traces were collected 70 days apart, accuracies were as low as 2.2%. Although there are studies that show that power-based side-channels can predict browsing activity on laptops, this paper is unique, because it is the first to study this side-channel on smartphones, under smartphone specific constraints. Further, we demonstrate that Websites can be correctly identified within a short time span of 2 × 6 seconds, which is in contrast with prior work, which uses 15-s traces. This is important, because users typically spend less than 15 s on a Webpage.

WiFi is a major source of energy consumption on smartphones. Unfortunately, a non-negligible portion of the WiFi energy consumption is spent for frames that are useless to the smartphone. For example, energy is wasted to receive WiFi broadcast frames that are not needed by any smartphone application. What's worse, in order to process the broadcast frames received, a smartphone in suspend mode switches from suspend mode to high power active mode and stays there for a while. As such, additional energy is wasted to do the processing. In this paper, we design a system, namely HIDE, to reduce smartphone energy wasted on useless WiFi broadcast traffic. With our system, smartphones in suspend mode do not receive useless broadcast frames or wake up to process useless broadcast frames. Our trace-driven simulation shows that the HIDE system saves 34%-75% energy for Nexus One when 10% of the broadcast frames are useful to the smartphone. Our overhead analysis demonstrates that our system has negligible impact on network capacity and packet round-trip time.

Our study reveals that facilitating warm launch of just five smartphone applications is extremely expensive, using up to 36 percent of memory. Further investigation of 20 popular applications indicates that rich multimedia applications have high heap usage and go above allowed boundaries, up to 5.63 times more heap than guaranteed by the system, and may cause crashes and erroneous behaviors. Therefore, we present iRAM, a personalized system that maintains optimal heap size limits to avoid crashes, efficiently maximizes free memory levels, and cleans low-priority processes to reduce application delays. The evaluation on memory hungry applications indicates that iRAM reduces application crashes by up to 14 percent, and reduces launch delays by up to 78.2 percent. In addition, the results confirm that iRAM increases free memory levels by up to 4.8 times. This performance gain comes with 3.5 percent of CPU overhead and 0.9 percent of power overhead.

As wearable devices become ubiquitous, there will be an increased demand for platforms that allow engineers and researchers to quickly prototype and evaluate new wearable devices. However, many of these platforms require that the hardware be configured before the code is written, thereby limiting the programmer to the limitations of the hardware. In this paper, we present a platform that allows researchers and engineers to quickly prototype new wearable devices using a code-first approach. This approach allows software developers to create new prototypes by first writing the code that the prototype is required to run. Once the code has been written, the hardware that is required to run the application can be generated by analyzing the code that the software developer has specified. This code-first approach is possible because of the system's architecture which is comprised of both a hardware and software component. The hardware component consists of a main board with four expansion ports, while the software platform is a modular middleware which consists of a collection of stateless libraries that abstract each hardware module. These modular abstractions allow us to synthesize the hardware configuration from the software definition. We evaluated our design using it to prototype three wearable devices: 1) an environmental exposure monitoring smartwatch; 2) an infrared indoor localization system; and 3) a step counter.

The ability to attach external devices to smartphones has revolutionized the role of smartphones by extending their capabilities beyond the limitations of commodity hardware. Developing external attachments that allow smartphones to sense the depth of an area will facilitate the development of new immersive applications and technologies. In this paper, we propose a smartphone compatible SONAR ranging attachment and address the compatibility problem by proposing a hybrid hardware/software modulator that allows a digital sensor to communicate with a smartphone via the 3.5-mm headphone jack, found on most smartphones. We evaluate the proposed sensor using two metrics, accuracy and spatial resolution. We evaluate the accuracy of this system by measuring known distances with the sensor and comparing them. We measure the sensor's spatial resolution by using ranging information from the SONAR module along with the phone's gyroscope, accelerometer, and magnetometer to construct a two-dimensional map of a space.

The importance of random number generators (RNGs) to various computing applications is well understood. To ensure a quality level of output, high-entropy sources should be utilized as input. However, the algorithms used have not yet fully evolved to utilize newer technology. Even the Android pseudo RNG (APRNG) merely builds atop the Linux RNG to produce random numbers. This paper presents an exploratory study into methods of generating random numbers on sensor-equipped mobile and Internet of Things devices. We first perform a data collection study across 37 Android devices to determine two things-how much random data is consumed by modern devices, and which sensors are capable of producing sufficiently random data. We use the results of our analysis to create an experimental framework called SensoRNG, which serves as a prototype to test the efficacy of a sensor-based RNG. SensoRNG employs collection of data from on-board sensors and combines them via a lightweight mixing algorithm to produce random numbers. We evaluate SensoRNG with the National Institute of Standards and Technology statistical testing suite and demonstrate that a sensor-based RNG can provide high quality random numbers with only little additional overhead.

Common smartphone authentication mechanisms (e.g., PINs, graphical passwords, and fingerprint scans) are not designed to offer security post-login. Multi-modal continuous authentication addresses this issue by frequently and unobtrusively authenticating the user via behavioral biometric signals, such as touchscreen interaction and hand movements. Because smartphones can easily fall into the hands of the adversary, it is critical that the behavioral biometric information collected and processed on these devices is secured. This can be done by offloading encrypted template information to a remote server, and then performing authentication via privacy-preserving protocols. In this paper, we demonstrate that the energy overhead of current privacy-preserving protocols for continuous authentication is unsustainable on smartphones. To reduce energy consumption, we design a technique that leverages characteristics unique to the authentication setting in order to securely outsource computation to an untrusted Cloud. Our approach is secure against a colluding smartphone and Cloud, thus making it well suited for authentication. We performed extensive experimental evaluation. With our technique, the energy requirement for running an authentication instance that computes Manhattan distance is 0.2 mWh, which corresponds to a negligible fraction of the smartphone's battery capacity. In addition, for Manhattan distance, our protocol runs in 0.72 and 2 s for 8 and 28 biometric features, respectively. We were also able to compute Hamming distance in 3.29 s, compared with 95.57 s achieved with the previous fastest outsourced computation protocol (Whitewash). These results demonstrate that ours is presently the only technique suitable for low-latency continuous authentication (e.g., with authentication scan windows of 60 s or shorter).

We introduce hand movement, orientation, and grasp (HMOG), a set of behavioral features to continuously authenticate smartphone users. HMOG features unobtrusively capture subtle micro-movement and orientation dynamics resulting from how a user grasps, holds, and taps on the smartphone. We evaluated authentication and biometric key generation (BKG) performance of HMOG features on data collected from 100 subjects typing on a virtual keyboard. Data were collected under two conditions: 1) sitting and 2) walking. We achieved authentication equal error rates (EERs) as low as 7.16% (walking) and 10.05% (sitting) when we combined HMOG, tap, and keystroke features. We performed experiments to investigate why HMOG features perform well during walking. Our results suggest that this is due to the ability of HMOG features to capture distinctive body movements caused by walking, in addition to the hand-movement dynamics from taps. With BKG, we achieved the EERs of 15.1% using HMOG combined with taps. In comparison, BKG using tap, key hold, and swipe features had EERs between 25.7% and 34.2%. We also analyzed the energy consumption of HMOG feature extraction and computation. Our analysis shows that HMOG features extracted at a 16-Hz sensor sampling rate incurred a minor overhead of 7.9% without sacrificing authentication accuracy. Two points distinguish our work from current literature: 1) we present the results of a comprehensive evaluation of three types of features (HMOG, keystroke, and tap) and their combinations under the same experimental conditions and 2) we analyze the features from three perspectives (authentication, BKG, and energy consumption on smartphones).

