Blockchain is a kind of non-tampered global database which provides digital booklets recording financial transactions and valuables such as birth certificates, asset ownerships, education diplomas, medical procedures, ballots and so on. Among the financial activities, coupons have longtime been used for marketing in business and mobile coupons are replacing paper ones due to the popularity of mobile devices gradually. The strong interactions between people and the use of mobile coupons have together inspired Peer to Peer (P2P) sharing and raised the usage of coupons with peer trusts. Adding bonus feedback mechanisms to mobile coupons may effectively prompt transfers of unused coupons to potential users. The valuable information including contents and feedback bonus in mobile coupons need to be protected against security threats like forgery or tampering. The limitations of storage space, computing power, and transmission bandwidth in mobile devices generally constrain strong protection mechanisms as used in desktop computers. Therefore, in this study, the technique of Hash Chain is combined with the blockchain technology to verify forgery coupons and to enable both storage and transmission of low-computation requirements. Our scheme exemplifies P2P sharing spirit with the nature of highly distribution in blockchain. Verifications are handed by the distributed miners in the P2P environment, which reduces the opportunity of conspiracy among evil-doers. In addition, eavesdroppers have no way to steal valuable information due to the anonymous nature of blockchains so that transmissions are entirely secured in the proposed scheme.
RSSI-based localization technology is widely concerned for WSN node localization. However, this kind of methods has relatively larger localization error, and its localization rate is low. In order to enhance the localization accuracy, the localization error is declined by considering the construction of signal transmission model as well as the ranging process, and a modified weighted centroid localization algorithm is proposed in this paper. The simulation results show that the proposed approach can effectively reduce the three-dimensional node localization error.
Redundant backup system is a common system in the real world, especially in aviation, electromechanics and other fields. Therefore, it is meaningful to give the parameter estimation of redundant systems. Previous research has given some results about the order and failure rate of the redundancy distributed redundant backup system, thus providing consumers with a computed and measurable product reliability calculation method. By using the irreplaceable censored life test, Providing a statistical significance of the estimation formula. However, the properties of estimation are not given and how to improve it. The study gives the estimation properties in probabilistic sense, which is asymptotic convergence in probability. Continue to improve the estimator on variance, so as to achieve asymptotic minimum consistent variance estimation. This ensures that the given estimation formula is available in practice.
The Mobile Ad-hoc NETwork (MANET) is a dynamic network topology, which is mainly achieved by mobile nodes through peer-to-peer communication or packet forwarding. However, an efficient routing algorithm between nodes is necessary to be proposed due to lack of physical backbone. As a result, a popular method which is called Connected Dominating Set (CDS) is proposed to be a virtual backbone of a MANET. The main concept of CDS is to reduce the search space for the route to the mobile node in the set. Subsequently, the CDS can be invoked to as the virtual backbone of MANET to reduce the cost of routing information and enhance the scalability of the network. However, the most algorithms cannot find the minimum CDS in a MANET. Therefore, the concept of nominate is used to our proposed algorithm to construct the minimum CDS in this paper.
With the development of the Internet and smart terminals, Quick Response (QR) code and its related applications become increasingly popular. Though there are plenty of advantages for the usage of QR code, but its security issue is always a problem that can not be overlooked. Aiming for embedding message into QR code, a Module Expansion (ME) based method is proposed in this paper. The core idea of ME is to expand a module to its neighbor if these two modules are in different colors. Experimental results show that (1) our capacity for embedding is about half of that for encoding of the same QR code, which is much higher comparing to the state-of-the-art methods, (2) for the QR codes of versions under 20, the average time cost for embedding and extraction is around 51 ms and 59 ms respectively, (3) the success ratio of message extraction under noise attack and print-and-scan process are both higher than 80%. The proposed method can be used in QR code applications such as the anti-counterfeiting for encoded message of a printed QR code, and others for which message sharing is needed.
Spoken dialog systems have become popular and are used in a home environment, such as smart speakers. A problem will occur when two or more smart speakers are in the same environment, in which a dialog system misdetects the other dialog systems voice as a users voice. In this paper, a method to mute synthesized speech is proposed to prevent a speech recognizer from recognizing speech uttered by a machine. The audio watermark technique is used to indicate that a machine utters the speech, and the speech recognizer attenuates the observed speech if it contains the watermark. The watermark is embedded in high frequency so that humans cannot perceive the watermark and the watermark is robustly extracted. From the experimental result, we found that the proposed method robustly determine the existence of the watermark when the SNR is no less than 0 dB.
