Blazing fast RDMA technology revolutionizes modern distributed systems and propels them to offload performance-critical data paths onto this network fabric. Designing an RDMA-optimized data path needs to clear a main hurdle—non-scalable distributed locks. Through a performance dissection of existing lock schemes, we find that software-based lock request ordering and polling-based lock ownership transfer scale poorly, leading to high NIC contention and heavy network congestion. To resolve these bottlenecks, this paper proposes StreamLock, a scalable lock primitive that co-designs the distributed lock protocol with fast RDMA networks. The core of StreamLock is a novel shared stream abstraction with two mechanisms: (i) scalable request ordering by repurposing the line-speed packet receiving provided by modern NICs; (ii) peer-to-peer notification to achieve one-round-trip-time lock ownership transfer. We implement StreamLock with off-the-shelf RDMA NICs and compare it with state-of-the-art distributed locks. Comprehensive experimental results showcase that StreamLock outperforms them significantly.

RFID has revolutionized item-level intelligence in IoT ecosystems, yet static localization with passive tags remains challenging due to multipath interference inherent in RF signals. We present LuxTag, the first system to enable visible light-based sensing and localization using standard, commercial RFID tags by transforming them into ambient light sensors. Our key insight leverages the discovery that photon-induced leakage currents in passive RFID ICs modulate their persistence time (i.e., the duration a tag remains operational after RF excitation ceases) proportional to ambient illuminance. LuxTag introduces two innovations: (i) a first-principles model characterizing how ambient light alters tag persistence time, enabling battery-free light sensing without hardware modifications; (ii) a differential measurement technique and zero-shot calibration method to isolate light effects and autonomously derive tag parameters, ensuring robust and accurate static localization system using COTS RFID infrastructure. Extensive experiments demonstrate that LuxTag achieves a mean light intensity error of 3.6 lux and 60.7% improvement over state-of-the-art static RFID localization. By synergizing the ubiquity of RFID with the multipath resilience of optical sensing, LuxTag opens new avenues for static RFID localization in smart warehouses, retails, and beyond.

Federated Learning (FL), a privacy-preserving decentralized machine learning framework, has been shown to be vulnerable to backdoor attacks. Current research primarily focuses on the Single-Label Backdoor Attack (SBA), wherein adversaries share a consistent target. However, a critical fact is overlooked: adversaries may be non-cooperative, have distinct targets, and operate independently, which exhibits a more practical scenario called Multi-Label Backdoor Attack (MBA). Unfortunately, prior works are ineffective in the MBA scenario since non-cooperative attackers exclude each other. In this work, we conduct an in-depth investigation to uncover the inherent constraints of the exclusion: similar backdoor mappings are constructed for different targets, resulting in conflicts among backdoor functions. To address this limitation, we propose Mirage, the first non-cooperative MBA strategy in FL that allows attackers to inject effective and persistent backdoors into the global model without collusion by constructing in-distribution (ID) backdoor mapping. Specifically, we introduce an adversarial adaptation method to bridge the backdoor features and the target distribution in an ID manner. Additionally, we further leverage a constrained optimization method to ensure the ID mapping survives in the global training dynamics. Extensive evaluations demonstrate that Mirage outperforms various state-of-the-art attacks and bypasses existing defenses, achieving an average ASR greater than 97% and maintaining over 90% after 900 rounds. This work aims to alert researchers to this potential threat and inspire the design of effective defense mechanisms. Code has been made open-source.

Federated Learning (FL) typically involves processing redundant data on resource-constrained edge devices, resulting in prolonged training time. A promising strategy to accelerate FL is on-device data sampling. SOTA methods generally select a subset based on sample importance before each local epoch. However, these methods, which employ mini-batch gradient descent on the static subset, suffer from outdated sample importance, leading to suboptimal sampling efficiency. Furthermore, these sampling methods fail to address the biased gradient expectation introduced by importance sampling, further degrading model accuracy. In this paper, we propose FedTT, a novel framework designed to accelerate FL with improved accuracy through pipelined intra-batch data sampling and model training. Specifically, at the algorithm level, to enable real-time data sampling with unbiased model updates, FedTT performs on-device sampling from the fixed-size mini-batch at each iteration and applies gradient correction to the variable-size sampled micro-batch. At the system level, to further accelerate FL training, FedTT implements a well-designed parallelism and synchronization mechanism that enables pipelined execution of intra-batch data sampling and model training on CPU-GPU architectures. Finally, we conduct extensive real-world experiments and simulations to demonstrate the effectiveness and universal adaptability of FedTT. Compared to the SOTAs, our evaluation on four datasets shows that FedTT improves time-to-accuracy performance by 1.23 × ∼ 2.51 × and model accuracy by 0.39% ∼ 11.24%, with real-world experiments further validating its real-world effectiveness by achieving a 1.75 × speedup and a 9.86% accuracy improvement. Integrated with various FL algorithms and importance criteria, FedTT consistently delivers performance gains. Code has been open sourced.

Manipulating articulated tools, such as tweezers or scissors, has rarely been explored in previous research. Unlike rigid tools, articulated tools change their shape dynamically, creating unique challenges for dexterous robotic hands. In this work, we present a hierarchical, goal-conditioned reinforcement learning (GCRL) framework to improve the manipulation capabilities of anthropomorphic robotic hands using articulated tools. Our framework comprises two policy layers: (1) a low-level policy that enables the dexterous hand to manipulate the tool into various configurations for objects of different sizes, and (2) a high-level policy that defines the tool’s goal state and controls the robotic arm for object-picking tasks. We employ an encoder, trained on synthetic pointclouds, to estimate the tool’s affordance states—specifically, how different tool configurations (e.g., tweezer opening angles) enable grasping of objects of varying sizes—from input point clouds, thereby enabling precise tool manipulation. We also utilize a privilege-informed heuristic policy to generate replay buffer, improving the training efficiency of the high-level policy. We validate our approach through real-world experiments, showing that the robot can effectively manipulate a tweezer-like tool to grasp objects of diverse shapes and sizes with a 70.8% success rate. This study highlights the potential of RL to advance dexterous robotic manipulation of articulated tools.

