The spatial features of emitted wireless signals are the basis of location distinction and determination for wireless indoor localization. Available in mainstream wireless signal measurements, the Received Signal Strength Indicator (RSSI) has been adopted in vast indoor localization systems. However, it suffers from dramatic performance degradation in complex situations due to multipath fading and temporal dynamics. Break-through techniques resort to finer-grained wireless channel measurement than RSSI. Different from RSSI, the PHY layer power feature, channel response, is able to discriminate multipath characteristics, and thus holds the potential for the convergence of accurate and pervasive indoor localization. Channel State Information (CSI, reflecting channel response in 802.11 a/g/n) has attracted many research efforts and some pioneer works have demonstrated submeter or even centimeter-level accuracy. In this article, we survey this new trend of channel response in localization. The differences between CSI and RSSI are highlighted with respect to network layering, time resolution, frequency resolution, stability, and accessibility. Furthermore, we investigate a large body of recent works and classify them overall into three categories according to how to use CSI. For each category, we emphasize the basic principles and address future directions of research in this new and largely open area.

From RSSI to CSI: Indoor Localization via Channel Response, A survey on indoor localization using PHY-layer information

A number of online services nowadays rely upon machine learning to extract valuable information from data collected in the wild. This exposes learning algorithms to the threat of data poisoning, i.e., a coordinate attack in which a fraction of the training data is controlled by the attacker and manipulated to subvert the learning process. To date, these attacks have been devised only against a limited class of binary learning algorithms, due to the inherent complexity of the gradient-based procedure used to optimize the poisoning points (a.k.a. adversarial training examples). In this work, we rst extend the de nition of poisoning attacks to multiclass problems. We then propose a novel poisoning algorithm based on the idea of back-gradient optimization, i.e., to compute the gradient of interest through automatic di erentiation, while also reversing the learning procedure to drastically reduce the attack complexity. Compared to current poisoning strategies, our approach is able to target a wider class of learning algorithms, trained with gradient- based procedures, including neural networks and deep learning architectures. We empirically evaluate its e ectiveness on several application examples, including spam ltering, malware detection, and handwritten digit recognition. We nally show that, similarly to adversarial test examples, adversarial training examples can also be transferred across di erent learning algorithms.

Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization

iRobot-Factory: An intelligent robot factory based on cognitive manufacturing and edge computing

The enabling Internet-of-Things (IoT) technology has inspired a large number of novel platforms and applications. One popular IoT platform is unmanned aerial vehicles (UAVs, also known as drone). Benefiting from the intrinsic flexibility, convenience, and low cost, UAVs have great potentials to be utilized in various civil applications, including parcel delivery. However, suffering from limited payloads and battery capacities, it is uneconomical for UAVs to perform parcel delivery tasks independently. To conquer the drawbacks of low payloads and battery capacities, people propose to employ both trucks and drones to construct truck-drone parcel delivery systems. However, previous works only leverage either independent drones or truck-carried drones to collaborate with trucks. In contrast, in this article we propose to simultaneously employ trucks, truck-carried drones, and independent drones to construct a more efficient truck-drone parcel delivery system. We claim that such a hybrid parcel delivery system can fully exploit the complementary benefits of the three platforms. We propose a novel routing and scheduling algorithm, referred to as hybrid truck-drone delivery (HTDD) algorithm, to solve the hybrid parcel delivery problem, wherein $M$ drones carried by $M$ trucks, together with $N$ independent drones, cooperate to deliver parcels to customers distributed in a wide region. The experimental results show that our algorithm outperforms the existing solutions which employ either independent drones or truck-carried drones.

