Loading...
ISSN 1000-1239 CN 11-1777/TP

Table of Content

01 May 2020, Volume 57 Issue 5
Computation Protocols: Analyzable Abstractions for Computing Systems
Xu Zhiwei, Wang Yifan, Zhao Yongwei, Li Chundian
2020, 57(5):  897-905.  doi:10.7544/issn1000-1239.2020.20200058
Asbtract ( 1012 )   HTML ( 25)   PDF (1539KB) ( 415 )  
Related Articles | Metrics
Computing systems research is entering an era of diversity. At the same time, systems research still mainly follows the prototype development and benchmark evaluation approach, making the research cost too high to address the diversity challenge. This dilemma calls for new analyzable abstractions of computing systems. When researching a new system, we can use its abstraction to analyze its characteristics to filter out inappropriate candidate systems before costly prototyping and benchmarking. We already have such a concept for computer applications, called algorithm. Before an algorithm’s implementation and benchmark evaluation, we can usually analyze its main properties, such as time complexity and space complexity. In this paper, we summarize seven advantages of the algorithm concept and propose a preliminary counterpart for computing systems, called computation protocol. Learning from six historical lessons from systems research, we discuss a general definition, a black-box representation, and a white-box representation of the computation protocol concept. We use preliminary examples to point out that computation protocol thinking may be helpful to propose computing systems conjecture, analyze new parallel computing model, extend existing systems architecture, and inspire new system evaluation method.
Brilliance and Darkness: Turing Test
Yu Jian
2020, 57(5):  906-911.  doi:10.7544/issn1000-1239.2020.20190794
Asbtract ( 1144 )   HTML ( 37)   PDF (441KB) ( 530 )  
Related Articles | Metrics
In this paper we discuss Turing Test and its modifications, study its theoretical presumption and practical feasibility, and briefly survey the development for Turing Test. By analyzing the presumptions of a classical concept definition, the basic assumptions of Turing Test are demonstrated. In this paper, it clearly shows that the basic assumptions of Turing Test are not consistent with humans daily life and social science, which brings greatly theoretical challenges for artificial intelligence research.
Survey on Secure Persistent Memory Storage
Yang Fan, Li Fei, Shu Jiwu
2020, 57(5):  912-927.  doi:10.7544/issn1000-1239.2020.20190820
Asbtract ( 1197 )   HTML ( 29)   PDF (1616KB) ( 425 )  
Related Articles | Metrics
With the rapid development of computer technology, computer security and data privacy protection have always been the focus of academic and industrial. By providing hardware-assisted confidentiality and integrity verification, memory security mechanism helps guarantee the security of application code and data, and prevent them from malicious memory disclosure and modification. The emerging persistent memory delivers a unique combination of affordable large capacity and support for data persistence and provides high-bandwidth and low-latency data access. It can be placed on the memory bus like DRAM and will be accessed via processor loads and stores. However, due to differences in media characteristics, DRAM-oriented memory security mechanisms cannot function efficiently on persistent memory and even have availability issues. Therefore, a secure memory storage system based on persistent memory will bring new opportunities for the secure and efficient memory storage of big data. Firstly, for the write characteristics of persistent memory, the reasons for low-efficiency in applying the security measure against traditional volatile memory to persistent memory are analyzed, and related work is expounded. Secondly, for persistent memory storage, we analyze the problems that need to be considered to ensure the security of persistent memory in its whole life cycle, and introduce research work on guaranteeing the consistency between data and corresponding metadata for security. Finally, we conclude the challenges and compare the related work in building secure memory storage based on persistent memory, and share our views on future research.
An Overview of Monaural Speech Denoising and Dereverberation Research
Lan Tian, Peng Chuan, Li Sen, Ye Wenzheng, Li Meng, Hui Guoqiang, Lü Yilan, Qian Yuxin, Liu Qiao
2020, 57(5):  928-953.  doi:10.7544/issn1000-1239.2020.20190306
Asbtract ( 1004 )   HTML ( 20)   PDF (2215KB) ( 891 )  
Related Articles | Metrics
Speech enhancement refers to the use of audio signal processing techniques and various algorithms to improve the intelligibility and quality of the distorted speech signals. It has great research value and a wide range of applications including speech recognition, VoIP, tele-conference and hearing aids. Most early work utilized unsupervised digital signal analysis methods to decompose the speech signal to obtain the characteristics of the clean speech and the noise. With the development of machine learning, some supervised methods which aim to learn the relationship between noisy and clean speech signals were proposed. In particular, the introduction of deep learning has greatly improved the performance. In order to help beginners and related researchers to understand the current research status of this topic, this paper conducts a comprehensive survey of the development process of the monaural speech enhancement, and systematically summarizes from the aspect of model methods, datasets, features, evaluation metrics, etc. First, we divide speech enhancement into noise reduction and de-reverberation, then respectively sort out the existing work of traditional and machine-learning-based methods in these two directions. Moreover, we briefly introduce the main ideas of typical solutions, and compare the performance of different methods. Then, commonly used datasets, features, learning objectives and evaluation metrics in experiments are enumerated and illustrated. Finally, four major challenges and corresponding issues in this area are summarized.