Discovering potential vulnerabilities has long been a fundamental goal in software security. Among them, bit flips, caused by hardware or environmental disturbances, are increasingly recognized as a new type of vulnerabilities that threaten program reliability at the instruction level. However, existing work is often restricted to individual programs and requires retraining when applied to unseen code, severely limiting their practicality and responsiveness. In this paper, we propose CIVP, a novel framework for context-aware instruction vulnerability prediction, generalizing to unseen programs without retraining. Specifically, to capture the rich contextual semantics of instructions, CIVP first leverages Large Language Models (LLMs) to accurately extract semantic embeddings of instructions. Then, CIVP further constructs an instruction execution graph containing complex relations of program execution, which implicates the potential path of error propagation. To improve instruction representation for vulnerability prediction, CIVP enhances GraphSAGE with multi-hop diffusion to capture inter-program structural patterns and contextual dependencies, and adopts pseudo-labeling to improve the model’s generalization for vulnerable instructions. Extensive experiments on a dataset of 26 real-world programs demonstrate that CIVP significantly outperforms the state-of-the-art approaches, achieving up to 20.5%↑ Recall and 18.5%↑ F1-score improvements. Notably, CIVP generalizes well to unseen programs, offering an efficient and scalable solution for proactive instruction-level hardening before software deployment.

Image stitching is often affected by obstructions such as pedestrians and vehicles. Traditional methods often ignore obstructions or retain them, leading to artifacts. They also cause edge distortion due to fixed-perspective stitching. To address above issues, this poster presents IIS, an IMU-assisted Image Stitching method. It uses IMU data for perspective pre-alignment, projecting images to an intermediate perspective to reduce edge distortion. It simultaneously removes obstructions and mitigating their impact. A weighted target scene stitching strategy is integrated to further enhance stitching quality. IIS effectively reduces artifacts and maintains high computational efficiency.

Image-based indoor localization using smartphones has become popular, leveraging visual landmarks and fingerprint extraction for localization. Fingerprint density significantly affects accuracy, but collecting dense, high-resolution fingerprints during on-site surveys is labor-intensive and incurs high computation/storage costs during matching. Additionally, efficient fingerprint extraction often constrains users to specific shooting poses, with deviations markedly reducing localization accuracy. To address these challenges, we introduce ARGILS, an Automated Real-time Generative Image Localization System. The key idea is to use cross sparse sampling instead of dense sampling, generate fingerprint features for missing locations, and quickly match locations through feature orthogonal decomposition. Cross sparse sampling ensures full coverage of scene features and helps to generate missing fingerprints. To maintain high localization resolution with sparse sampling, we designed a distance-constrained generative adversarial network to generate fingerprints for unsampled locations. Additionally, we developed an orthogonal fingerprint extraction method to decompose image features into horizontal and vertical directions in 2D space. To improve robustness against obstacles, we implemented a scanning localization scheme using key frame filtering and clustering. We have implemented ARGILS and performed extensive real-world evaluations. Experiment results show that when reducing 75% site survey effort, the average location error of ARGILS is around 2.5m in a shopping mall, 48% higher than state-of-the-art methods. ARGILS can also efficiently speed up localization process, with the time consumption ranging from 0.1 to 0.3 seconds on smartphones of various configurations.

Unmanned aerial vehicles (UAVs) equipped with sensors, computing units, and communication modules, together with ground devices, constitute a ubiquitous integrated low-altitude network, which can provide users with sustainable computing and communication services in areas where terrestrial infrastructure has been compromised or rendered inoperable. Federated learning-enabled UAV (FL-UAV) wireless networks fully utilize the computational and communication capabilities of UAVs to protect user data privacy by exchanging model updates with ground devices. However, facing the challenges of low energy utilization efficiency and high training latency caused by UAV deployment, resource allocation, and communication overhead in FL-UAV. Existing solutions do not achieve efficient communication and resource scheduling to solve the energy and delay optimization issues in FL-UAV wireless networks. In this paper, we propose an air-to-ground integrated federated distillation (AirFD) framework for UAV-assisted mobile computing and communication networks, which significantly reduces communication overhead between UAV and ground devices by introducing knowledge distillation to transmit average logits instead of model parameters. Furthermore, we formulate cross-layer resource scheduling in AirFD as a non-convex optimization problem to achieve a trade-off between energy consumption and delay. To solve this nonlinear coupling and NP-complete problem, we use successive convex approximation and greedy algorithm to obtain the local optimal solution. Simulation evaluation and field experiments confirm the effectiveness of our proposed method in reducing communication costs and training delays by nearly 40%, and increasing energy utilization by about 50%.

Recent years have witnessed increasing privacy concerns towards machine learning. To protect privacy in machine learning, federated learning has been proposed as a decentralized privacy-preserving framework where clients upload the parameters rather than private data. However, training a fair federated learning model in heterogeneous environments is still challenging. First, heterogeneous data distributions lead the global model fail to show high accuracy on all distributions. Second, the federated learning training process exposes and exacerbates potential biases in heterogeneous training data. Third, the local bias of each client can be propagated through parameter sharing, biasing the global model. In this work, we propose a two-stage fairness-aware federated learning framework (HeteroFair) to achieve fairness under heterogeneous data distributions. Initially, we introduce the fairness constraint to the loss function and propose a local adaptive weighting algorithm to adjust the proportion of the fairness constraint, achieving fair training in heterogeneous environments. Then, we present a fairness-aware aggregation reweighting algorithm that reduces the mismatch between local and global fairness to achieve fair federated learning. Extensive evaluation results demonstrate the effectiveness of our proposed framework in achieving fairness and high accuracy under het-eroaeneous data distributions.

As a computing paradigm tailored for resource-constrained client devices, federated learning based on model pruning compresses the model size by removing unimportant parameters in the neural network, which has shown outstanding results in improving model efficiency and reducing computing costs. However, previous works simply customized a unified static model pruning rate, ignoring the heterogeneous capabilities of clients and the impact of pruning on different layers of the model during continuous iteration. In this paper, we design a novel Federated learning framework based on Dynamic Layer-wise Pruning, named FedDLP, which is capable of pruning at the hierarchical level depending on the client’s capability and model similarity to improve model efficiency and maintain model performance. This framework consists of two parts. First, pre-training customizes the initial pruning rate: We set the initial pruning rate for each layer according to the different capabilities of heterogeneous clients during the pre-training stage. Second, adaptively optimize the pruning rate: We use cosine similarity to quantify the contribution of each layer of the client model to the global model, thereby adaptively and dynamically optimizing the model pruning rate. Experimental results verify that our proposed method improves model efficiency by 2 to 3.5× compared to the state-of-the-art baselines, while achieving an accuracy difference of no more than 2% compared to the unpruned model.