Routing and Scheduling for Hybrid Truck-Drone Collaborative Parcel Delivery With Independent and Truck-Carried Drones

Recently, device-free human behavior recognition has become a hot research topic and has achieved significant progress in the field of ubiquitous computing. Among various implementation, behavior recognition based on WiFi CSI (channel state information) has drawn wide attention due to its major advantages. This paper investigates more than 100 latest CSI based behavior recognition applications within the last 6 years and presents a comprehensive survey from every aspect of human behavior recognition. Firstly, this paper reviews general behavior recognition applications using the WiFi signal and presents the basic concept of CSI and the fundamental principle of CSI-based behavior recognition. This paper analyzes the key components and core characteristics of the system architecture of human behavior recognition using CSI. Afterward, we divide the sensing procedures into many steps and summarize the typical studies from these steps, including base signal selection, signal preprocessing, and identification approaches. Next, based on the recognition technique, we classify the applications into three groups, including pattern-based, model-based, and deep learning-based approach. In every group, we categorize the state-of-the-art applications into three groups, including coarse-grained specific behavior recognition, fine-grained specific behavior recognition, and activity inference. It elaborates the typical behavior recognition applications from five aspects, including experimental equipment, experimental scenario, behavior, classifier, and system performance. Then, this paper presents comprehensive discussions about representative applications from the implementation view and outlines the major consideration when developing a recognition system. Finally, this article concludes by analyzing the open issues of CSI-based behavior recognition applications and pointing out future research directions.

/pdf/a-survey-on-human-behavior-recognition-using-channel-state-17xs5detky.pdf

A Survey on Human Behavior Recognition Using Channel State Information

Conventional sensing methodologies for smart home are known to be labor-intensive and complicated for practical deployment. Thus, researchers are resorting to alternative sensing mechanisms. Wi-Fi is one of the key technologies that enable connectivity for smart home services. Apart from its primary use for communication, Wi-Fi signal has now been widely leveraged for various sensing tasks, such as gesture recognition and fall detection, due to its sensitivity to environmental dynamics. Building smart home based on Wi-Fi sensing is cost-effective, non-invasive, and enjoys convenient deployment. In this paper, we survey the recent advances in the smart home systems based on the Wi-Fi sensing, mainly in such areas as health monitoring, gesture recognition, contextual information acquisition, and authentication.

http://summit.sfu.ca/system/files/iritems1/19643/Smart%20Home%20Based%20on%20WiFi%20Sensing_%20A%20Survey%20-%20IEEE%20Journals%20%2526%20Magazine.pdf

Smart Home Based on WiFi Sensing: A Survey

The 5G networks are broadly characterized by three unique features: ubiquitous connectivity, extremely low latency, and extraordinary high-speed data transfer. The challenge of 5G is to assure the network performance and different quality of service (QoS) requirements of different services, such as machine type communication (MTC), enhanced mobile broad band (eMBB), and ultra-reliable low latency communications (URLLC) over 5G networks. Unlike the previous ”one size fits all” system, the softwarization, slicing and network capability exposure of 5G provide dynamic programming capabilities for QoS assurance. With the increasing complexity and dynamics of the network behaviors, it is non-trivial for a programmer to develop traditional software codes to schedule the network resources based on expert knowledge, especially when there is no quantitative relationship among the network events and the QoS anomalies. Machine learning is a computer technology that gives computer systems the ability to learn with data and improve performance and accuracy of decision making on a specific task, without being explicitly programmed. The areas of machine learning and communication technology are converging. Supervised learning based QoS assurance architecture for 5G networks was proposed in this paper. The supervised machine learning mechanisms can intelligently learn the network environment and react to dynamic situations. They can learn from the fore passed QoS related information and anomalies, and further reconstruct the relationship between the fore passed data and the current QoS related anomalies automatically and accurately. They, then, can trigger automatic mitigation or provide suggestions. The supervised machine learning mechanisms can also predict future QoS related anomalies with high confidence. In this paper, a case study for QoS anomaly root cause tracking based on decision tree was given to validate the proposed framework architecture.