Research on Node Importance Fused Multi-Information for Multi-Relational Social Networks
Luo Hao, Yan Guanghui, Zhang Meng, Bao Junbo, Li Juncheng, Liu Ting, Yang Bo, Wei Jun
2020, 57(5):  954-970.  doi:10.7544/issn1000-1239.2020.20190331
Asbtract ( 581 )   HTML ( 11)   PDF (6032KB) ( 335 )  
Related Articles | Metrics
Identifying critical nodes is one of the principal tasks of social network analysis, and it is essential to understand the structure and dynamic characteristics of the complex networks. However, the analysis framework of node importance mainly focuses on single-relational networks. As a typical model of the real world, the multi-relational network has become one of the hot topics in the field of network science, in which the research on node importance lacks systematic research. Focusing on the study of node importance in multi-relational social networks, we create the directed multiplex network model to describe a multi-relational network and use the representation framework based on tensor algebra to analyze it. Meanwhile, we propose a measure of node importance considered the influence of centrality, prestige, transitivity in multi-relational social networks. Considering the influence of coupling information and the difference of transmission mechanism for node importance on multi-relational networks, in this work we improve the method and propose another more efficient method called IOMEC to evaluate the node importance. Experimental results on four real networks show that the method of information fusion can effectively eliminate the influence on node importance evaluation, which is caused by the coupling information and the transmission mechanism of the multi-relational network. The IOMEC method can measure the importance of nodes more accurately and has lower time complexity. The experimental results demonstrate that centrality and prestige are the main factors to evaluate the node importance and the necessity of considering the transitivity of nodes. In this work we not only provide new ideas and methods for evaluating node importance for multi-relational networks but also expand the application of information fusion technology.
A Reasoning Method for Qualitative Distance Change Based on OPRA\-4 Direction Relations
Dong Yiqun, Liu Jiandong, Xu Wenxing, Wang Shuhong
2020, 57(5):  971-983.  doi:10.7544/issn1000-1239.2020.20190442
Asbtract ( 512 )   HTML ( 5)   PDF (1500KB) ( 160 )  
Related Articles | Metrics
Spatial information includes many relations such as direction, topology, shape, distance, etc. Qualitative spatial representation and reasoning has become an important subfield of artificial intelligence, and has gained increasing popularity in recent years with applications in spatial information systems, robot navigation, natural language understanding, intelligent transportation system and so on. Previous studies are mostly oriented to static spatial objects, and focus on a single kind of spatial relation. The research on constraints between different kinds of spatial relations is insufficient, and it is difficult to make use of one kind of spatial relation to reason about the evolution of another spatial relation effectively. In this study, we focus on the qualitative direction relations and distance changes of moving spatial objects. Firstly, the relative moving direction between two spatial objects is described by a combination of the position relations between the corresponding ray and the circle. Secondly, the restrictions of the combinations on the qualitative distance changes, and the corresponding relationship between the combinations and oriented point algebra with granularity of 4 (OPRA\-4) are studied respectively. And then the connection between the OPRA\-4 direction relations and the qualitative distance changes is established. Finally, an approach to reasoning about the qualitative distance changes with basic OPRA\-4 direction relations is presented. The correctness and effectiveness of the approach are illustrated by an example of continuous k nearest neighbor queries of moving objects in traffic field.