Mobile video chat apps offer users an approachable way to communicate with others. As high-speed 4G networks are being deployed worldwide, the number of mobile video chat app users increases. However, video chatting on mobile devices brings users financial concerns, since streaming video demands high bandwidth and can use up a large amount of data in dozens of minutes. Lowering the bandwidth usage of mobile video chats is challenging since video quality may be compromised. In this paper, we attempt to tame this challenge. Technically, we propose a context-aware frame rate adaption framework, named low-bandwidth video chat (LBVC). It follows a sender-receiver cooperative principle that smartly handles the tradeoff between lowering bandwidth usage and maintaining video quality. We implement LBVC by modifying an open-source app–Linphone– and evaluate it with both objective experiments and subjective studies.

Extensive empirical studies presented in this article confirm that the quality of radio communication between low-power sensor devices varies significantly with time and environment. This phenomenon indicates that the previous topology control solutions, which use static transmission power, transmission range, and link quality, might not be effective in the physical world. To address this issue, online transmission power control that adapts to external changes is necessary. This article presents ATPC, a lightweight algorithm for Adaptive Transmission Power Control in wireless sensor networks. In ATPC, each node builds a model for each of its neighbors, describing the correlation between transmission power and link quality. With this model, we employ a feedback-based transmission power control algorithm to dynamically maintain individual link quality over time. The intellectual contribution of this work lies in a novel pairwise transmission power control, which is significantly different from existing node-level or network-level power control methods. Also different from most existing simulation work, the ATPC design is guided by extensive field experiments of link quality dynamics at various locations over a long period of time. The results from the real-world experiments demonstrate that (1) with pairwise adjustment, ATPC achieves more energy savings with a finer tuning capability, and (2) with online control, ATPC is robust even with environmental changes over time.

Existing research has demonstrated that inter-body sensor network (inter-BSN) information sharing among coexisting BSNs can enhance applications' performance and save energy. However, how to achieve effective inter-BSN information sharing through wireless communication is a challenging task. On one hand, a BSN should be able to discover neighboring BSNs and establish inter-BSN links with quality of service (QoS) assurances. On the other hand, a BSN should be able to prevent the QoS of intraand inter-BSN links from being degraded by multiple BSNs' mutual interference. In this paper, we propose BuddyQoS, a framework that provides network throughput assurances for coexisting and shared buddy BSNs. In particular, BuddyQoS accurately estimates and adaptively schedules wireless resources to meet the throughput requirements of all interand intra-BSN links. Our trace-driven experiment results demonstrate that BuddyQoS outperforms the default CSMA solution in the standard TinyOS-2.x releases in terms of providing throughput assurances.

Frequent route is an important individual outdoor behavior pattern that many trajectory-based applications rely on. In this paper, we propose a novel framework for extracting frequent routes from personal GPS trajectories. The key idea of our design is to accurately detect road corners and utilize these new metaphors to tackle the problem of frequent route extraction. Concretely, our framework contains three phases: 1) characteristic point (CP) extraction; 2) corner detection; and 3) trajectory mapping. In the first phase, we present a linear fitting-based algorithm to extract CPs. In the second phase, we develop a multiple density level DBSCAN (density-based spatial clustering of applications with noise) algorithm to locate road corners by clustering CPs. In the third phase, we convert each trajectory into an ordered sequence of road corners and obtain all routes that have been traversed by an individual for at least ${F}$ (frequency threshold) times. We evaluate the framework using real-world trajectory datasets of individuals for one year and the experimental results demonstrate that our framework outperforms the baseline approach by 7.8% on average in terms of precision and 21.9% in terms of recall.

Data transmission over WiFi quickly drains the batteries of mobile devices. Although the IEEE 802.11 standards provide power save mode (PSM) to help mobile devices conserve energy, PSM fails to bring expected benefits in many real scenarios. In particular, when multiple PSM mobile devices associate to a single access point (AP), PSM does not work well under transmission contention. Optimizing power saving of multiple PSM clients is a challenging task, because each PSM client expects to complete data transmission early so that it can turn to low power mode. In this paper, we define an energy conserving model to describe the general PSM traffic contention problem. We prove that the optimization of energy saving for multiple PSM clients under constraint is an NPcomplete problem. Following this direction, we propose a solution called harmonious power saving mechanism (HPSM) to address one specific case, in which multiple PSM clients associate to a single AP. In HPSM, we first use a basic sociological concept to define the richness of a PSM client based on the link resource it consumes. Then, we separate these poor PSM clients from rich PSM clients in terms of link resource consumption and favor the former to save power when they face PSM transmission contention. We implement prototypes of HPSM based on the open source projects Mad-wifi and NS-2. Our evaluations show that HPSM can help the poor PSM clients effectively save power while only slightly degrading the rich PSM clients' performance in comparison with the existing PSM solutions.

As one of the most popular exercises, running is accomplished through a tight cooperation between the respiratory and locomotor systems. Research has suggested that a proper running rhythm -- the coordination between breathing and strides -- helps improve exercise efficiency and postpone fatigue. This paper presents RunBuddy -- the first smartphone-based system for continuous running rhythm monitoring. RunBuddy is designed to be a convenient and unobtrusive exercise feedback system, and only utilizes commodity devices including smartphone and Bluetooth headset. A key challenge in designing RunBuddy is that the sound of breathing typically has very low intensity and is susceptible to interference. To reliably measure running rhythm, we propose a novel approach that integrates ambient sensing based on accelerometer and microphone, and a physiological model called Locomotor Respiratory Coupling (LRC), which indicates possible ratios between the stride and breathing frequencies. We evaluate RunBuddy through experiments involving 13 subjects and 39 runs. Our results show that, by leveraging the LRC model, RunBuddy correctly measures the running rhythm for indoor/outdoor running 92:7% of the time. Moreover, RunBuddy also provides detailed physiological profile of running that can help users better understand their running process and improve exercise self-efficacy.

Smartphones save energy by entering a low power suspend mode (120mW). As a result, power consumption increases dramatically. To improve energy efficiency, some phones employ a hardware broadcast filter in the WiFi driver. All UDP broadcast frames other than Multicast DNS frames are blocked, thus none is received by upper layer applications. So, we have a dilemma of handling WiFi broadcast traffic during smartphone suspend mode: either receive all of them suffering high power consumption, or receive none of them sacrificing functionalities. In this paper, we propose Software Broadcast Filter (SBF) to address the dilemma. SBF is smarter than the hardware broadcast filter as it only filters out useless broadcast frames and does not impair functionalities of applications. SBF is also more energy efficient than the “receive all” method. Our trace driven evaluation shows that SBF can save up to 52% energy consumption than the “receive all” method.

Video chat apps enable users to stay in touch with their family, friends and colleagues. However, they consume a lot of bandwidth and hence can quickly use up a monthly data plan quota, which is a high-cost resource on smartphones. In this paper, we propose LBVC (Low-bandwidth Video Chat), a user-guided vibration-aware frame rate adaption framework. LBVC takes a sender-receiver cooperative approach and reduces bandwidth usage as well as alleviates video quality degradation for video chat apps on smartphones. We implement LBVC on the Android platform and evaluate its performance on a series of experiments and user study with 21 pairs of subjects. Compared to the default solution, LBVC decreases bandwidth usage by 35% and at the same time maintains good video quality without introducing extra power consumption under typical video chat scenarios.

The smartphone has become an important part of our daily lives. However, the user experience is still far from being optimal. In particular, despite the rapid hardware upgrades, current smartphones often suffer various unpredictable delays during operation, e.g., when launching an app, leading to poor user experience. In this paper, we investigate the behavior of reads and writes in smartphones. We conduct the first large-scale measurement study on the Android I/O delay using the data collected from our Android application running on 2611 devices within nine months. Among other factors, we observe that reads experience up to 626% slowdown when blocked by concurrent writes for certain workloads. Additionally, we show the asymmetry of the slowdown of one I/O type due to another, and elaborate the speedup of concurrent I/Os over serial ones. We use this obtained knowledge to design and implement a system prototype called SmartIO that reduces the application delay by prioritizing reads over writes, and grouping them based on assigned priorities. SmartIO issues I/Os with optimized concurrency parameters. The system is implemented on the Android platform and evaluated extensively on several groups of popular applications. The results show that our system reduces launch delays by up to 37.8%, and run-time delays by up to 29.6%.