Watching other people playing games on live streaming platforms have become more popular. In published literature, most of researches on live streaming focused on predicting the number of viewers in the live streaming period, explaining the high peak of the audience in a game, and finding out popular live streamers, and discussing usage behaviors such as exploring the gift giving. However, from available literature, relatively few works focus on discussing the text chats/comments which can affect other users’ watching behaviors. Therefore, this study aims to find important terms that affect viewing of live streaming. We used live game streaming as our study target. Using the comments of the audience in the chat room of the Twitch live streaming platform as experimental samples. Text mining and feature selection methods, including Least Absolute Shrinkage and Selection Operator (LASSO), Support Vector Machine-Recursive Feature (SVM-RFE) and chi-square test (χ^2 test), to find important terms that affect viewing of live streaming.
The purpose of this investigation is to verify the accuracy and stability of a novel detection scheme based on pulsation micro-vibration signals. Different from traditional heart rate measurement, this scheme has the advantage of convenience in comparison with the pulsations determined by EKG and PPG. Actually, the heartbeat pulsation measurement based on EKG is usually served as the ground true. However, the pulsation measurement based on PPG was popularly fulfilled on many wearable devices such as rings or watches. To assure the measuring effectiveness of heartbeat pulsations, the scheme is compared to the pulsation measurement based on EKG and PPG. After several experiments and signal processing steps, the statistics of correlation coefficients based on the Pearson correlation coefficient were obtained, and the correlation coefficients among the signals due to EKG, PPG, and micro-vibration are as high as 0.98. Since modern smart phones have the same inertial sensors, which are used in this investigation to detect the micro-vibration signals from heartbeats, it is very promising to realize an APP of smart phones to detect heartbeats more conveniently.
In this paper, we deal with melody completion, a technique which smoothly completes melodies that are partially masked. Melody completion can be used to help people compose or arrange pieces of music in several ways, such as editing existing melodies or connecting two other melodies. In recent years, various methods have been proposed for realizing high-quality completion via neural networks. Therefore, in this research, we examine a method of melody completion based on an image completion network. We represent melodies of a certain length as images and train a completion network to complete those images. The completion network consists of convolution layers and is trained in the framework of generative adversarial networks. We also consider chord progression from musical pieces as conditions.
In this study, we propose a voice conversion technique with two-stage conversion, which is realized by using two models consisting of U-Net and pix2pix. Using U-Net, we tried to reproduce intonation of a target speaker by performing low-dimensional feature conversion considering the time direction. We introduced pix2pix for the task of spectrogram enhancement. The pix2pix is trained to map from low definition spectrogram to high definition spectrogram (low-to-high spectrogram mapping). Low definition spectrogram is reconstructed from low dimensional mel-cepstrum converted by U-Net and high definition spectrogram is extracted from natural speech. In objective evaluations, we showed that the proposed method was effective in improvement of mel-cepstral distance (MCD) and Log F0 RMSE. Subjective evaluations revealed that the use of the proposed method had a certain effect in improving speech individuality while maintaining the same level of naturalness as the conventional method.
Over the recent years, ontologies are widely used in the biomedical domains. However, biomedical ontology heterogeneity problem hamper the cooperation between intelligent applications based on biomedical ontologies. It is crucial to establish correspondences between the heterogeneous biomedical concepts in different ontologies, which is so-called biomedical ontology matching. Approaches based on Multi-Objective Evolutionary Algorithm (MOEA), such as NSGA-II, are emerging as a new methodology to solve the ontology matching problem. In this paper, to further improve the quality of biomedical ontology alignments, a hybrid NSGA-II is proposed, which modifies the knee solutions in the Pareto front by using a local search method. Experiment utilizes two biomedical ontology matching tracks provided by Ontology Alignment Evaluation Initiative (OAEI 2017). The experimental results show that our approach outperforms the participants of OAEI 2017 and NSGA-II based ontology matching technique.
This paper proposed an attack pattern mining algorithm based on improved fuzzy clustering and sequence pattern mining. The method combines the advantage of fuzzy clustering to describe the similarity between security logs and the advantage of sequence pattern to describe the logical relationship in attacking steps. The experimental results show that the algorithm can effectively mine the attack pattern, improve the accuracy and generate more effective attack pattern.