app Popularity Prediction with Multi-Level Attention Networks
Zhang Yixuan, Guo Bin, Liu Jiaqi, Ouyang Yi, Yu Zhiwen
2020, 57(5):  984-995.  doi:10.7544/issn1000-1239.2020.20190672
Asbtract ( 660 )   HTML ( 11)   PDF (2747KB) ( 320 )  
Related Articles | Metrics
The popularity prediction of mobile apps provides substantial value to a broad range of applications, ranging from operational strategy optimization to targeted advertising and investment. This work includes leveraging the rich data provided by the app market to mine the dynamic correlation between different factors and popularity, so as to predict the app popularity over the next period of time, which creates great value for developers, investors and the app market. However, the evolution of app popularity is highly dynamic, and its influence factors are very complex, including the iterative evolution of the app itself, user feedback, and competition for similar products and so on. At present, there are relatively few research studies on app popularity modeling and prediction. Most of them construct artificial features and capture its association with popularity, and there is room for improvement in terms of computational performance, prediction accuracy, and interpretability of results. In this paper, we propose DeePOP, an attention based neural network for app popularity modeling and prediction, which performs hierarchical modeling for complex influence factors. First, we propose the time-level self-sequence module to capture the long-term dependence on historical popularity, and propose the local and global feature level modules to capture the nonlinear relationship between features and app popularity. Second, the attention mechanisms provide adaptive capabilities for different modules to capture most relevant historical states and provide explanation for prediction. Last, the experimental results show that DeePOP outperforms the state-of-the-art methods and the root mean square error of prediction reaches up to 0.089.
A Pedestrian Tracking Algorithm Based on Multi-Granularity Feature
Wang Ziye, Miao Duoqian, Zhao Cairong, Luo Sheng, Wei Zhihua
2020, 57(5):  996-1002.  doi:10.7544/issn1000-1239.2020.20190280
Asbtract ( 697 )   HTML ( 17)   PDF (1972KB) ( 408 )  
Related Articles | Metrics
Recently in some popular applications, such as video scene surveillance, long-term effective pedestrian tracking is the basis of these applications. Although the related technology of target detection and target tracking have a long history, how to achieve real-time and accurate pedestrian tracking is still an active research field and needs to be solved. At present, most pedestrian tracking methods only use hand-designed features to track or only use deep learning to extract features, which are not good ways to represent the features of the target because the use of one single feature will restrict the expression of the features. Therefore, multi-granularity hierarchical features are used in this paper to achieve more stable pedestrian tracking. This paper proposes an improved pedestrian tracking algorithm. The algorithm adopts the idea of multi-granularity, combines convolutional feature with bottom color feature, makes decision on the tracking result obtained by GOTURN, a tracking algorithm based on deep learning, and modifies the tracking result with target detection. This paper uses Pascal VOC data set for model training, and uses OTB-100 and VOT 2015 data sets for testing. The experimental results show that the tracking algorithm based on multi-granularity decision can track target pedestrians more accurately than a single tracking algorithm and the tracking accuracy is improved obviously.
Survey on Biometrics Template Protection
Wang Huiyong, Tang Shijie, Ding Yong, Wang Yujue, Li Jiahui
2020, 57(5):  1003-1021.  doi:10.7544/issn1000-1239.2020.20190371
Asbtract ( 782 )   HTML ( 20)   PDF (4239KB) ( 359 )  
Related Articles | Metrics
Biometric authentication (BA) has become an important means of identity authentication. However, many BA systems deployed at present do not take enough consideration in protecting the security and privacy of users biometric data, which has become a main obstacle to the popularization and application of the BA technology. BA systems may face various attacks from software or hardware implementations, among which, template attack is the main consideration. Many technical literatures have been devoted to dealing with this type of attacks. However, existing review literatures suffer from incomplete descriptions or conflicting discussions. In order to systematically summarize the attacking and protection technologies against biometric templates, some related concepts of the BA system is introduced at first, as well as the architecture of a BA system and the connotation of BA security and privacy. Then, template protection technologies for a BA system are classified into two main categories for description: the transformation-based methods and the crypto-based methods, which solves some conflictions in existing literatures. Afterwards, some classical methods and emerging technologies in each category are expounded and analyzed, as well as some subsequent evaluations and improvements. Finally, several major difficulties and the corresponding possible solutions for building a secure BA system are pointed out.