No abstract available.

We live in a 3-D world. However, the smart phones that we use every day are incapable of sensing depth, without the use of custom hardware. By creating new depth sensors, we can provide developers with the tools that they need to create immersive mobile applications that take advantage of the 3-D nature of our world. In this paper, we propose a new sonar sensor for smart phones. This sonar sensor does not require any additional hardware, and utilizes the phone's microphone and rear speaker. The sonar sensor calculates distances by measuring the elapsed time between the initial pulse and its reflection. We evaluate the accuracy of the sonar sensor by using it to measure the distance from the phone to an object. We found that we were able to measure the distances of objects accurately with an error bound of 12 cm.

The articles in this special section are focused on two major aspects of Internet of things (IoT) technologies for smart and connected health services (SCH): 1) monitoring and assisting individuals by means of smart systems including sensors, devices, and robotics; and 2) creating interoperable digital health information infrastructures to increase medical/health information availability and use. The papers published in this SI provide evidence that SCH tools that rely upon IoT technologiescould significantly improve clinical outcomes and thequality of life of individuals undergoing monitoring.

Many applications in wireless sensor networks require communication performance that is both consistent and of high quality. Unfortunately, performance of current network protocols can vary significantly because of various interferences and environmental changes. Current protocols estimate link quality based on the reception of probe packets over a short time period. This method is neither efficient nor accurate enough to capture the dramatic variations of link quality. Therefore, we propose a link metric called competence that characterizes links over a longer period of time. We combine competence with current short-term estimations in routing algorithm designs. To further improve network performance, we have designed a distributed route maintenance framework based on feedback control solutions. This framework allows every link along an end-to-end (E2E) path to adjust its link protocol parameters, such as transmission power and number of retransmissions, to ensure specified E2E reliability and latency under dynamic link qualities. Our solutions are evaluated in both extensive simulations and real system experiments. In real system evaluations with 48 T-Motes, our overall solution improves E2E packet delivery ratio over existing solutions by up to 40% while reducing transmission energy consumption by up to 22%. Importantly, our solution also achieves more stable and better transient performance than current approaches.

WiFi effectively has two extremes: low power consumption and high latency, or low latency and high power consumption. WiFi Power Save Mode saves energy by trading added latency for less power consumption. Minimal latency but maximum power, on the other hand, is consumed with WiFi Active Mode. While research has advanced in mitigating these extremes, certain types of network traffic such as constant bitrate streaming make the contrast unavoidable. We introduce Bluesaver, which provides low latency and low energy by maintaining a Bluetooth and WiFi connection simultaneously. Bluesaver is designed at the MAC layer and is able to opportunistically select the most efficient connection for packets while still assuring acceptable latency. We implement Bluesaver on an Android phone and Access Point and show that we can save more than 25% energy over existing solutions and attain the capability of quickly adapting to changes in network traffic.

The increasing availability of inexpensive off-the-shelf 802.11 hardware has made it possible to deploy access points (APs) densely to ensure the coverage of complex enterprise environments such as business and college campuses. However, dense AP deployment often leads to increased level of wireless contention, resulting in low system throughput. A promising approach to address this issue is to enable the transmission concurrency of exposed terminals in which two senders lie in the range of one another but do not interfere each other's receiver. However, existing solutions ignore the rate diversity of 802.11 and hence cannot fully exploit concurrent transmission opportunities in a WLAN. In this paper, we presentTRACK-TransmissionRateAdaptation forColliding linKs, a novel protocol for harnessing exposed terminals with a rate adaptation approach in enterprise WLANs. Using measurement-based channel models, TRACK can optimize the bit rates of concurrent links to improve system throughput while maintaining link fairness. Our extensive experiments on a testbed of 17 nodes show that TRACK improves system throughput by up to 67% and 35% over 802.11 CSMA and conventional approaches of harnessing exposed terminals.

Body area networks are used extensively in the medical field and elderly care. These networks perform a collection of roles including monitoring an individual's activities, through a process known as activity recognition. Current approaches to activity recognition require specialized hardware for advanced sensors which puts stress on battery life. We show that it is theoretically possible to distinguish between different human activities/postures by using radio signal propagation only and provide strategies for doing this. This is immensely beneficial to the field of sensor networks for two reasons: 1) It removes the need for more energy intensive components thus reducing strain on limited power resources and thus 2) reduces form factor for sensor network nodes. We show that activity recognition can be done using only radio signals at low transmit power levels with good accuracy. Lower power consumption and reduced form factor are desirable features for body networks and have been identified as being essential for wide adoption of body networks.

No abstract available.

The smartphone has become an important part of our daily lives. However, the user experience is still far from being optimal. In particular, despite the rapid hardware upgrades, current smartphones often suffer various unpredictable delays during operation, e.g., when launching an application, leading to poor user experience. This video features our study of storage impact on smartphone application delay. We conduct the first large-scale measurement study on the I/O delay of Android using the data collected from our application running on 1009 devices within 130 days. We observe that Android devices spend up to 58% of their CPU active time waiting for storage I/Os to complete. This negatively affects the smartphone's overall application performance, and results in slow response time. Further investigation, among others, reveals that reads experience up to a 626% slowdown in the presence of concurrent writes. The obtained knowledge is used to design and implement a system called SmartIO that reduces the application delay by prioritizing reads over writes, and grouping them based on assigned priorities. SmartIO is implemented on the Android platform and evaluated extensively on several groups of popular applications. The results from the 20 researched applications demonstrate that SmartIO reduces launch delays by up to 37.8%, and run-time delays by up to 29.6%.

Continuous authentication modalities allow a device to authenticate users transparently without interrupting them or requiring their attention. This is especially important on smartphones, which are more prone to be lost or stolen than regular computers, and carry plenty of sensitive information. There is a multitude of signals that can be harnessed for continuous authentication on mobile devices, such as touch input, accelerometer, and gyroscope, etc. However, existing public datasets include only a handful of them, limiting the ability to do experiments that involve multiple modalities. To fill this gap, we performed a large-scale user study to collect a wide spectrum of signals on smartphones. Our dataset combines more modalities than existing datasets, including movement, orientation, touch, gestures, and pausality. This dataset has been used to evaluate our new behavioral modality named Hand Movement, Orientation, and Grasp (H-MOG). This poster reports on the data collection process and outcomes, as well as preliminary authentication results.

With newly developed wireless neuroheadsets, electroencephalography (EEG) neurometrics can be incorporated into in situ and ubiquitous physiological monitoring for human mental health. As a resource constraint system providing critical health services, the EEG headset design must consider both high application fidelity and energy efficiency. However, through empirical studies with an off-the-shelf Emotiv EPOC Neuroheadset, we uncover a mismatch between lossy EEG sensor communication and high neurometric application fidelity requirements. To tackle this problem, we study how to learn the sensitivity of neurometric application fidelity to EEG data. The learned sensitivity is used to develop two algorithms: 1) an energy minimization algorithm minimizing the energy usage in EEG sampling and networking while meeting applications' fidelity requirements and 2) a fidelity maximization algorithm maximizing the sum of all applications' fidelities through the incorporation and optimal utilization of a limited data buffer. The effectiveness of our proposed solutions is validated through trace-driven experiments.

The primary goal of breadcrumb trail sensor networks is to transmit in real-time users' physiological parameters that measure life-critical functions to an incident commander through reliable multihop communication. In applications using breadcrumb solutions, there are often many users working together, and this creates a well-known body shadowing effect (BSE). In this article, we first measure the characteristics of body shadowing for 2.4GHz sensor nodes. Our empirical results show that the body shadowing effect leads to severe packet loss and consequently very poor real-time performance. Then we develop a novel Intentional Forwarding solution. This solution accurately detects the shadowing mode and enables selected neighbors to forward data packets. Experimental results from a fully implemented testbed demonstrate that Intentional Forwarding is able to improve the end-to-end average packet delivery ratio (PDR) from 58% to 93% and worst-case PDR from 45% to 85%, and is able to meet soft real-time requirements even under severe body shadowing problems.