The advent of smart devices has revolutionized our daily lives, especially through mobile sensing, with Human Activity Recognition (HAR) emerging as a key area of interest. Its applications span healthcare, indoor localization, and smart environments. Despite its promise, HAR's deployment faces numerous obstacles, such as personalization loss, scarce labeling, long training time, and poor edge device adaptation. Current research often addresses only a subset of these challenges, leaving a gap for a holistic solution. In response, we introduce Meta-Sense-Federated-Learning (MSFL), an innovative federated sensing system designed to personalize deep sensing models to individual users in a shorter time. MSFL leverages meta-learning to navigate data diversity, integrating physical principles to mitigate labeling biases and support real-world applications with minimal user input. Our framework synergizes a server-side model with mobile user engagement, where meta-learning crafts tasks that refine the base model and recalibrate it for novel scenarios. MSFL stands out by extracting uniform activity insights from varied Inertial Measurement Unit (IMU) data, employing data augmentation aligned with physical laws. This approach requires minimal calibration, addressing mobile sensing data's heterogeneity effectively. Users get their personalized model in a short time without having to spend extra time on the federal process. Our empirical findings underscore MSFL's superiority, showcasing a significant boost in model accuracy—over 10% higher than traditional techniques under conditions of limited labels, and 11% higher than MAML in the cross-dataset test. Furthermore, MSFL achieves local adaptation more rapidly, cutting down time expenditure by approximately 30%.

Asynchronous Federated Learning (AFL) is a promising technique to enable efficient and flexible distributed learning across heterogeneous devices. However, AFL faces the challenge of handling stragglers, i.e., devices that have high latency or low participation rate, which can degrade the convergence speed and accuracy of the global model. Existing solutions either discard or penalize the updates from stragglers, which may result in losing valuable information or introducing bias. In this paper, we propose FedVDA, a novel AFL framework that significantly improved the QoS of AFL in terms of model accuracy and training time cost by effectively utilizing and compensating for the updates from stragglers. Specifically, Fed-VDA introduced the dynamic window protocol that dynamically adjusts the server’s waiting time in each round based on the estimated completion time of the devices. We further designed a version control mechanism that corrects the stale gradients of the stragglers by supplementing the missing training rounds. Extensive experiments on three public datasets demonstrate that FedVDA achieves, on average, 2.7× faster convergence speed and 5.1% higher accuracy than state-of-the-art AFL methods. Moreover, we open-sourced FedVDA1and show that it is non-intrusive and highly scalable, which enables easy integration with other AFL algorithms and improves their QoS with no-pain in large-scale federated learning systems.

Mobile crowdsensing (MCS) has been applied for signal map construction in smart city. MCS leverages the mobility of users and the sensors embedded in mobile phones to collect and transfer sensing data. However, it is still costly for MCS to cover large-scale regions. Accordingly, data recovery algorithms are proposed, which allow participants to collect only few signal data and infer the rest by leveraging spatial-temporal correlation of signals. However, existing work only considered the temporal and 2-D spatial correlation in the plane, while the altitude dimension is not exploited. In this paper, we give an attempt to exploit the 3-D spatial-temporal correlation of signals to infer missing data and reconstruct ground signal map, where UAVs can be used to collect signals from the air. Two UAV-assisted ground signal map construction schemes are proposed based on propagation loss (PL) and convolution neural network (CNN). To further reduce the number of aerial samples required and reduce the cost incurred by UAV, a random-active sampling strategy is proposed to select more valuable signals in our work. Extensive simulations are performed which show that the proposed framework and schemes perform well under extremely high missing rate situations and outperform pure ground-based data recovery schemes. In addition, experiments for both indoor and outdoor are conducted to further verify the effectiveness of the proposed schemes.

As a privacy-preserving distributed learning paradigm, federated learning (FL) has been proven to be vulnerable to various attacks, among which backdoor attack is one of the toughest. In this attack, malicious users attempt to embed backdoor triggers into local models, resulting in the crafted inputs being misclassified as the targeted labels. To address such attack, several defense mechanisms are proposed, but may lose the effectiveness due to the following drawbacks. First, current methods heavily rely on massive labeled clean data, which is an impractical setting in FL. Moreover, an in-avoidable performance degradation usually occurs in the defensive procedure. To alleviate such concerns, we proposeBadCleaner, a lossless and efficient backdoor defense scheme via attention-based federated multi-teacher distillation. Firstly,BadCleanercan effectively tune the backdoored joint model without performance degradation, by distilling the in-depth knowledge from multiple teachers with only a small part of unlabeled clean data. Secondly, to fully eliminate the hidden backdoor patterns, we present an attention transfer method to alleviate the attention of models to the trigger regions. The extensive evaluation demonstrates thatBadCleanercan reduce the success rates of state-of-the-art backdoor attacks without compromising the model performance.

With an increasing number of mobile devices equipped with screens and cameras, screen-camera communication (SCC) systems enable data exchange between devices conveniently and efficiently. By encoding data with spatial and temporal diversity on a screen, multiple users with a camera can receive data without setting up a wireless network. However, as the transmitter pushes the limits of increasing throughput with a high display rate, the receiver actually suffers from a low goodput caused by composite frames. Those frames cannot be decoded correctly with existing methods. To address this problem, we propose a reliable data recovery scheme named RescQR. In RescQR, a mixture separation scheme coupled with a dedicated frame border is proposed to separate composite frames. A Viterbi-based data recovery scheme is proposed to recover data from blurred regions in composite frames. Additionally, an auto-configuration method with the help of a front camera is proposed to adjust parameters automatically according to the estimated distance between the screen and the camera. Our prototype and experiments demonstrate that RescQR achieves a data goodput of 400+kbps even with standard QR codes, which significantly outperforms previous solutions.

