High-Level Synthesis (HLS) tools convert C/C++ design code into Hardware Description Language (HDL) code automatically, which are often used for Field Programmable Gate Array (FPGA) design. HLS tools provide many pragmas, which are a kind of directive to be inserted into C/C++ code, for designers to efficiently control the synthesis of code components (e.g., arrays and loops) to generate FPGA implementations with varying performances and costs. However, the use of some pragmas may trigger HLS tool bugs (e.g., tool crashes). Although many formal methods have been proposed to verify the correctness of various HLS phases, no relevant work addresses the problem on detecting HLS tool pragma-related bugs. To resolve this problem, two challenges need to be addressed, namely the selection of testing programs and the acquisition of pragma combinations, due to the enormous number of testing programs and pragma combinations. In this paper, we propose TEPACS, a TEsting Program and prAgma Combination Selection-based framework, to construct diverse testing programs with pragmas for effectively detecting HLS tool pragma-related bugs. TEPACS follows the idea of fuzzing, which is a widely used technique in software testing. First, TEPACS selects the representative testing program according to the cosine distance between the code component vectors of testing programs. Then, for a selected program, TEPACS generates its golden output and uses the pragma combination selection method based on combinatorial testing to generate a set of programs with different pragmas. TEPACS uses the HLS tool under test to convert these testing programs into HDL codes and obtains the simulation results of the HDL code. Finally, based on differential testing, TEPACS identifies HLS tool bugs triggered if the simulation result and golden output are inconsistent. We evaluate TEPACS and its five variants on Vitis HLS, a widely used FPGA HLS tool. Experimental results show that TEPACS outperforms the baselines by at least 11.17% in terms of the bug-finding capability. In one month, TEPACS detected 34 bugs on the latest version of Vitis HLS, of which 9 bugs have been confirmed.

As an instance segmentation model, Mask R-CNN can be well applied to text detection tasks, but the accuracy and efficiency of its backbone network, such as ResNet or ResNeXt, are relatively low. To improve the accuracy and computational efficiency, we propose a novel backbone network for Mask R-CNN, called ResNect (Residual Network with channel mixing). ResNect increases model accuracy (reflected by the F1 score on MTWI dataset) by mixing multi-scale features, and improves the efficiency (reflected by the runtime tested on CIFAR-100 dataset) of the backbone network by reducing the module expansion. Through these two methods, the computational requirements are reduced while increasing the accuracy. The experimental results show that, compared with the backbone networks ResNet, ResNeXt and Res2Net, the runtime of ResNect tested on CIFAR-100 is reduced by 15.8%, 34.5% and 29.3%, and the Mask R-CNN with ResNect as the backbone network also has the highest F1 score on MTWI.

In Industrial Internet of Things (IIoT), a large number of wireless nodes communicate through limited channel resources. Using IEEE802.11ac/ah with multiple users multiple-input–multiple-output (MU-MIMO) and channel bonding technology in IIoT, bonding multiple channels for transmission links can improve data transmission quality. How to reasonably bond limited channel resources on demand for data transmissions in IIoT has become a key issue to improve network performance and ensure communication quality of the nodes. In this article, the definition of outage probability is extended from the amount of information to the signal noise ratio (SNR), which changes the outage probability from a statistical quantity to a quantity that can be directly calculated. This article further analyzes various factors affecting the outage probability, and derives the direct calculation formula of the outage probability in single-hop and multihop data transmission, which allows the outage probability to be directly calculated by some simple parameters. Based the outage probability, this article proposes a dynamic channel bonding algorithm based on outage probability (DCB-OP), which can bond multiple channels for the data transmissions with a high outage probability to improve the success rate of data transmissions. The experimental results show that in IEEE 802.11ac/ah networks, the DCB-OP algorithm can improve the utilization of channel resources and increase the throughput by about 40% compared with no channel bonding. Compared with the general channel bonding algorithm, DCB-OP can make the network throughput higher.

Generative Adversarial Networks (GANs) effectively capture the true posterior distribution. When applied to Collaborative Filtering (CF), GANs can generate a recommendation list through implicit feedback. However, the discriminators in the existing GANs-based CF methods are not utilized fully, and the generators perform poorly on sparse data mining. In this paper, we propose an improved collaborative filtering framework based on variational denoising for GANs (VDGAN). Specifically, VDGAN integrates the variational encoder and the self-attention mechanism into the GANs. By using the positive-negative sampling mechanism to add specific noise to the input data, the variational encoder obtains a robust feature matrix and improves the sparse data processing capability of the generator. In VDGAN, the denoising generator reconstructs the user-items interaction matrix through the feature matrix. And the discriminator is composed of the self-attention mechanism to obtain the explicit features of user preferences, which extends the ability of the discriminator. Furthermore, reinforcement learning replaces the traditional objective function of GANs, which better optimizes the generator and further improves the recommendation accuracy of the model. From our comprehensive experiments on three real-world datasets, we demonstrate that the performance of VDGAN significantly outperforms the state-of-the-art methods based on GANs and Auto-Encoders.