Wireless sensor network applications, such as those for natural disaster warning, vehicular traffic monitoring, and surveillance, have stringent accuracy requirements for detecting or classifying events and demand long system lifetimes. Through quantitative study, we show that existing event detection approaches are challenged to explore the sensing capability of a deployed system and choose the right sensors to meet user-specified accuracy. Event detection systems are also challenged to provide a generic system that efficiently adapts to environmental dynamics and works easily with a range of applications, machine learning approaches, and sensor modalities. Consequently, we propose Watchdog, a modality-agnostic event detection framework that clusters the right sensors to meet user-specified detection accuracy during runtime while significantly reducing energy consumption. Watchdog can use different machine learning techniques to learn the sensing capability of a heterogeneous sensor deployment and meet accuracy requirements. To address environmental dynamics and ensure energy savings, Watchdog wakes up and puts to sleep sensors as needed to meet user-specified accuracy. Through evaluation with real vehicle detection trace data and a building traffic monitoring testbed of IRIS motes, we demonstrate the superior performance of Watchdog over existing solutions in terms of meeting user-specified detection accuracy, energy savings, and environmental adaptability.

In this paper, to our best knowledge, we are first to provide an experimental study on how storage techniques affect power levels in smartphones and introduce energy-efficient approaches to reduce energy consumption. We evaluate power degradation at several layers of block I/O, focusing on the block layer and device driver. At each level, we investigate the amount of energy that can be saved, and use that to design and implement a prototype with optimal energy savings named SmartStorage. The system tracks the run-time I/O pattern of a smartphone that is then matched with the closest pattern from the benchmark table. After having obtained the optimal parameters, it dynamically configures storage parameters to reduce energy consumption. We evaluate our prototype by using the 20 most popular Android applications, and our energy-efficient approaches achieve from 23% to 52% of energy savings compared to using the current techniques.

In many body sensor network (BSN) applications, such as activity recognition for assisted living residents or physical fitness assessment of a sports team, users spend a significant amount of time with one another while performing many of the same activities. We exploit this physical proximity with Remora, a smartphone-based Body Sensor Network activity recognition system which shares sensing resources among neighboring BSNs. Compared to other resource sharing approaches, Remora provides both increased accuracy and significant energy savings. To increase classification accuracy, Remora BSNs share sensors by overhearing neighbors' sensor data transmissions. When sharing, fewer on-body sensors are needed to achieve high accuracy, resulting in energy savings by turning off unneeded sensors. To save phone energy, neighboring BSNs share classifiers: only one classifier is active at a time classifying activities for all neighbors. Remora addresses three major challenges of sharing with physical neighbors: 1) Sharing only when the energy benefit outweighs the cost, 2) Finding and utilizing the shared sensors and classifiers which produce the best combination of accuracy improvement and energy savings, and 3) Providing a lightweight and collaborative classification approach, without the use of a backend server, which adapts to the dynamics of available neighbors. In a two week evaluation with 6 subjects, we show that Remora provides up to a 30% accuracy increase while extending phone battery lifetime by over 65%.

In a Body Sensor Network (BSN) activity recognition system, sensor sampling and communication quickly deplete battery reserves. While reducing sampling and communication saves energy, this energy savings usually comes at the cost of reduced recognition accuracy. To address this challenge, we propose AdaSense, a framework that reduces the BSN sensors sampling rate while meeting a user-specified accuracy requirement. AdaSense utilizes a classifier set to do either multi-activity classification that requires a high sampling rate or single activity event detection that demands a very low sampling rate. AdaSense aims to utilize lower power single activity event detection most of the time. It only resorts to higher power multi-activity classification to find out the new activity when it is confident that the activity changes. Furthermore, AdaSense is able to determine the optimal sampling rates using a novel Genetic Programming algorithm. Through this Genetic Programming approach, AdaSense reduces sampling rates for both lower power single activity event detection and higher power multi-activity classification. With an existing BSN dataset and a smartphone dataset we collect from eight subjects, we demonstrate that AdaSense effectively reduces BSN sensors sampling rate and outperforms a state-of-the-art solution in terms of energy savings.

The quality of sleep is an important factor in maintaining a healthy life style. To date, technology has not enabled personalized, in-place sleep quality monitoring and analysis. Current sleep monitoring systems are often difficult to use and hence limited to sleep clinics, or invasive to users, e.g., requiring users to wear a device during sleep. This paper presents iSleep -- a practical system to monitor an individual's sleep quality using off-the-shelf smartphone. iSleep uses the built-in microphone of the smartphone to detect the events that are closely related to sleep quality, including body movement, couch and snore, and infers quantitative measures of sleep quality. iSleep adopts a lightweight decision-tree-based algorithm to classify various events based on carefully selected acoustic features, and tracks the dynamic ambient noise characteristics to improve the robustness of classification. We have evaluated iSleep based on the experiment that involves 7 participants and total 51 nights of sleep, as well the data collected from real iSleep users. Our results show that iSleep achieves consistently above 90% accuracy for event classification in a variety of different settings. By providing a fine-grained sleep profile that depicts details of sleep-related events, iSleep allows the user to track the sleep efficiency over time and relate irregular sleep patterns to possible causes.

In this paper, we present an optimal packet size solution that optimizes the communication energy consumption in the heterogeneous wireless networks. More specifically, we consider a heterogeneous network system composed of a body sensor network (BSN) and a WiFi network. Then, based on the analysis of data communication in the BSN and WiFi (BSN-WiFi) network, we formulate a communication energy consumption optimization model with the constraints of throughput and time delay. Mathematically, we convert this model into a geometric programming problem, which is then numerically solved. The optimal solution can be applied in both BSN and WiFi network to dynamically select packet payload sizes according to real-time packet delivery ratios (PDRs). Since PDRs are time-varying, we tabulate a packet payload size lookup table for online packet size selection using PDRs as indices. Finally, we collect PDRs from a deployed BSN-WiFi network and evaluate the energy optimization model. The performance evaluation results show that, in comparison with fixed packet size solutions, our optimal solutions achieve up to 70 percent energy savings in a BSN(TDMA)-WiFi network and 68 percent in a BSN(CSMA)-WiFi network.

Effective WiFi power management can strongly impact the energy consumption on Smartphones. Through controlled experiments, we find that WiFi power management on a wide variety of Smartphones is a largely autonomous process that is processed completely at the driver level. Driver level implementations suffer from the limitation that important power management decisions can be made only by observing packets at the MAC layer. This approach has the unfortunate side effect that each application has equal opportunity to impact WiFi power management to consume more energy, since distinguishing between applications is not feasible at the MAC layer. The power cost difference between WiFi power modes is high (a factor of 20 times when idle), therefore determining which applications are permitted to impact WiFi power management is an important and relevant problem. In this paper we propose SAPSM: Smart Adaptive Power Save Mode. SAPSM labels each application with a priority with the assistance of a machine learning classifier. Only high priority applications affect the client's behavior to switch to CAM or Active mode, while low priority traffic is optimized for energy efficiency. Our implementation on an Android Smartphone improves energy savings by up to 56% under typical usage patterns.