Narrowband Internet of Things (NB-IoT) is an low-power wide area network based on cellar architecture. The repetition scheme is a key solution to achieve enhanced coverage with low complexity in the uplink. However, the impact of the current repetition scheme on energy consumption and coverage performance of NB-IoT are still unclear. In this paper, we conduct field measurements of the repetition scheme in terms of energy efficiency. We find that most of repetition values configured by the eNodeB lead to non-optimal energy efficiency. Then we propose an adaptive repetition scheme based on a regression block delivery rate (BDR) model which can be derived from a theoretical model and a small number of measurements. We conduct simulations based on real-world measurement data. The results show that the proposed adaptive repetition scheme outperforms the default repetition scheme in both energy efficiency and data transmission rate.

Deep learning-based visual localization methods use convolutional neural networks to directly regress the position of a target. However, previous studies only consider localization in a single scene and neglect personalized localization in multiple scenes. Furthermore, changes in the scene result in a reduced accuracy due to the model’s lack of adaptability. Moreover, traditional centralized training methods pose data privacy concerns. In this paper, we propose a personalized semi-supervised federated learning framework with cloud-edge-end collaboration, called FedVL. The hierarchical architecture extends single-scene localization to multiple scenes, while the federated learning mechanism ensures data privacy. In this framework, we apply personalized federated learning to achieve scene-specific model and employ semi-supervised federated learning to allow the localization model to adapt to scene changes. Experiments conducted on indoor and outdoor datasets demonstrate the effectiveness of this approach.

Personalized federated learning (PFL) allows distributed clients and the server to share model parameters instead of raw data, aiming to customize a personalized model for each client. However, the naive design of weighted aggregation in previous studies can easily transfer the deviated sample knowledge to some local models, while ignoring the implicit relationship between the model layer and the training samples that match each other. Furthermore, PFL requires frequent parameter exchange between clients and the server to accurately obtain a personalized model suitable for its needs, which further exacerbates the problem of communication overhead. In this paper, we present a PFL framework, named DhcPFL, which is featured by the fine-grained observation that the updated parameters of each layer of the model implicitly provide information about the distribution of training samples, while achieving both improvements in model performance and communication cost. We manage to do these by innovating in the following aspects. Specifically, we propose a hierarchical clustering method based on Wasserstein distance for model parameters, which can fine-grained match clients with similar individual needs through the update of parameter layers. Based on this, we further propose a novel layered aggregation rule, which allows model parameters of partial layer aggregation instead of complete model parameters to be exchanged between client and server, effectively alleviating the problem of excessive communication overhead. Experiments on three public datasets demonstrate that our proposed method achieves about $2 \%$ performance improvement and $3.5 \times$ communication efficiency compared to the current baseline.

Keystroke snooping is an effective way to steal sensitive information from the victims. Recent research on acoustic emanation-based techniques has greatly improved the accessibility by non-professional adversaries. However, these approaches either require multiple smartphones or require specific placement of the smartphone relative to the keyboards, which tremendously restricts the application scenarios. In this article, we propose UltraSnoop, a training-free, transferable, and placement-agnostic scheme that manages to infer user’s input using a single smartphone placed within the range covered by a microphone and speaker. The innovation of Ultrasnoop is that we propose an ultrasonic anchor-keystroke positioning method and a Mel Frequency Cepstrum Coefficients clustering algorithm, synthesis of which could infer the relative position between the smartphone and the keyboard. Along with the keystroke time difference of arrival, our method could infer the keystrokes and even gradually improve the accuracy as the snooping proceeds. Our real-world experiments show that UltraSnoop could achieve more than 85% top-3 snooping accuracy when the smartphone is placed within the range of 30–60 cm from the keyboard.

With the dawn of 5G network, a new set of requirements for site spectrum monitoring, location-based services (LBS), network construction, and cellular planning are emerging, all of which are relying on fine-grained signal map. Although with significant importance, the traditional signal map construction, e.g., through full site survey, could be time-consuming and labor-intensive as the signal varies frequently over time and the accuracy requirement grows rapidly with the emergence of new applications. The state-of-the arts usually employ crowdsourcing scheme and matrix completion algorithm to solve the dilemma. However, the crowdsourcing scheme usually suffers from uneven distributed and inadequate participants, while the matrix completion methods do not take the specific signal map features into account, thus suffering from sub-optimal recovery results. To this end, in this paper, we study how to effectively reconstruct and update the signal map in the case of partially measured signal maps with smaller cost and propose a GAN-based active signal map reconstruction method (GSMAC). Our method is mainly innovative in two parts: GSMC, GAN-based signal map construction, and ACS, an active crowdsourcing scheme. Specifically, GSMC can effectively update the signal map with only a small number of observations while also fully using the incomplete historical signals to effectively update the signal map online. Meanwhile, ACS consists of a reinforce learning-based active query mechanism which quantitatively evaluates the most valuable measurement site for reconstruction, which further reduces the measurement cost to minimum. The simulation results and real implemented data driven experiments demonstrate the advantages and effectiveness of our approach in both accuracy and cost.

Mobile onboard target detection system with autonomous drone-assisted Internet of Things, due to its inherent agility and coverage-effective deployment, is beneficial in city management, ecosystem monitoring, etc. However, most advanced detection methods become inefficient or even malfunction, since the computation resources for online detection in such a high-altitude dynamic environment are far beyond the capability of the drone. To this end, we design and implement ODTDS—an online drone-based target detection system that performs online data processing and autonomous navigation simultaneously with restricted resources. Specifically, to prolong detection durations with limited energy providing for continuous processing and flying, we develop an adaptive motion planner for autonomous and energy-efficient navigation. Meanwhile, to perform online target detection from complex environments, we propose a hybrid method integrating feature pyramid feedback with the speed up robust features, to enhance the background information to achieve high accuracy on the restricted computation platform of the drone. Based on these two designs, our system can address resource-constrained challenges to accomplish detection missions autonomously. Finally, we implement an ODTDS prototype and evaluate it through outdoor high-altitude experiments. The results of various performance evaluations can demonstrate the effectiveness of our system.