According to pixels out of bounds and the serious distortion caused by too much pixel value modification in traditional PVD steganography, This paper proposes a steganography algorithm based on pixel block difference and variable modulus function (PVBD). The algorithm uses the modular function of dynamic parameters to optimize the amplitude of the pixel value modification, and at the same time the related parameters of the mode function are dynamically variable, which increases the security of the algorithm on the basis of the diversity of the embedded modes. Experiments show that the algorithm this paper proposed has better performance in the embedded capacity and digital image quality as well as other properties than the existing PVD algorithm.
Due to Digital Selective Calling (DSC) under Global Marine Distress and Safety System (GMDSS), the existing software and equipment need to rely on foreign manufacturers. The maintenance cost is high, and the existing display interface of the software can only indicate the information of the vessels in distress by text message. In order to improve above-mentioned shortcoming. The Automatic Identification System (AIS) and DSC real-time information integration system was constructed. Through the customized system, the operators in the coast station can understand the vessels in distress and the relevant surrounding information of the vessels at the first time and thus shorten the emergency response time.
In the era of blossoming computer sciences and internet technology, people cannot abolish network in our lives. However, the large number of users, website services will make itself became the most favorite targets for hackers. Although these malicious behaviors can be detected by network intrusion detection system, it is difficult to generate accuracy result owing to the shortage of data. This paper proposed a solution using host intrusion detection system that focus on the host log detection of webserver. Besides using port monitoring to monitor network environment, this paper also collected signatures of web attack and malicious activities by using signature-based approach. Furthermore, this research will find out the source of the malicious files with file monitoring function, and take appropriate action to protect web services. By using the proposed mechanism of host-based intrusion detection methods, it can provide a high accuracy to bring safety for managers and users.
To concise the truck's box volume measuring system, some techniques such as calculation of camera intrinsic parameters, estimation of truck's position and orientation, and calculation the vertex coordinates on the truck's box are discussed in details. A novel method based on one single image is proposed under the constraint information of orthogonal vanishing points and circle projection priori. Firstly, three orthogonal vanishing points are detected, the intrinsic parameters of the camera are calculated based on the vanishing points. Secondly, the world coordinate system is established at the center of one of the truck's wheels the truck's position and orientation related to the camera's plane and the external facade equation of truck's box can be figured out with the camera's intrinsic parameters and the prior of wheel's single circular projection. Finally, the vertex coordinates of the truck's box are calculated based on the external facade equation and matrix equation of projection algorithm. Then the volume of truck's box can be calculated. Experimental results show that the error of the truck's box volume is within 5%. The proposed approach is more effective and lower cost than other state of art methods, it is a competitive approach for real-time measurement of truck's box volume.
Among current information hiding methods, the least significant bits (LSBs) substitution is the most common. By modifying the least significant bit that has no effect on the image, secret information can be hidden in image media and cannot be perceived by the human visual system, making information transmission via the Internet more secure. Not all blocks in the image have the ability to carry secret messages of the same length and high confidence. Also, success can be achieved by hiding more secret information in complex areas of the image. This research proposes a hybrid steganographic method by using LSB substitution and pixel-value differencing (PVD) to increase the embedding capacity for the digital image. Moreover, the proposed method can effectively resist an RS detection attack and effectively improve the security of transmitting secret messages.
The controller area network (CAN) has been widely used in the modern automotives for interconnecting electrical components such as air bag system, anti-lock braking system (ABS), electronic dashboard, fuel injection system, and etc. In order to make sure the urgent message, e.g., ABS, could be processed in the shortest time, CAN bus protocol establishes the priority of the message, and allows certain messages can take priority over others. It goes without saying that this design is quite desirable for vehicular applications; however, it also provides vulnerability for Denial-of-Service (DoS) attacks. It is possible for malicious adversaries to cause major damage by exploiting flaws in the CAN protocol design or implementation. Some of these attacks can lead to catastrophic consequences for both the vehicle and the driver. This paper proposed a study on the impact of such priority based DoS attacks. Experimental results shown that a significant impact on the CAN bus efficiency of priority- based DoS attacks. In addition, a single attacker could block an entire CAN network just using fake CAN message with continuous injection.
