Cloud computing built on virtualization technologies can provide Internet service providers (SPs) with elastic virtualized node and link resources. SPs can outsource their virtualized resources as customized virtual networks (VNs) to end users. Hence, how to efficiently embed these VNs is the core issue in virtualization research. This technical issue is virtual network embedding (VNE). Since the issue inception, multiple mapping algorithms have been studied, including the reinforcement learning (RL) approach of machine learning. However, prior mapping algorithms are mostly static. Existing dynamic mapping algorithms just focus on accepting as many VNs as possible. No existing dynamic algorithm considers optimizing the quality of service (QoS) performance of each accepted VN. Optimizing the VN QoS performance is beneficial to guaranteeing service quality in cloud computing environment. On these backgrounds, we jointly investigate the dynamic VN embedding and optimize the QoS performance of each accepted VN. A dynamic heuristic algorithm is proposed in order to be evaluated in continuous time. When one VN service is requested, the VN will be mapped by the dynamic heuristic algorithm. If the QoS demand of the VN is not guaranteed, the re-embedding scheme of the heuristic algorithm will be driven. Certain virtual elements of the VN will be adjusted. The dynamic embedding algorithm ensures flexible VN assignment and fulfills customized QoS demands. Finally, simulation results are illustrated in order to validate the strength of our dynamic algorithm. We perform the comparison with multiple existing dynamic algorithms. For instance, VN acceptance ratio of our dynamic heuristic algorithm improves at least 13%.

In wireless device-to-device (D2D) networks, devices are reluctant to forward packets because of limited energy and possible delays for their own data. The incentive mechanisms that motivate devices to constitute direct communication for wireless multimedia quality optimality in D2D systems have been overlooked in the past. In this paper, we propose a new low-complexity distributed game-theoretic source selection and power control scheme that enhances the multimedia transmission quality with latency constraints. This approach has two major contributions. First, the proposed approach optimally selects the most beneficial source devices by analyzing the interactions between the base station's (BS's) rewarding strategies (denoted by price) and the devices' contributing behaviors (denoted by transmission power) using a Stackelberg game model. Second, optimal transmission power is adjusted for each selected source device in D2D networks by deriving Stackelberg equilibrium, wherein the BS and the device both achieve maximum utility. Computer simulations demonstrate that significant improvement in D2D multimedia transmission quality can be obtained by deploying the proposed scheme.

Reconfigurable intelligent surface (RIS) is considered as a promising technology, which can effectively control the phase of signals with low-cost reflective elements. In this paper, we investigate the RIS-assisted D2D communication system under Nakagami-m fading. Overlay mode and underlay mode are discussed respectively. To characterize the performance of proposed communication system, we propose novel closed-form analytical expressions for outage probability under two modes, respectively. Based on the analytical results, we derive the diversity order of outage probability. We show that under overlay D2D mode, the diversity order depends on the number of elements and fading parameters. For the underlay D2D case, 0 diversity order is obtained due to the existence of interference. Furthermore, we study the ergodic achievable rate for each mode and the corresponding tight bounds are derived in closed form. High SNR slopes achieved by users under two modes are also discussed. We observe a high SNR slope of order 1 for overlay D2D mode and 0 for underlay D2D mode. In addition, we also investigate the energy efficiency (EE) of the considered system. Numerical results show that the optimal phase shift proposed in our paper enhance system performance effectively comparing to the traditional D2D communication and random phase shift scheme. Significant EE gains can be obtained for low to middle SNR region by passive reflectors with reconfigurable phase. Unlike relay system, the nearer RIS to the D2D transmitter/receiver, the better performance is obtained.

Specific emitter identification (SEI) uses advanced techniques to identify radio equipment by analyzing unique characteristics in radio frequency signals. Recently, deep learning (DL) has been considered a promising tools for designing various intelligent SEI methods. This is primarily due to its ability to fully exploit hidden data features and make autonomous classification decisions, leading to effective performance. The existing DL-SEI methods are based on the availability of extensive labeled datasets, however, collecting and annotating such data is challenging and time-consuming in real-world scenarios. Furthermore, these datasets often contain both device-specific and irrelevant features, which limits the adaptability of models to fixed channels. To overcome these challenges, we propose a robust variable-channel SEI (VC-SEI) method. This method uses semantic consistency-powered semi-supervised domain adaptation (SSDA). We introduce domain adversarial training to ensure global semantic consistency (GSC), allowing the extraction of channel-irrelevant features. Additionally, we design two loss functions to maintain local semantic consistency (LSC) for extracting category-relevant features. This approach enables effective domain adaptation. Our SSDA-based VC-SEI method has been rigorously evaluated using the ORACLE RF fingerprinting datasets from 16 USRP X310 radios. When only 1% of training samples in the target domain are labeled, our method achieves 84.20% identification accuracy in the target domain and 92.00% identification accuracy in the source domain. These results surpass those of current state-of-the-art methods. Simulation results confirm the robust identification performance of our proposed VC-SEI method in both source and target domains across all scenarios. Our code can be downloaded from https://github.com/frownean/VC-SEI-based-SSDA.

Malware traffic classification (MTC) is one of the important research topics in the field of cyber security. Existing MTC methods based on deep learning have been developed based on the assumption of enough high-quality samples and powerful computing resources. However, both are hard to obtain in real applications especially in availability of IoT. In this paper, we propose a few-shot MTC (FS-MTC) method combining knowledge transfer and neural architecture search (i.e. NAS-based FS-MTC) with limited training samples as well as acceptable computational resources, in order to mitigate the identified challenges. Specifically, our proposed method first converts the raw network traffic into traffic images through data pre-processing to serve as input data for the neural network. Second, we use neural architecture search to adaptively search for the effective feature extraction model on the source domain (including Edge-IIoTset, Bot-IoT, and benign USTC-TFC2016). Third, the searched model is pre-trained on source task to achieve the generic feature representation of malware traffic. Finally, we only use few-shot malware traffic samples to fine-tune the pre-trained model to quickly adapt to new types of MTC tasks in realistic network environments. The experimental results show that the proposed NAS-based FS-MTC method has great scalability and classification performance in different FS-MTC tasks, including 5-way K-shot USTC-TFC2016 dataset and 10-way K-shot CIC-IoT dataset. Compared with state-of-the-art methods in the field of malware classification, the proposed NAS-based FS-MTC has higher classification accuracy. Especially in the 1-shot case of the USTC-TFC2016 dataset, its average accuracy is as high as 86.91%.