Federated learning (FL) has been widely adopted to train machine learning models over massive data in edge computing. However, machine learning faces critical challenges, e.g., data imbalance, edge dynamics, and resource constraints, in edge computing. The existing FL solutions cannot well cope with data imbalance or edge dynamics, and may cause high resource cost. In this paper, we propose an adaptive asynchronous federated learning (AAFL) mechanism. To deal with edge dynamics, a certain fraction $\alpha$ of all local updates will be aggregated by their arrival order at the parameter server in each epoch. Moreover, the system can intelligently vary the number of local updated models for global model aggregation in different epochs with network situations. We then propose experience-driven algorithms based on deep reinforcement learning (DRL) to adaptively determine the optimal value of $\alpha$ in each epoch for two cases of AAFL, single learning task and multiple learning tasks, so as to achieve less completion time of training under resource constraints. Extensive experiments on the classical models and datasets show high effectiveness of the proposed algorithms. Specifically, AAFL can reduce the completion time by about 70 percent and improve the learning accuracy by about 28 percent under resource constraints, compared with the state-of-the-art solutions.

WiFi connection has been suffering from MAC forgery attacks due to the loose authentication mechanism between access points (APs) and clients. To address this problem, the physical (PHY) layer information-based fingerprint has been adopted for safe WiFi authentication. Since such a fingerprint is constant and unique for each specific network interface card (NIC), it can effectively prevent MAC forgery attacks. However, the PHY layer information-based fingerprint is still vulnerable to malicious attacks as it is extracted from Channel State Information (CSI), and its stability can be affected by the wireless environment. In this paper, we propose a novel undetectable attack framework, called PhyFinAtt, base on which the attacker can undermine the stability of the PHY layer-based authentication fingerprints through human movement and further attack the WiFi authentication protocols. Specifically, we first demonstrate that human movement at a designated location can affect the PHY fingerprint. We then illustrate the impact of human movement on the PHY fingerprint and the relationship between the movement and the channel quality to ensure that the PHY fingerprint is destroyed by the movement in an undetected way without affecting normal communication. Extensive experiments in real-world scenarios show that our proposed attack can effectively disrupt the stability of the PHY fingerprints and significantly degrade the performance of the authentication protocols based on such fingerprints. To the best of our knowledge, this is the first study on effective attacks against the PHY information-based WiFi authentication protocols. Furthermore, we also present a practical defense mechanism without involving any additional equipment to mitigate attacks similar to PhyFinAtt.

Federated learning (FL) has been widely adopted to train machine learning models over massive data in edge computing. Most works of FL employ pre-defined model architectures on all participating clients for model training. However, these pre-defined architectures may not be the optimal choice for the FL setting since manually designing a high-performance neural architecture is complicated and burdensome with intense human expertise and effort, which easily makes the model training fall into the local suboptimal solution. To this end, Neural Architecture Search (NAS) has been applied to FL to address this critical issue. Unfortunately, the search space of existing federated NAS approaches is extraordinarily large, resulting in unacceptable completion time on the resource-constrained edge clients, especially under the non-independent and identically distributed (non-IID) setting. In order to remedy this, we propose a novel framework, called FINCH, which adopts hierarchical neural architecture search to enhance federated learning. In FINCH, we first divide the clients into several clusters according to the data distribution. Then, some subnets are sampled from a pre-trained supernet and allocated to the specific client clusters for searching the optimal model architecture in parallel, so as to significantly accelerate the process of model searching and training. The extensive experimental results demonstrate the high effectiveness of our proposed framework. Specifically, FINCH can reduce the completion time by about 30.6%, and achieve an average accuracy improvement of around 9.8% compared with the baselines.

As an important technology in the field of Internet of Things, activity recognition is a well-researched topic. Due to its characteristics of no-invasion, no privacy leakage, good user experience, and high recognition accuracy rate, Channel State Information (CSI)-based method is one of the most widely used approaches for the activity recognition problem. However, the commodity WiFi devices for identification are usually in a complex electromagnetic environment, where the interference caused by WiFi devices from channel overlap is common and severe. Furthermore, our extensive experiments show that the performance of the state-of-the-art recognition systems may degrade significantly with co-channel interference (CCI). To solve the above problem, we propose WiAnti, a CSI-based activity recognition system which is robust to CCI. Two adaptive subcarrier selection algorithms, WiAnti-Pearson and WiAnti-DTW, are proposed to mitigate the impact of CCI and to improve the activity recognition performance. As demonstrated in the experimental results, WiAnti-Pearson yields 95% activity recognition accuracy rate on average, which can improve up to 14% recognition accuracy rate in the presence of constant CCI. Moreover, WiAnti-DTW achieves 8% higher recognition accuracy rate in the scenario with varying CCI, reaching 94%.

The broad spectrum of applications of WiFi sensing technology, like gait and gesture recognition, has received widespread attention in recent years. Though most WiFi sensing systems may achieve impressive performance, the challenge lies in making good use of the amplitude and phase information of the Channel State Information (CSI) retrieved from commodity WiFi devices to carry out sensing tasks. To address this issue, we develop an attention-based framework to properly track the importance of amplitude and phase information to adaptively extract distinguishing features related to gestures. Specifically, we first use the CSI ratio instead of the original CSI as the basic signal, which not only eliminates most of the noise, but also contains the complete information of the CSI signal corresponding to human motion. Then, we use the self-attention module to learn the coarse attention weights of amplitude and phase information of the CSI ratio. Moreover, the relation-attention module is used to integrate features to further refine the attention weight. In this way, we proposed a framework that can adaptively learn distinctive feature representations and thus facilitate ubiquitous gesture recognition. Extensive experiments demonstrate the effectiveness of method for gesture recognition under various conditions on the open Widar3.0 dataset. The proposed method achieves 99.69% in-domain recognition accuracy, 96.95% cross-location recognition accuracy and 93.71% cross-orientation recognition accuracy, outperforming the state-of-the-art solutions.