Body sensor networks (BSNs) and WiFi networks have been widely investigated due to the availability of sensor motes and WiFi devices, but they are commonly deployed separately. In this paper we propose to optimize the total communication energy consumption of BSN and WiFi (BSN-WiFi) networks using joint data rate adaptation. More specifically, we first elaborate the BSN-WiFi network system in four consecutive phases. Then based on the system, we analyze the communication energy consumption, throughput and time delay, and provide a signal-to-noise ratio and packet delivery ratio (SNR-PDR) mappings of BSN and WiFi networks. Next, we build an energy optimization model with constraints of SNR-PDR mappings, throughput, and time delay to minimize the total communication energy consumption in BSN-WiFi networks. With the input of SNR values, we solve this model by cvx to obtain the output of optimal data rates associated with SNR values, which are then tabulated for online data rate adaptation. Finally, we collect 20-minute traces from a specific BSN-WiFi network system for performance evaluation, and the results demonstrate that our optimal data rate solution achieves up to 86% energy savings comparing with the solutions using fixed data rates.

Automatically recognizing human activities in a body sensor network (BSN) enables many human-centric applications. Many current works recognize human activities through collecting and analyzing sensor readings from on-body sensor nodes. These sensing-based solutions face a dilemma. On one hand, to guarantee data availability and recognition accuracy, sensing-based solutions have to either utilize a high transmission power or involve a packet retransmission mechanism. On the other hand, enhancing the transmission power increases a sensor node's energy overheads and communication range. The enlarged communication range in consequence increases privacy risks. A packet retransmission mechanism complicates on-body sensor nodes' MAC layer and hence increases energy overheads. In contrast to the sensing-based solutions, we build Radio Sense, a prototype system that exploits wireless communication patterns for BSN activity recognition. Using Radio Sense, we benchmark three system parameters (transmission (TX) power, packet sending rate, and smoothing window size) to design algorithms for system parameter selection. The algorithms aim to balance accuracy, latency, and energy overheads. In addition, we investigate the minimal amount of training data needed for reliable performance. We evaluate our Radio Sense system with multiple subjects' data collected over a two-week period and demonstrate that Radio Sense achieves reliable performance in terms of accuracy, latency, and battery lifetime.

Since one-third of a smart phone's battery energy is consumed by its WiFi interface, it is critical to switch the WiFi radio from its active or Constantly Awake Mode (CAM), which draws high power (726mW with screen off), to its sleep or Power Save Mode (PSM), which consumes little power (36mW). Applications like VoIP do not perform well under PSM mode however, due to their real-time nature, so the energy footprint is quite high. The challenge is to save energy while not affecting performance. In this paper we present SiFi: Silence prediction based WiFi energy adaptation. SiFi examines audio streams from phone calls and tracks when silence periods start and stop. This data is stored in a prediction model. Using this historical data, we predict the length of future silence periods and place the WiFi radio to sleep during these periods. We implement the design on an Android Smart phone and acheive 40% energy savings while maintaining high voice fidelity.

Wireless sensor networks for human health monitoring, military surveillance, and disaster warning all have stringent accuracy requirements for detecting or classifying events while maximizing system lifetime. We define meeting such user accuracy requirements as confident sensing. To perform confident sensing and reduce energy, we must address sensing diversity: sensing capability differences among heterogeneous and homogeneous sensors in a specific deployment. We are among the first to explore the impact of sensing diversity on sensor collaboration, exploit diversity for sensing confidence, and apply diversity exploitation for confident sensing coverage. We show that our diversity-exploiting confident coverage problem is NP-hard for any specific deployment and present a practical solution, Wolfpack. Through a distributed and iterative sensor collaboration approach, Wolfpack maximizes a specific deployment's capability to meet user detection requirements and save energy by powering off unneeded nodes. Using real vehicle detection trace data, we demonstrate that Wolfpack provides confident event detection coverage for 30% more detection locations, using 20% less energy than a state of the art approach.

Body sensor networks (BSNs) have been developed for a set of performance-critical applications, including smart healthcare, assisted living, emergency response, athletic performance evaluation, and interactive controls. Many of these applications require stringent performance assurance in terms of communication throughput and bounded time delay. While solutions exist in literature for providing joint throughput and time delay assurance by proposing specific MAC protocols or extensions, we provide this joint assurance in a novel radio-agnostic manner. In our approach, the underlying MAC and PHY layers can be heterogeneous and their details do not need to be known to upper layers like the resource management. Such a radio-agnostic performance assurance is critical because a range of radio platforms are adopted for practical body sensor usage. Our approach is based on a group-polling scheme that is essential for radio-agnostic BSN design. Through theoretical analysis, we prove that with the group-polling scheme, achieving joint throughput and time delay assurance is an NP-hard problem. For practical system deployment, we propose the BodyT2 framework that assures throughput and time delay performance in a heterogeneous BSN. Through both TelosB mote lab tests and real body experiments in an Android phone-centric BSN, we demonstrate that BodyT2 achieves superior performance over existing solutions.

In this paper, we propose to optimize energy consumption in heterogeneous wireless networks through joint packet size optimization. Specifically, we consider a two-hop data communication system composed of a body sensor network (BSN) and a WiFi network. Within the system, we formulate an energy consumption optimization problem with the constraints of both throughput and time delay. Mathematically, we convert this problem into a geometric programming (GP) problem, which is then numerically solved. The solutions can be used by both the BSN and the WiFi network to dynamically change their packets' payload sizes based on their current packet delivery ratios (PDRs). Since the PDRs are time-varying, we tabulate an offline payload size lookup table for online packet size selection using PDRs as indices. Finally, we collect PDRs from a deployed two-hop BSN-WiFi network and simulate the energy consumption. The performance evaluation results show that our solution achieves up to 70% energy savings compared with solutions that use fixed packet sizes.

The vast array of small wireless sensors is a boon to body sensor network applications, especially in the context awareness and activity recognition arena. However, most activity recognition deployments and applications are challenged to provide personal control and practical functionality for everyday use. We argue that activity recognition for mobile devices must meet several goals in order to provide a practical solution: user friendly hardware and software, accurate and efficient classification, and reduced reliance on ground truth. To meet these challenges, we present PBN: Practical Body Networking. Through the unification of TinyOS motes and Android smartphones, we combine the sensing power of on-body wireless sensors with the additional sensing power, computational resources, and user-friendly interface of an Android smartphone. We provide an accurate and efficient classification approach through the use of ensemble learning. We explore the properties of different sensors and sensor data to further improve classification efficiency and reduce reliance on user annotated ground truth. We evaluate our PBN system with multiple subjects over a two week period and demonstrate that the system is easy to use, accurate, and appropriate for mobile devices.

As wireless devices and sensors are increasingly deployed on people, researchers have begun to focus on wireless body-area networks. Applications of wireless body sensor networks include healthcare, entertainment, and personal assistance, in which sensors collect physiological and activity data from people and their environments. In these body sensor networks, quality of service is needed to provide reliable data communication over prioritized data streams. This article proposes BodyQoS, the first running QoS system demonstrated on an emulated body sensor network. BodyQoS adopts an asymmetric architecture, in which most processing is done on a resource-rich aggregator, minimizing the load on resource-limited sensor nodes. A virtual MAC is developed in BodyQoS to make it radio-agnostic, allowing a BodyQoS to schedule wireless resources without knowing the implementation details of the underlying MAC protocols. Another unique property of BodyQoS is its ability to provide adaptive resource scheduling. When the effective bandwidth of the channel degrades due to RF interference or body fading effect, BodyQoS adaptively schedules remaining bandwidth to meet QoS requirements. We have implemented BodyQoS in NesC on top of TinyOS, and evaluated its performance on MicaZ devices. Our system performance study shows that BodyQoS delivers significantly improved performance over conventional solutions in combating channel impairment.