Unlike the traditional dock-based systems, dockless bike-sharing systems are more convenient for users in terms of flexibility. However, the flexibility of these dockless systems comes at the cost of management and operation complexity. Indeed, the imbalanced and dynamic use of bikes leads to mandatory rebalancing operations, which impose a critical need for effective bike traffic flow prediction. While efforts have been made in developing traffic flow prediction models, existing approaches lack interpretability, and thus have limited value in practical deployment. To this end, we propose an Interpretable Bike Flow Prediction (IBFP) framework, which can provide effective bike flow prediction with interpretable traffic patterns. Specifically, by dividing the urban area into regions according to flow density, we first model the spatio-temporal bike flows between regions with graph regularized sparse representation, where graph Laplacian is used as a smooth operator to preserve the commonalities of the periodic data structure. Then, we extract traffic patterns from bike flows using subspace clustering with sparse representation to construct interpretable base matrices. Moreover, the bike flows can be predicted with the interpretable base matrices and learned parameters. Finally, experimental results on real-world data show the advantages of the IBFP method for flow prediction in dockless bike sharing systems. In addition, the interpretability of our flow pattern exploitation is further illustrated through a case study where IBFP provides valuable insights into bike flow analysis.

Potential crowd flow prediction for new planned transportation sites is a fundamental task for urban planners and administrators. Intuitively, the potential crowd flow of the new coming site can be implied by exploring the nearby sites. However, the transportation modes of nearby sites (e.g. bus stations, bicycle stations) might be different from the target site (e.g. subway station), which results in severe data scarcity issues. To this end, we propose a data-driven approach, named MOHER, to predict the potential crowd flow in a certain mode for a new planned site. Specifically, we first identify the neighbor regions of the target site by examining the geographical proximity as well as the urban function similarity. Then, to aggregate these heterogeneous relations, we devise a cross-mode relational GCN, a novel relation-specific transformation model, which can learn not only the correlation but also the differences between different transportation modes. Afterward, we design an aggregator for inductive potential flow representation. Finally, an LTSM module is used for sequential flow prediction. Extensive experiments on real-world data sets demonstrate the superiority of the MOHER framework compared with the state-of-the-art algorithms.

Recent research advancement of wireless sensing technology has made device-free interaction in the WiFi-enabled IoT environment possible. Although gesture-based interaction with such a smart environment greatly improves usability, it also introduces many security problems, such as shoulder surfing attacks. By spoofing the gestures of legitimate users, the attacker could easily access private information or services and cause even worse consequences. A secure interaction mechanism for this environment is required to prevent attackers without compromising the usability, while the limited recognition ability and low robustness of WiFi sensing make this target extremely challenging. To this end, we propose a secure interaction mechanism called secure interaction via WiFi Signal (SiWi), which provides the ability to resist shoulder surfing attacks without compromising the usability by using just WiFi signals. SiWi innovates in a concurrent interaction/authentication framework with only three elemental gestures (push, swing, and wave) and four types of identity-related imperceptible/hidden features (time distribution, direction, angle, and distance). HMM and Fresnel model-based algorithms are used to recognize the gestures and extract hidden features robustly and efficiently. Extensive experiments in a real implemented system were conducted to investigate the effectiveness of the proposed secure interaction system. The results show that our system can achieve an average accuracy of 93% to identify legitimate users and 97% to resist the spoofer.

Abstract Smartwatches have arguably become a popular wearable device nowadays. It is important to protect privacy data stored in smartwatches from being stolen. This study proposes a novel smartwatch user authentication technique based on the arm-raising gesture, which is the process of moving the arm from one side of the body to the chest height. We conducted two experiments to verify the effectiveness of the proposed technique. In Experiment 1, we investigated the performance of identifying users with the arm-raising gesture. We selected a set of features and applied them to five basic machine learning algorithms (i.e. random forest, simple logistic, naive Bayes, multilayer perceptron and linear classifier). Results with 32 participants show that with combined features, these classifiers generally achieved high authentication accuracy with high true accept rate (TAR) ($\geq $92.1% for random forest, simple logistic and multilayer perceptron), low false accept rate (FAR) ($\leq $0.6%) and large area under the curve (AUC) of receiver operating characteristics) ($\geq $92.4%). In Experiment 2, we examined the performance of identifying the arm-raising gesture across different day-to-day gestures. Results show that the arm-raising gesture can be identified from other eight common gestures with high TAR ($\geq $99.5%), low FAR ($\leq $3.6%) and large AUC ($\geq $99%). Overall, the results indicate that our technique could be a viable alternative for smartwatch user authentication.

Mobile crowdsensing (MCS) has been applied for signal map construction in smart city. However, it is still costly for MCS to cover large-scale regions. Accordingly some data recovery algorithms are proposed, which allow participants to collect only few signal data and infer the rest of missing data by leveraging spatial-temporal correlation of signals. However, existing work only considered the temporal and 2-D spatial correlation in the plane, while the altitude dimension is not exploited. In this paper, we give a first attempt to exploit the 3-D spatial-temporal correlation of signals to infer missing data in the ground and reconstruct ground signal map. An UAV-assisted ground signal map construction scheme is proposed based on matrix completion (MC). Specifically, UAVs can be used to collect signals in the air and aerial-ground signal mappings are performed to assist the ground signal inference. Extensive simulations are performed which show that the proposed scheme performs well under extremely high missing rate situations and outperform pure ground-based data recovery schemes.

With the rise of privacy concerns in traditional centralized machine learning services, the federated learning, which incorporates multiple participants to train a global model across their localized training data, has lately received signifi-cant attention in both industry and academia. However, recent researches reveal the inherent vulnerabilities of the federated learning for the membership inference attacks that the adversary could infer whether a given data record belongs to the model’s training set. Although the state-of-the-art techniques could successfully deduce the membership information from the centralized machine learning models, it is still challenging to infer the membership to a more confined level, user-level. In this paper, We propose a novel user-level inference attack mechanism in federated learning. Specifically, we first give a comprehensive analysis of active and targeted membership inference attacks in the context of the federated learning. Then, by considering a more complicated scenario that the adversary can only passively observe the updating models from different iterations, we incorporate the generative adversarial networks into our method, which can enrich the training set for the final membership inference model. The extensive experimental results demonstrate the effectiveness of our proposed attacking approach in the case of single-label and multi-label.