Due to rarely considering document popularity and personalization issues at the same time in the extraction of semantic features based semantic matching algorithms, the accuracy is low in the field of information retrieval. To solve the problem, an improved personalized information retrieval algorithm DSMN is proposed. DSMN bases on the deep struct semantic model (DSSM), uses independent recurrent neural network (IndRNN) to extract semantic features, and process long sequences. In DSMN, the self-attention mechanism is used to further extract the features, and the semantic similarity is calculated. By combining the semantic similarity with the processed user characteristics and document popularity, the relevance score of the query and the document is calculated to improve the efficiency of information retrieval. Experiments are done on four datasets, and the results show that the performance of DSMN is significantly better than the other state-of-the-art information retrieval algorithms based on semantic matching.

Neural networks have been widely used in privacy protection and intrusion detection. As the core algorithm of neural network optimization parameters, the gradient descent algorithm is an important reason for the wide application of neural networks. For the problem that Adam may have a low learning rate in the later stage and the possibility of surpassing the global optimum, this paper proposes a new algorithm IdiffGrad based on the ratio of the first-order moment square to the second-order moment on the basis of diffGrad. The algorithm adjusts the learning rate of different dimensions based on the local variation of the gradient of this dimension and the ratio of the square of the first moment to the second moment, so as to better meet the requirement of this dimension for learning rate. Comparing IdiffGrad with Adam and diffGrad through simulation. The simulation results show that the IdiffGrad algorithm is better in comprehensive performance. It has broad prospects in various intrusion detection schemes using neural networks.

It is difficult for users to find the required information promptly in the massive data. The collaborative filtering is an effective way to help users get the proper information. To achieve better performance, an improved framework based on variational Generative Adversarial Networks with self-attention (VGCF) is proposed. In VGCF, self-attention mechanism and Variational Autoencoders are combined to form self-attention variational encoder, which improves the ability of acquiring explicit and implicit features of sparse data and obtains the personal preference and the correlation between users. By utilizing Generative Adversarial Networks for prediction, the generative model uses the compression matrix obtained by self-attention variational encoder to generate the predicted user-items interaction matrix. The discriminative model provides a better approximation for the posterior and maximum-likelihood assignment, which makes the generated result closer to the real data distribution. Finally, we show that the performance of VGCF is significantly better than the state-of-the-art recommendation methods on several real-world datasets.

The central component of E-Health is electronic health records, however, the number of electronic records that need to be stored is increasing day by day. The electronic records can be stored with the help of DBaaS in the Cloud. Electronic health records contain a lot of private information, so it is unreasonable to store the plaintext records in the cloud. There will be additional computing costs if using traditional encrypted storage to retrieve all the records in the cloud database and retrieve them after decryption. In order to improve the retrieval efficiency of electronic health records as much as possible while ensuring user privacy, ISDB expands on the model of the secure query model SDB, so that the improved model supports the data types of varchar, date and date time in SQL Encryption. In ISDB, the sting-like types of exact matching function is designed, so that the model can be more effectively used for encrypted storage of electronic health records and ciphertext retrieval. By analysis and experiments, the ISDB can meet the security ciphertext retrieval of commonly used data types in electronic health records, and it is a light model and is effective.

WMN (Wireless Mesh Network) is a useful wireless multi-hop network with tremendous research value. A good routing mechanism for WMN can use the whole bandwidth of the network and can assure the quality of service of traffic. Through simulations we verify that the routing metric ETX (Expected Transmission Count) cannot assure good quality of wireless links. To improve the routing performance, a QoS-assured opportunistic routing mechanism is proposed in this paper. This mechanism can choose the highest throughput links to improve the performance of routing over WMN and then reduce the energy consumption of mesh routers. The analyses show that the opportunistic routing mechanism is better than the mechanism with the routing metric of ETX.

ABSTRACT Trust mechanism has become a research focus in recent years as a novel and valid way to ensure the transaction security in peer‐to‐peer file sharing networks. Nevertheless, some fundamental challenges still exist, for example: How can malicious peers be effectively isolated? How can various threats of manipulation by strategic peers be resisted? What strategy should be used to ensure that the service providers are authentic peers? Considering these challenges in our minds, in this paper, we propose a new trust mechanism based on the maximum flow theory. We firstly add a few prestigious peers into a cluster as the original members according to their transaction behaviors in a period; then, we perform maximum flow algorithm and identify those peers that still link from (to) the peers in the cluster as new members, which is carried out repeatedly, and almost every normal peer would finally become the member of the cluster. Each request peer has the priority to select downloading sources from this cluster according to our trust mechanism. In this way, the malicious peers are isolated, and their transaction behaviors are also confined largely even though they have high reputation. Extensive experimental results confirm the efficiency of our trust mechanism against the threats of exaggeration, cheat, collusion, and disguise. Copyright © 2013 John Wiley & Sons, Ltd.

Cooperation within selfish individuals can be promoted by natural selection only in the presence of an additional mechanism. In this paper, we focus on an indirect reciprocity mechanism in dynamical structured populations. In social networks rational individuals update their strategies and adjust their social relationships. We propose a three-strategy prisoner's dilemma game model to investigate the evolution of cooperation on dynamical networks. In the coevolution of state and structure process, reciprocators adapt their behaviors and switch their partners based on reputation. Simulation results show that the dynamics of strategies and links can promote cooperation provided the partners switch proceeds much faster than the strategy updating.