Recent years have witnessed the increasing adoption of ZigBee technology for performance-sensitive applications such as wireless patient monitoring in hospitals. However, operating in unlicensed ISM bands, ZigBee devices often yield unpredictable throughput and packet delivery ratio due to the interference from ever increasing WiFi hotspots in 2.4 GHz band. Our empirical results show that, although WiFi traffic contains abundant white space, the existing coexistence mechanisms such as CSMA are surprisingly inadequate for exploiting it. In this paper, we propose a novel approach that enables ZigBee links to achieve assured performance in the presence of heavy WiFi interference. First, based on statistical analysis of real-life network traces, we present a Pareto model to accurately characterize the white space in WiFi traffic. Second, we analytically model the performance of a ZigBee link in the presence of WiFi interference. Third, based on the white space model and our analysis, we develop a new ZigBee frame control protocol called WISE, which can achieve desired trade-offs between link throughput and delivery ratio. Our extensive experiments on a testbed of 802.11 netbooks and 802.15.4 TelosB motes show that, in the presence of heavy WiFi interference, WISE achieves 4× and 2× performance gains over B-MAC and a recent reliable transmission protocol, respectively, while only incurring 10.9% and 39.5% of their overhead.

Packet collision causes packet loss and wastes resources in wireless networks. It becomes even worse in dense WSNs, due to burst-traffic and congestion around sinks. In this paper, we propose a novel protocol to recover collided packets. Our experiments on a testbed reveal that collisions between long packets and short packets cause a partial error pattern on collided packets, which can be used for efficient recovery. We give a theoretical analysis that demonstrates that combining such collision recovery with CSMA protocols achieves a significant performance improvement. Then, we design ACR, an Active Collision Recovery protocol, which actively converts most potential collisions into LS-collisions, and then applies a lightweight FEC scheme to recover collided packets with such partial error patterns. We implement ACR on a Tmote testbed, and compare its performance with other packet recovery schemes. Results show that ACR significantly reduces the number of retransmissions, and achieves around 25% improvement on transmission efficiency over other schemes.

Many mission-critical applications such as military surveillance, human health monitoring, and obstacle detection in autonomous vehicles impose stringent requirements for event detection accuracy and demand long system lifetimes. Through quantitative study, we show that traditional approaches to event detection have difficulty meeting such requirements. Specifically, they cannot explore the detection capability of a deployed system and choose the right sensors, homogeneous or heterogeneous, to meet user specified detection accuracy. They also cannot dynamically adapt the detection capability to runtime observations to save energy. Therefore, we are motivated to propose Watchdog, a modality-agnostic event detection framework that clusters the right sensors to meet user specified detection accuracy during runtime while significantly reducing energy consumption. Through evaluation with vehicle detection trace data and a building traffic monitoring testbed of IRIS motes, we demonstrate the superior performance of Watchdog over existing solutions in terms of meeting user specified detection accuracy and energy savings.

Multifrequency media access control has been well understood in general wireless ad hoc networks, while in wireless sensor networks, researchers still focus on single frequency solutions. In wireless sensor networks, each device is typically equipped with a single radio transceiver and applications adopt much smaller packet sizes compared to those in general wireless ad hoc networks. Hence, the multifrequency MAC protocols proposed for general wireless ad hoc networks are not suitable for wireless sensor network applications, which we further demonstrate through our simulation experiments. In this article, we propose MMSN, which takes advantage of multifrequency availability while, at the same time, takes into consideration the restrictions of wireless sensor networks. In MMSN, four frequency assignment options are provided to meet different application requirements. A scalable media access is designed with efficient broadcast support. Also, an optimal nonuniform back-off algorithm is derived and its lightweight approximation is implemented in MMSN, which significantly reduces congestion in the time synchronized media access design. Through extensive experiments, MMSN exhibits the prominent ability to utilize parallel transmissions among neighboring nodes. When multiple physical frequencies are available, it also achieves increased energy efficiency, demonstrating the ability to work against radio interference and the tolerance to a wide range of measured time synchronization errors.

Smart wireless sensor devices are becoming increasingly ubiquitous and are expected to be embedded in everyday objects in the near future. When these devices are deployed in overlapping or adjacent geographic areas, the unlicensed 2.4 GHz ISM band will be crowded. To deal with the crowded spectrum, we propose SAS, a self-adaptive spectrum management middleware for wireless sensor networks. SAS enables single-frequency MAC protocols with multi-frequency capability, so that an existing MAC protocol, like B-MAC, can automatically adapt to the least congested physical channel at runtime. We implemented SAS in TinyOS 2.1 with nesC and evaluated its performance with TelosB motes. Our performance results demonstrate that SAS improves the performance of existing single-frequency MAC protocols, like B-MAC. The use of SAS results in higher packet reception ratio and system throughput, and a lower packet delay and energy consumption.

This paper presents C-MAC, a new MAC protocol designed to achieve high-throughput bulk communication for data-intensive sensing applications. C-MAC exploits concurrent wireless channel access based on empirical power control and physical interference models. Nodes running C-MAC estimate the level of interference based on the physical signal-to-interference-plus-noise-ratio (SINR) model and adjust the transmission power accordingly for concurrent channel access. C-MAC employs a block-based communication mode that not only amortizes the overhead of channel assessment, but also improves the probability that multiple nodes within the interference range of each other can transmit concurrently. C-MAC has been implemented in TinyOS-1.x and extensively evaluated on Tmote nodes. Our experiments show that C-MAC significantly outperforms the state-of-art CSMA protocol in TinyOS with respect to system throughput, delay and energy consumption.

Many applications in wireless sensor networks require communication performance that is both consistent and high quality. Unfortunately, performance of current network protocols can vary significantly because of various interferences and environmental changes. Current protocols estimate link quality based on the reception of probe packets over a short time period. This method is neither efficient nor accurate enough to capture the dramatic variations of link quality. Therefore, we propose a link metric called competence that characterizes links over a longer period of time. We combine competence with current short term estimations in routing algorithm designs. To further improve network performance we have designed a distributed route maintenance framework based on feedback control solutions. In real system evaluations with 48 T-Motes, our overall solution improves end-to-end packet delivery ratio over existing solutions by up to 40%, while reducing energy consumption by up to 22%. Importantly, our solution also achieves more stable and better transient performance than current approaches.

Multi-channel design has received significant attention for low-power wireless networks (LWNs), such as 802.15.4-based wireless sensor networks, due to its potential of mitigating interference and improving network capacity. However, recent studies reveal that the number of orthogonal channels available on commodity wireless platforms is small, which significantly hinders the performance of existing multi-channel protocols. A promising solution is to explore the use of partially overlapping channels for communications. However, this approach faces several key challenges such as increased inter-channel interference and significantly higher overhead of channel measurement. In this paper, we systematically study the inter-channel interference and its impact on link capacity and the performance of multi-channel protocols in LWNs. First, we develop empirical models for characterizing inter-channel signal attenuation based on experiments on TelosB motes. We then propose a novel measurement algorithm which can significantly reduce the overhead of multi-channel interference measurement by exploiting the spectral power density (SPD) of the transmitter. Finally, we apply our interference models to both link capacity analysis and channel assignment protocols. Our extensive experiments on a testbed of 30 TelosB motes show that our interference measurement algorithm has an average error of 2.95%. Our results also demonstrate that multi-channel protocols for LWNs can significantly benefit from using overlapping channels.

In-situ data collection for mobile wireless sensor network deployments has received little study, such as in the case of floating sensor networks for storm surge and innundation monitoring. We demonstrate through quantitative study that traditional approaches to routing in mobile environments do not work well due to volatile topology changes. Consequently, we propose Sidewinder, a predictive data forwarding protocol for mobile wireless sensor networks. Like a heat-seeking missile, data packets are guided towards a sink node with increasing accuracy as packets approach the sink. Different from conventional sensor network routing protocols, Sidewinder continuously predicts the current sink location based on distributed knowledge of sink mobility among nodes in a multi-hop routing process. Moreover, the continuous sink estimation is scaled and adjusted to perform with resource-constrained wireless sensors. Our design is implemented with nesC and evaluated in TOSSIM. The performance evaluation demonstrates that Sidewinder significantly outperforms state-of-the-art solutions in packet delivery ratio, time delay, and energy efficiency.