Among the numerous indoor localization technologies, image-based solution has great advantages on convenient access from smartphone and its infrastructure-less deployment. However, the image-based localization also suffers from two key disadvantages, which hinders the universal application. Firstly, it requires a large amount of computing and storage resources, which is difficult to achieve for the mobile device, while cloud-based scheme incurs unacceptable delay. Secondly, although this solution doesn't require infrastructure, it still suffers from labor intensive image-database construction and updates. To overcome these limitations, we propose an image-based indoor localization method featured with realtime localization and labor-less image database update. This method mainly innovates in two aspects. First, we propose a mobile device compatible image database compressing framework, which enable realtime and accurate on-device image searching even in a large scenario. Our localization method achieves resource efficiency (in terms of storage and processing) by only keeping image feature vectors, and employing the efficient k-mean tree to search for the best matched image. Secondly, to achieve labor-less image database updating, we mainly add high-quality and informative query image into the database. These query image could compensate the missing information or changed scenario in a up-to-date manner. We conduct real experiments in Android Platform to verify the feasibility and performance of the localization method. Experiment results show that our method has good accuracy (90% location errors are within 1.5m) and high real-time performance (average location delay is less than 0.5s).

With the explosive growth of massive data generated by smart Internet of Things (IoT) devices, federated learning has been envisioned as a promising technique to provide distributed machine learning services while protecting training data privacy. However, conventional federated learning protocols have shown significant drawbacks in regards of efficiency and scalability. First, since the synchronous communication model of federated learning and the computation capability of each device is different, the straggled users could severely desegregate the efficiency. Second, in synchronous communication, there is no effective client selection mechanism to make the model perform better in the early stage. Third, how to coordinate the communication of various nodes to accelerate global convergence is also one of the issues that need to be considered. To solve the above-mentioned problems, we propose a semi-asynchronous federated learning mechanism where a data expansion method is used to effectively reduce the stragglers existing in both synchronous and asynchronous communication models. Moreover, we also designed a priority function to make the accuracy increase rapidly in the early stage. Experimental results demonstrate that our proposed method have higher accuracy and faster convergence time compared with existing synchronization methods.

The Aerial Internet-of-Things (Aerial-IoT) systems, deploying sensors on high-altitude platforms, e.g., drones, parachutes, and aircrafts, are a crucial monitor due to its agile maneuverability and augmentation of observation, collection, and communication. As such, the measurement accuracy and requirements of Aerial-IoT are far beyond the ability of general commercial-off-the-shelf sensors, especially in the high-altitude environment, where environmental factors (air pressure, temperature, humidity, wind movement, etc.) tend to change rapidly and lead to highly deviated readings. In this article, we tackle this challenge. First, we introduce our designed measurement system for Aerial-IoT. Then, to compensate for the low data quality and calibrate the deviation data from sensors, we take into account the inherent correlations and interaction between sensor data and environmental factors, and construct a data calibration model, called data calibration based on the neural network (DC-NN). Finally, to illustrate the effectiveness of our system, we carry out a real-world implementation by deploying sensors on the surface of parachutes in a dynamic airdrop environment. Extensive experiments on temperature-humidity-material-tensile-testing (THMTT) and high-altitude airdrop are conducted to show the significant improvements of our proposed DC-NN model.

Edge computing is envisioned to be a powerful platform that provides efficient data storage and computation services in the smart Internet-of-Things (IoT) systems. In this data-intensive architecture, protecting user-side data privacy is one of the most critical concerns to prevent privacy leakage from any other untrusted entities. Aiming to resist this concern, many privacy-preserving data aggregation (PPDA) schemes have been proposed for various cloud-enabled IoT applications. However, due to the resource-constrained nature of the smart IoT devices, the conventional PPDA solutions, in terms of both privacy and performance requirements, are unsuitable in edge computing. To address this challenge, we propose a lightweight and verifiable PPDA scheme, named LVPDA, for the edge-computing-enabled IoT system, where the Paillier homomorphic encryption method and an online/offline signature technique are combined to ensure the privacy preserving and integrity verification during the data aggregation process. A detailed security analysis indicates that LVPDA is existentially unforgeable under the chosen message attack (EU-CMA) and the data integrity can be guaranteed with formal proof under q -strong Diffie-Hellman (q -SDH) assumptions. Compared with other PPDA methods, our scheme can achieve lightweight PPDA in terms of less computational complexity and communication overhead.

Indoor localization is of great significance to a wide range of applications in the era of mobile computing. The maturity of the computer vision techniques and the ubiquity of embedded sensors in commercial off-the-shelf (COTS) smartphones shed the light on the submeter localization services for indoor environment. The state-of-the-art indoor localization works suffer from high-cost deployment and inaccurate results due to the coarse readings from internal measurement units (IMUs) sensors in the smartphones. In this article, we mainly innovate in introducing the WiFi-sensing technology to extract the distance information in a low-cost and device-free manner. Along with the computer vision technology, we model and implement an accurate and easy-to-deploy system for indoor localization. This system enhances indoor localization with multimodal sensing via two images, IMU sensors reading and CSI of WiFi signal. Specifically, we first model and design camera-based, sensor and WiFi-assisted indoor localization and propose several algorithms in this model. We then implement a prototype with smartphones and commercial WiFi devices and evaluate it in several distinct indoor environments. The experimental results show that 92-percentile error is within 0.2 m for indoor targets which sheds light on submeter indoor localization.

As a decentralized ledger to record all transaction information, blockchain can be applied to address the security and privacy issues in mobile application system. We term the blockchain applied to mobile applications as mobile blockchain. The mining process in mobile blockchain requires high computing capacity and energy which could overwhelm that mobile devices can offer. In this case, mobile edge computing servers (MESs) can be involved to offer computation services to miners in mobile blockchain. Note that the resources of MESs are also limited, MESs could further request resources from the cloud computing server (CCS). Accordingly, in this paper, both mobile edge computing and cloud computing are considered to support the mobile blockchain applications which makes the problem a hierarchical one. Naturally, the issue of hierarchical resource allocation arises. And a hierarchical combinatorial auction model is proposed to solve this problem, based on which an efficient and truthful framework is provided. Specifically, we formulate winner determination problems (WDPs) for mobile edge computing service providers and cloud computing service provider, and computationally tractable algorithms to address both problems are proposed. Finally, numerical analysis shows the effectiveness of the proposed scheme.