/pdf/a-supervised-learning-based-qos-assurance-architecture-for-582pz08uz9.pdf

A Supervised Learning Based QoS Assurance Architecture for 5G Networks

Federated learning (FL) has recently emerged as a promising distributed machine learning (ML) paradigm. Practical needs of the "right to be forgotten" and countering data poisoning attacks call for efficient techniques that can remove, or unlearn, specific training data from the trained FL model. Existing unlearning techniques in the context of ML, however, are no longer in effect for FL, mainly due to the inherent distinction in the way how FL and ML learn from data. Therefore, how to enable efficient data removal from FL models remains largely under-explored. In this paper, we take the first step to fill this gap by presenting FedEraser, the first federated unlearning method-ology that can eliminate the influence of a federated client’s data on the global FL model while significantly reducing the time used for constructing the unlearned FL model. The basic idea of FedEraser is to trade the central server’s storage for unlearned model’s construction time, where FedEraser reconstructs the unlearned model by leveraging the historical parameter updates of federated clients that have been retained at the central server during the training process of FL. A novel calibration method is further developed to calibrate the retained updates, which are further used to promptly construct the unlearned model, yielding a significant speed-up to the reconstruction of the unlearned model while maintaining the model efficacy. Experiments on four realistic datasets demonstrate the effectiveness of FedEraser, with an expected speed-up of 4× compared with retraining from the scratch. We envision our work as an early step in FL towards compliance with legal and ethical criteria in a fair and transparent manner.

FedEraser: Enabling Efficient Client-Level Data Removal from Federated Learning Models

Recently, edge computing has attracted significant interest due to its ability to extend cloud computing utilities and services to the network edge with low response times and communication costs. In general, edge computing requires mobile users to upload their raw data to a centralized data server for further processing. However, these data usually contain sensitive information about mobile users that the users do not want to reveal, such as sexual orientation, political stance, health status, and service access history. The transmission of user data increases the leakage risk of data privacy since many extra devices can get access to these data. In this article, we attempt to keep the data of edge devices and end users on their local storage to resist the leakage of user privacy. To this end, we integrate federated learning and edge computing to propose P2FEC, a privacy-preserving framework that can construct a unified deep learning model across multiple users or devices without uploading their data to a centralized server. Furthermore, we use membership inference attacks as a case study for the privacy analysis of edge computing. The experiments show that the model constructed by our framework can achieve similar prediction performance and stricter protection of data privacy, compared to the model trained by standard edge computing.

Keep Your Data Locally: Federated-Learning-Based Data Privacy Preservation in Edge Computing

Internet of Things (IoT) applications bring in a great convenience for human’s life, but users’ data privacy concern is the major barrier toward the development of IoT. ${k}$ -anonymity is a method to protect users’ data privacy, but it is presently known to suffer from inference attacks. Thus far, existing work only relies on a number of experimental examples to validate ${k}$ -anonymity’s performance against inference attacks, and thereby lacks of a theoretical guarantee. To tackle this issue, in this paper we propose the first theoretical foundation that gives a nonasymptotic bound on the performance of ${k}$ -anonymity against inference attacks, taking into consideration of adversaries’ background information. The main idea is to first quantify adversaries’ background information, and from the point of the view of adversaries, classify users’ data into four kinds: 1) independent with unknown data values; 2) local dependent with unknown data values; 3) independent with certain known data values; and 4) local dependent with certain known data values. We then move one step further, theoretically proving the bound on the performance of ${k}$ -anonymity corresponding to each of the four kinds of users’ data through cooperating with the noiseless privacy. We argue that such a theoretical foundation links ${k}$ -anonymity with noiseless privacy, theoretically proving ${k}$ -anonymity provides noiseless privacy. Additionally, this paper theoretically explains why ${k}$ -anonymity is vulnerable to inference attacks using the modified Stein method. Simulations on real check-in dataset from the location-based social network have validated our results. We believe that this paper can bridge the gap between design and evaluation, enabling a designer to construct a more practical ${k}$ -anonymity technique in real-life scenarios to resist inference attacks.

Yang Yang

Papers

Smart Home Based on WiFi Sensing: A Survey

A Supervised Learning Based QoS Assurance Architecture for 5G Networks

FedEraser: Enabling Efficient Client-Level Data Removal from Federated Learning Models

Keep Your Data Locally: Federated-Learning-Based Data Privacy Preservation in Edge Computing

On the Performance of $k$ -Anonymity Against Inference Attacks With Background Information