Energy efficiency is a fundamental issue for outdoor sensor network systems. This article presents the design and implementation of multidimensional power management strategies in VigilNet, a major recent effort to support long-term surveillance using power-constrained sensor devices. A novel tripwire service is integrated with an effective sentry and duty cycle scheduling in order to increase the system lifetime, collaboratively. The tripwire service partitions a network into distinct, nonoverlapping sections and allows each section to be scheduled independently. Sentry scheduling selects a subset of nodes, the sentries, which are turned on while the remaining nodes save energy. Duty cycle scheduling allows the active sentries themselves to be turned on and off, further lowering the average power draw. The multidimensional power management strategies proposed in this article were fully implemented within a real sensor network system using the XSM platform. We evaluate key system parameters using a network of 200 XSM nodes in an outdoor environment, and an analytical probabilistic model. We evaluate network lifetime using a simulation of a 10,000-node network that uses measured XSM power values. These evaluations demonstrate the effectiveness of our integrated approach and identify a set of lessons and guidelines, useful for the future development of energy-efficient sensor systems. One of the key results indicates that the combination of the three presented power management techniques is able to increase the lifetime of a realistic network from 4 days to 200 days.

In this paper, we analyze the performance of group based detection in sparse sensor networks, when the system level detection decision is made based on the detection reports generated from multiple sensing periods. Sparse deployment is essential for reducing cost of large scale sensor networks, which cover thousands of square miles. In a sparse deployment, the sensor field is only partially covered by sensorspsila sensing ranges, resulting in void sensing areas in the region, but all nodes are connected through multi-hop networking. Further, due to the unavoidable false alarms generated by a single sensor in a network, many deployed systems use group based detection to reduce system level false alarms. Despite the popularity of group based detection, few analysis works in the literature deal with group based detection. In this paper, we propose a novel approach called Markov chain based Spatial approach (MS-approach) to model group based detection in sensor networks. The M-S-approach successfully overcomes the complicated conditional detection probability of a target in each sensing period, and reduces the execution time of the analysis from many days to 1 minute. The analytical model is validated through extensive simulations. This analytical work is important because it provides an easy way to understand the performance of a system that uses group based detection without running countless simulations or deploying real systems.

As wireless devices and sensors are increasingly deployed on people, researchers have begun to focus on wireless body-area networks. Applications of wireless body sensor networks include healthcare, entertainment, and personal assistance, in which sensors collect physiological and activity data from people and their environments. In these body sensor networks, quality of service is needed to provide reliable data communication over prioritized data streams. This paper proposes BodyQoS, the first running QoS system demonstrated on an emulated body sensor network. BodyQoS adopts an asymmetric architecture, in which most processing is done on a resource rich aggregator, minimizing the load on resource limited sensor nodes. A virtual MAC is developed in BodyQoS to make it radio-agnostic, allowing a BodyQoS to schedule wireless resources without knowing the implementation details of the underlying MAC protocols. Another unique property of BodyQoS is its ability to provide adaptive resource scheduling. When the effective bandwidth of the channel degrades due to RF interference or body fading effect, BodyQoS adaptively schedules remaining bandwidth to meet QoS requirements. We have implemented BodyQoS in NesC on top of TinyOS, and evaluated its performance on MicaZ devices. Our system performance study shows that BodyQoS delivers significantly improved performance over conventional solutions in combating channel impairment.

Traditional fall detection is only based on acceleration analysis. In this work we present a novel fall detection method that also utilizes posture and context information. This information can help reduce both false positives and negatives. Our solution also strives for low computational cost and fast response.

Extensive empirical results reveal that interference can cause link qualities to change quickly and dramatically. For such highly dynamic links, the short term link quality estimations widely used in existing protocols require frequent measurements and may not be accurate. As a result, when these links are selected, end-to-end communication quality varies significantly. Also, route changes occur frequently, introducing traffic oscillation and excessive overhead in network protocols. To achieve good and stable network performance, it is not enough to use short term link estimation. It is essential to characterize a link's capacity to perform well at a desired level in the presence of interference and environmental changes. Therefore, we propose a performance metric called competence. We have incorporated the competence metric into routing algorithm designs. We have also designed and implemented a maintenance framework that stabilizes performance at both link and network layers. This framework allocates the desired performance level among multiple links along an active route by using an end-to-end feedback loop, and enforces the performance level of each link through adaptive transmission power control and retransmission control. In real system evaluations with 48 TMotes, our solution outperforms previous protocols significantly and achieves end-to-end stable performance for more than 99% of the time over 24 hours.

This work presents a TDMA based multi-channel MAC protocol called TMMAC for Ad Hoc Networks. TMMAC requires only a single half-duplex radio transceiver on each node. In addition to explicit frequency negotiation which is adopted by conventional multi-channel MAC protocols, TMMAC introduces lightweight explicit time negotiation. This two-dimensional negotiation enables TMMAC to exploit the advantage of both multiple channels and TDMA, and achieve aggressive power savings by allowing nodes that are not involved in communication to go into doze mode. Moreover, TMMAC dynamically adjusts its negotiation window size based on different traffic patterns, which further improves communication throughput and energy savings. In this paper, the performance of TMMAC is analyzed and evaluated. The evaluations show that TMMAC achieves up to 113% higher communication throughput while consuming 74% less per packet energy over the state-of-the-art multi-channel MAC protocols for single-transceiver wireless devices.

No abstract available.

In this paper we address three challenges that are present when building and analyzing wireless sensor networks (WSN) as part of ubiquitous computing environment: the need for an in-situ user interface, a data logger, and a power consumption meter. Solutions for the above have been presented using laptops, personal digital assistants (PDA), onboard flash memory chips of limited size (usually 1MB), and laboratory test equipment. All of them have a good utility for the right applications. However, considering a certain variety of WSNs, where size, battery life, and cost are crucial, none of the above solutions is satisfactory. In this paper we present a compact, lightweight, low power, and low cost multimodal sensor module SeeMote that meets the stated challenges, and is compatible with the popular MICAz mote. Our module has the following components: (1) a graphical user interface component that combines a color liquid crystal display (LCD) and 5-way buttons, (2) a power meter component that is reconfigurable for attaching various low-power devices, and (3) a data logger component that is interfaced to a removable secure digital (SD) or multimedia memory card (MMC). The module dimensions are 34times58times12 mm. This paper describes the hardware and software design and experiences while developing and using the device. The device is evaluated by comparing its parameters and functionality to laptop and PDA solutions. We conclude that SeeMote is preferred for certain WSNs, such as very large scale, difficult to reach, and wearable WSNs. We also present several applications that use the LCD module, such as the portable frequency spectrum analyzer and remote sensory data display device.

Abstract — Energy efficiency is a fundamental issue for out-door sensor network systems. This paper presents the design and implementation of multi-dimensional power management strategies in VigilNet, a major recent effort to support long-term surveillance using power-constrained sensor devices. We integrate a novel tripwire service with an effective sentry and duty cycle scheduling in order to increase the system lifetime, collaboratively. Through extensive system implementation, we demonstrate the feasibility to achieve high surveillance perfor-mance and energy efficiency, simultaneously. We invest a fair amount of effort to evaluate our architecture with a network of 200 XSM motes in an outdoor environment, an extensive simulation with 10,000 nodes, as well as an analytical probabilistic model. These evaluations demonstrate the effectiveness of our integrated approach and identify many interesting lessons and guidelines, useful for the future development of energy-efficient sensor systems. I.