Signal map is of great importance, especially in the dawn of 5G network, for site spectrum monitoring, location-based services (LBS), network construction, and cellular planning. Despite its significance, the traditional signal map construction, e.g., through full site survey, could be time-consuming and labor-intensive as the signal varies frequently over time and the accuracy requirement grows rapidly with the emergence of new applications. Even with crowdsourcing scheme, the participants tend to be unevenly distributed in space while the encouragement budgets for the participants could be far from enough to collect adequate high-quality measurements. Therefore, the signal map constructed by crowdsourcing is often sparse and incomplete. To this end, in this paper, we study how to effectively reconstruct and update the signal map in the case of partially measured signal maps with minimum cost and propose an auto-encoder-based active signal map reconstruction method (AER). Our method is mainly innovative in three parts. Firstly, AER can effectively update the signal map with only a small number of observations while also fully using the incomplete historical signals to effectively update the signal map online. Secondly, AER consists of an active query mechanism which quantitatively evaluates the most valuable measurement site for reconstruction, which further reduces the measurement cost to a large extent. Thirdly, to cope with the measurement dynamics, we give a new signal map model describing not only the signal strength but also the signal dynamics, based on which an advanced AER algorithm is proposed. The simulation results demonstrate the advantages and effectiveness of our approach in both accuracy and cost.

Wireless virtualization has become a key concept in future cellular networks which can provide multiple virtualized wireless networks for different mobile virtual network operators (MVNOs) over the same physical infrastructure. Resource allocation problem is a main challenge for wireless virtualization for which auction approaches have been widely used. However, for most existing auction-based allocation schemes, the objective is to maximize the social welfare (i.e., the sum of all valuations of winning bidders) due to its simplicity. While in reality, MVNOs are more interested in maximizing their own revenues. However, the revenue-maximization auction problem is much more complex since the price is unknown before calculation. In this paper, we give a first attempt for designing a revenueoptimal auction mechanism for resource allocation in wireless virtualization. Considering the complexity in revenue maximization, we apply the deep learning techniques. Specifically, we construct a multi-layer feed-forward neural network based on the analysis of optimal auction design. The neural network adopts users' bids as the input and the allocation rule and conditional payment rule for the users as the output. The training set of this neural network is the users' valuation profiles. The proposed auction mechanism possesses several satisfactory properties, e.g., individual rationality and incentive compatibility. Finally, simulation results demonstrate the effectiveness of the proposed scheme.

Wireless energy transfer has been widely studied in recent decades, with existing works mainly focused on maximizing network lifetime, optimizing charging efficiency, and optimizing charging quality. All these works use a charging model with the linear superposition, which may not be the most accurate. We apply a nonlinear superposition model, and we consider the Fast Charging Scheduling problem (FCS): Given multiple chargers and a group of sensors, how can the chargers be optimally scheduled over the time dimension so that the total charging time is minimized and each sensor has at least energy E ? We prove that FCS is NP-complete and propose a 2-approximation algorithm to solve it in one-dimensional (1D) line. In a 2D plane, we first consider a special case of FCS, where the initial phases of all chargers are the same, and propose an algorithm to solve it, which has a bound. Then we propose an algorithm to solve FCS in a general 2D plane. Unlike other algorithms, our algorithm does not need to calculate the combined energy of every possible combination of chargers in advance, which greatly reduces the complexity. Extensive simulations demonstrate that the performance of our algorithm performs almost as good as the optimal algorithm.

Detection of moving targets is a basic but challenging function of drone-based surveillance systems (DBSSs), which could give rise to various potential applications in smart city and intelligent transportation. However, how to detect moving targets in the dynamic high-altitude environment with restricted computational resources is one of the most critical ones. Furthermore, during the detection, the moving target could travel in dynamic speed and be blocked by dense-obstructer, such as woods and buildings. In this paper, we develop an online drone-based moving target detection (ODMTD)system, which performs moving target detection in dense-obstructer areas. Specifically, first, our proposed system simultaneously performs adaptive path planning and autonomous drone flight via the combination of historical path cost and energy loss computation by using drone attitudes. Second, to detect moving targets in the dynamic background, we develop an algorithm of combing the speed up robust features and approximate nearest neighbors, shortly SURF-ANN, for estimating and compensating the global motion of the background. Finally, in order to calibrate distorted images taken by the camera obliquely, we utilize perspective transformation to remap images into another plane, and then detect moving targets by subtracting registered images (SRI). Furthermore, real-time outdoor high-altitude experiments, by comprising with the state-of-art methods, demonstrate the effectiveness of our ODMTD system.

One of the major challenges in indoor localization is the matching difficulty and prediction accuracy of anchor points. In this work, we innovate in proposing a camera-based, sensor- and WiFi-assisted, and easy-to-deploy system for localization. The proposed method is based on muliti-modal sensing to enhancing localization measurement. We implement a prototype with smartphones and commercial WiFi devices and evaluate it in distinct indoor environments. Experimental results show that the 85-percentile error is within 0.21m for indoor POIs that sheds light on sub-meter level localization.

Edge computing has emerged as the key enabling technology that empowers the IoT with intelligence and efficiency. In this data enriched infrastructure, privacy-preserving data aggregation (PPDA) is one of the most critical services. However, the security and privacy-preserving requirements and online computational cost still present practical concerns in edge computing for resource-constraint edge terminals. To cope with this challenge, we present a lightweight privacy-preserving data aggregation scheme named LPDA-EC for edge computing system by employing the online/offline signature technique, Paillier homomorphic cryptosystem, and double trapdoor Chameleon hash function in this paper. The proposed LPDA-EC scheme can achieve data confidentiality and privacy-preserving, ensuring that the edge server and control center are agnostic of the user's private information during the whole aggregation process. Through detailed analysis, we demonstrate that our scheme is existentially unforgeable under chosen message attack (EU-CMA) and ensures data integrity with formal proofs under q-Strong Diffie-Hellman (q-SDH) assumptions. Numerical results indicate that the LPDA-EC scheme has less computational and communication overheads.