Abstract — Sensing events from dynamic environments are normally asynchronous and non-repeatable. This lack of repeatability makes it particularly difficult to statistically evaluate the performance of sensor network applications. Hence, it is essential to have the capability to capture and replay sensing events, providing a basis not only for system evaluation, but also for realistic protocol comparison and parameter tuning. To achieve that, we design and implement EnviroLog, a distributed service that improves repeatability of experimental testing of sensor networks via asynchronous event recording and replay. To use EnviroLog, an application programmer needs only to specify two types of simple annotations to the source code. Automatically, the preprocessor embeds EnviroLog into any desired level of an event-driven architecture. It records all events generated by lower layers and can replay them later to upper layers on demand. We validate the accuracy and performance of recording and replay through a set of microbenchmarks, using the latest XSM platforms. We further demonstrate the strength of EnviroLog in system tuning and performance evaluation for sensor network applications in an outdoor environment with 37 XSMs. I.

Abstract — Multi-frequency media access control has been well understood in general wireless ad hoc networks, while in wireless sensor networks, researchers still focus on single frequency solutions. In wireless sensor networks, hardware devices are equipped with very limited communication ability and applications adopt much smaller packet sizes compared to those in general wireless ad hoc networks. Hence, the multi-frequency MAC protocols proposed for general wireless ad hoc networks are not suitable for wireless sensor network applications, which we further demonstrate through our simulation experiments. In this paper, we propose MMSN, the first multi-frequency MAC protocol for wireless sensor networks. In the MMSN protocol, four frequency assignment options are provided to meet different application requirements. A scalable media access is designed with efficient broadcast support. Also, an optimal non-uniform backoff algorithm is derived and its lightweight approximation is implemented in MMSN, which significantly reduces congestion in the time synchronized media access design. Through extensive experiments, MMSN exhibits prominent ability to utilize parallel transmission among neighboring nodes. It also achieves increased energy efficiency when multiple physical frequencies are available. I.

Extensive empirical studies presented in this paper confirm that the quality of radio communication between low power sensor devices varies significantly with time and environment. This phenomenon indicates that the previous topology control solutions, which use static transmission power, transmission range, and link quality, might not be effective in the physical world. To address this issue, online transmission power control that adapts to external changes is necessary. This paper presents ATPC, a lightweight algorithm of Adaptive Transmission Power Control for wireless sensor networks. In ATPC, each node builds a model for each of its neighbors, describing the correlation between transmission power and link quality. With this model, we employ a feedback-based transmission power control algorithm to dynamically maintain individual link quality over time. The intellectual contribution of this work lies in a novel pairwise transmission power control, which is significantly different from existing node-level or network-level power control methods. Also different from most existing simulation work, the ATPC design is guided by extensive field experiments of link quality dynamics at various locations and over a long period of time. The results from the real-world experiments demonstrate that 1) with pairwise adjustment, ATPC achieves more energy savings with a finer tuning capability and 2) with online control, ATPC is robust even with environmental changes over time.

This article describes one of the major efforts in the sensor network community to build an integrated sensor network system for surveillance missions. The focus of this effort is to acquire and verify information about enemy capabilities and positions of hostile targets. Such missions often involve a high element of risk for human personnel and require a high degree of stealthiness. Hence, the ability to deploy unmanned surveillance missions, by using wireless sensor networks, is of great practical importance for the military. Because of the energy constraints of sensor devices, such systems necessitate an energy-aware design to ensure the longevity of surveillance missions. Solutions proposed recently for this type of system show promising results through simulations. However, the simplified assumptions they make about the system in the simulator often do not hold well in practice, and energy consumption is narrowly accounted for within a single protocol. In this article, we describe the design and implementation of a complete running system, called VigilNet, for energy-efficient surveillance. The VigilNet allows a group of cooperating sensor devices to detect and track the positions of moving vehicles in an energy-efficient and stealthy manner. We evaluate VigilNet middleware components and integrated system extensively on a network of 70 MICA2 motes. Our results show that our surveillance strategy is adaptable and achieves a significant extension of network lifetime. Finally, we share lessons learned in building such an integrated sensor system.

In this article, we investigate the impact of radio irregularity on wireless sensor networks. Radio irregularity is a common phenomenon that arises from multiple factors, such as variance in RF sending power and different path losses, depending on the direction of propagation. From our experiments, we discover that the variance in received signal strength is largely random; however, it exhibits a continuous change with incremental changes in direction. With empirical data obtained from the MICA2 and MICAZ platforms, we establish a radio model for simulation, called the Radio Irregularity Model (RIM). This model is the first to bridge the discrepancy between the spherical radio models used by simulators and the physical reality of radio signals. With this model, we investigate the impact of radio irregularity on several upper layer protocols, including MAC, routing, localization and topology control. Our results show that radio irregularity has a relatively larger impact on the routing layer than the MAC layer. It also shows that radio irregularity leads to larger localization errors and makes it harder to maintain communication connectivity in topology control. To deal with these issues, we present eight solutions to deal with radio irregularity. We evaluate three of them in detail. The results obtained from both the simulations and a running testbed demonstrate that our solutions greatly improve system performance in the presence of radio irregularity.

In wireless sensor networks, many protocols assume that if node A is able to interfere with node B's packet reception, node B is within node A's communication range. It is also assumed that if node B is within node A's communication range, node A is able to interfere with node B's packet reception from any transmitter. While these assumptions may be useful in protocol design, they are not valid, according to the real experiments we conducted in MICA2 platform. For a strong link that has a high packet delivery ratio, the interference range is observed smaller than the communication range, while for a weak link that has a low packet delivery ratio, the interference range is larger than the communication range. So using communication range information alone is not enough to design real collision-free media access control protocols. This paper presents a radio interference detection protocol (RID) and its variation (RID-B) to detect run-time radio interference relations among nodes. The interference detection results are used to design real collision-free TDMA protocols. With extensive simulations in GlomoSim, and with sensor network application scenarios, we observe that the TDMA which uses the interference detection results has 100% packet delivery ratio, while the traditional TDMA has packet loss up to 60%, in heavy load. In addition to the scheduling-based TDMA protocols, we also explore the application of interference detection on contention-based MAC protocols.

Battlefield surveillance often involves a high element of risk for military operators. Hence, it is very important for the military to execute unmanned surveillance by using large-scale wireless sensor systems. This invited paper summarizes the architecture of the VigilNet system - a long-term real-time networked sensor system for military surveillance. Specifically, we review the design of several major subsystems within VigilNet including sensing and classification, localization, tracking, networking, power management, reconfiguration, graphic user interface, and the debugging subsystem. High-level programming abstractions are also presented. This is a balanced design to achieve realtime response, high confidence detection, accurate tracking and energy efficiency simultaneously.

In this paper we present a balanced data replication scheme that provides real-time latency bounds on content retrieval in content distribution networks. Many network applications have ever-increasing requirements on latency sensitive data services. Data replication services have been widely used as an important performance enhancement mechanism to reduce data access latency and throughput. We investigate the problem of provisioning an underlying balanced data replication service to provide a global latency bound on data retrieval in content distribution networks. The solution involves constructing an overlay network based on the given latency bound, and a mechanism to assign content objects to the network nodes so that the workload of all the network nodes is balanced. Our evaluation results drawn from detailed simulations show the efficacy of our load-balancing scheme in meeting the latency bound requirements with high confidence under heavy load.

In this paper, we investigate the impact of radio irregularity on the communication performance in wireless sensor networks. Radio irregularity is a common phenomenon which arises from multiple factors, such as variance in RF sending power and different path losses depending on the direction of propagation. From our experiments, we discover that the variance in received signal strength is largely random; however, it exhibits a continuous change with incremental changes in direction. With empirical data obtained from the MICA2 platform, we establish a radio model for simulation, called the Radio Irregularity Model (RIM). This model is the first to bridge the discrepancy between spherical radio models used by simulators and the physical reality of radio signals. With this model, we are able to analyze the impact of radio irregularity on some of the well-known MAC and routing protocols. Our results show that radio irregularity has a significant impact on routing protocols, but a relatively small impact on MAC protocols. Finally, we propose six solutions to deal with radio irregularity. We evaluate two of them in detail. The results obtained from both the simulation and a running testbed demonstrate that our solutions greatly improve communication performance in the presence of radio irregularity.