With the dense deployment of wireless infrastructure such as radio towers and WiFi access points, wireless network optimization becomes very important for improving the network capacity and enhancing the communication quality of wireless links. Most optimization algorithms rely on the conflict graph to describe the interference situation. However, building a conflict graph requires exhaustive measurements of the whole network and the existing estimation approaches are static and inaccurate. In this paper, we propose a conflict graph embedding approach to assess network interference situations by representing the wireless nodes with low-dimensional vectors while preserving their conflict relationships. Specifically, our approach introduces a sliding-window based partial measurement method to sample the interference graph in the network, then adopts a learning algorithm to obtain the vector representation of the nodes, and then infers the interference situations by exploring the feature vectors. The proposed approach has been proved to be low measurement overhead, low computational cost, and self-adaptive, which is suitable for large-scale dynamic wireless networks. We illustrate that conflict graph embedding can be used for interference-aware wireless network optimizations. We conduct extensive experiments based on real wireless network datasets, which show the efficiency of the proposed approach.

The receiving signal strength (RSS) is crucial for the Internet of Things (IoT), as it is the key foundation for communication resource allocation, localization, interference management, sensing, and so on. Aside from its significance, the measurement process could be tedious, time consuming, inaccurate, and involving human operations. The state-of-the-art works usually applied the fashion of “measure a few, predict many,” which use measurement calibrated models to generate the RSS for the whole networks. However, this kind of methods still cannot provide accurate results in a short duration with low measurement cost. In addition, they also require careful scheduling of the measurement which is vulnerable to measurement conflict. In this paper, we propose a compressive sensing- (CS-) based RSS measurement solution, which is conflict-tolerant, time-efficient, and accuracy-guaranteed without any model-calibrate operation. The CS-based solution takes advantage of compressive sensing theory to enable simultaneous measurement in the same channel, which reduces the time cost to the level of O(log⁡N) (where N is the network size) and works well for sparse networks. Extensive experiments based on real data trace are conducted to show the efficiency of the proposed solutions.

A traditional approach to solving the large delay problem of campus networks is to upgrade the link connecting the gateway to the Internet. Inspired by the emerging content-centric network (CCN) and software defined network (SDN) architecture, we propose an alternative solution where the campus network uses a few CCN routers with caching ability, so that duplicate requests for the same content can be satisfied locally without traffic from the Internet. In our solution, we formulate the problem of deciding the cached content at each router to minimize the total delay of all requests. We prove that the problem is NP-hard, and no polynomial time algorithm can provide a constant approximation ratio, unless P=NP. We then propose an exponential-time exact algorithm and three polynomial-time heuristic algorithms. Numerical results show that our solution can reduce network delay significantly, compared to existing cache decision algorithms.

Conflict graph has been widely used for wireless network optimization in dealing with the issues of channel assignment, spectrum allocation, links scheduling and etc. Despite its simplicity, the traditional conflict graph suffers from two drawbacks. On one hand, it is a rough representation of the interference condition, which is inaccurate and will cause suboptimal results for wireless network optimization. On the other hand, it only defines the interference between two entities, which neglects the accumulative effect of small amount interference. In this paper, we propose the model of quantized conflict graph (QCG) to tackle the above issues. The properties, usage and construction methods of QCG are explored. We show that in its matrix form, a QCG owns the properties of low-rank and high-similarity. These properties give birth to three complementary QCG estimation strategies, namely low-rank approximation approach, similarity based approach, and comprehensive approach, to construct the QCG efficiently and accurately from partial interference measurement results. We further explore the potential of QCG for wireless network optimization by applying QCG in minimizing the total network interference. Extensive experiments using real collected wireless network are conducted to evaluate the system performance, which confirm the efficiency of the proposed algorithms.

Collecting the RSS between all pair of nodes in the networks is very significant for wireless network optimization, localization, interference management and etc. Aside from its significances, the measurement process could be tedious, time consuming and involving human operations. The state-of-art works usually applied the fashion of “measure a few, predict many”, which use measurement calibrated models to generate the RSS for the whole networks. However, this kind of methods still cannot provide accurate results in a short duration and low measurement cost. In addition, they also require careful scheduling of the measurement which is vulnerable to measurement conflict. In this paper, we propose a compressive sensing (CS)-based RSS measurement solution, which is conflict-tolerant, time-efficient and accuracy-guaranteed without any model-calibrate operation. The CS-based solution takes advantage of compressive sensing theory to enable simultaneous measurement in the same channel, which reduces the time cost to the level of O(logN) (where N is the network size) and works well for sparse networks. Extensive experiments based on real data trace are conducted to show the efficiency of the proposed solutions.

Cognitive radio networks (CRNs) promise to enable the next generation of communication networks. The channel assignment (CA) problem is one of the most important issues in CRNs. In this paper, our goal is to design highly efficient and localized protocols for CA. In addition, we want to maximize node connectivity after CA, which is important for packet delivery. To this end, we design two basic algorithms and an advanced algorithm framework. Within this framework, we can change the edge priority in CA to meet different requirements. Simulation results show that the proposed framework is fast (two rounds of communication among nodes, regardless of network size) and outperforms an existing method.

Recently, the SINR-model has been widely utilized in link scheduling, spectrum allocation and other applications. The SINR model requires the receiving power information of all potential link peers, which is usually assumed to be known as priori or following a uniform propagation model. We have performed experiments to illustrate how real power data could improve the performance of the SINR-based applications with considerable margin. Thus, obtaining the real power data through measurements is promising. However, this method faces many challenges. We propose a pathloss model based solution, including a representative link selection method to cut down the measurement pairs; accuracy control to determine the sample size; and a measurement distribution method to shorten the measurement duration. Our experiments show that our solution significantly improves the SINR-based scheduling's performance.

Interactive wireless applications, like VoIP over wireless networks, desire high-quality links and smooth connectivity during user movement. In order to support seamless roaming in wireless networks, handoff optimization has attracted a lot of attention recently. Most existing approaches aim at reducing handoff latency in communication protocols. While these methods provide significant savings in handoff latency, frequent handoffs could still be crucial and problematic for interactive applications. In this paper, we propose a new perspective for handoff optimization by introducing navigation guidance to minimize the handoff frequency. We first formulate the navigation-driven handoff minimization problem, then propose an optimal algorithm and a localized algorithm to solve it. The optimal algorithm assumes global knowledge of AP locations and uses a navigation graph to find a minimal handoff frequency path. The localized algorithm, however, only uses neighbor AP locations for route selection, which is more practical in real applications. Implementation issues of the proposed algorithms are discussed and simulations based on real world AP deployment are used to evaluate their performance. Experiment results show that our algorithms reduce handoff frequency by at most 42% compared to existing strategies.