A Fast Traffic Sign Detection Algorithm Based on Three-Scale Nested Residual Structures
Li Xudong, Zhang Jianming, Xie Zhipeng, Wang Jin
2020, 57(5):  1022-1036.  doi:10.7544/issn1000-1239.2020.20190445
Asbtract ( 684 )   HTML ( 31)   PDF (4085KB) ( 419 )  
Related Articles | Metrics
Automatic driving technology has high requirements for real-time and robustness of traffic sign detection in real world. The YOLOv3-tiny model is a lightweight network with good real-time performance in the object detection, but its accuracy is not high. In this paper, we use YOLOv3-tiny as the basic network and propose a fast traffic sign detection algorithm with three-scale nested residual structure. Firstly, shortcut based on pixel by pixel addition is employed in the YOLOv3-tiny network. It does not increase the number of feature map channels, and a small residual structure is formed in the network at the same time. Secondly, the predictive output with higher spatial resolution is also added through the shortcut, which contains more abundant spatial information, thus forming a large residual structure. Finally, the two residual structures are nested to form a three-scale predictive nested residual network, which makes the main network of Tiny located in these two residual structures and the parameters can be adjusted three times. The results show that the proposed algorithm can quickly and robustly detect traffic signs in real scenes. The F\-1 value of total traffic signs achieves 91.77% on German traffic sign detection benchmark and the detection time is 5ms. On CSUST Chinese traffic sign detection benchmark, F\-1 values of the Mandatory, the Prohibitory and the Warning are 92.41%, 93.91% and 92.03% respectively, and the detection time is 5ms.
Multi-Modal Knowledge-Aware Attention Network for Question Answering
Zhang Yingying, Qian Shengsheng, Fang Quan, Xu Changsheng
2020, 57(5):  1037-1045.  doi:10.7544/issn1000-1239.2020.20190474
Asbtract ( 928 )   HTML ( 23)   PDF (1982KB) ( 699 )  
Related Articles | Metrics
With the popularity of the Internet, more people choose to search online to find the solutions when they feel sick. With the emergence of reliable medical question answering websites, e.g. Chunyu Doctor, XYWY, patients can communicate with the doctor one-one at home. However, existing question answering methods focus on word-level interaction or semantics, but rarely notice the hidden rationale with doctors’ commonsense, while in the real scenes, doctors need to acquire plenty of domain knowledge to give advice to the patients. This paper proposes a novel multi-modal knowledge-aware attention network (MKAN) to effectively exploit multi-modal knowledge graph for medical question answering. The incorporation of multi-modal information can provide more fine-grained information. This information shows how entities in the medical graph are related. Our model first generates multi-modal entity representation with a translation-based method, and then defines question-answer interactions as the paths in the multi-modal knowledge graph that connect the entities in the question and answer. Furthermore, to discriminate the importance of paths, we propose an attention network. We build a large-scale multi-modal medical knowledge graph based on Symptom-in-Chinese, as well as one real-world medical question answering datasets based on Chunyu Doctor website. Extensive experiments strongly evidence that our proposed model obtains significant performance compared with state-of-the arts.
Detection of Persistent Elements in Distributed Monitoring System
Lu Le, Sun Yu’e, Huang He, Wang Runzhi, Cao Zhen
2020, 57(5):  1046-1056.  doi:10.7544/issn1000-1239.2020.20190287
Asbtract ( 422 )   HTML ( 16)   PDF (1079KB) ( 137 )  
Related Articles | Metrics
The detection of persistent elements has many important applications in the fields of detecting intrusions in a distributed system, finding common interests, measuring the traffic, etc. Most of the existing state-of-the-art studies of detecting persistent elements have some problems such as false or missing report, high communication cost and great limitation so that they can hardly satisfy the requirement of some distributed applications. To solve these problems, the paper proposes a scheme to detect persistent elements in distributed monitoring system with the goal of minimizing the total communication cost during the whole detection process. First, the scheme filters out most of the irrelevant elements to reduce the overall communication overhead through multiple rounds of compressed data transferring between all monitors and the central coordinator. Then, we ensure that each round of the filtering is necessary and can achieve the best performance by adjusting parameters of filtering according to the theoretical analysis and derivation. With the technology of extended Bloom filter and the persistent spread estimation function, the scheme can work well no matter in the balanced environment or in the unbalanced environment. Finally, we perform extensive simulations to study the performance of the proposed mechanism, and the simulation results show the effectiveness of our scheme.
Updatable Attribute-Based Encryption Scheme Supporting Dynamic Change of User Rights
Yan Xincheng, Chen Yue, Ba Yang, Jia Hongyong, Wang Zhonghui
2020, 57(5):  1057-1069.  doi:10.7544/issn1000-1239.2020.20190254
Asbtract ( 561 )   HTML ( 13)   PDF (2306KB) ( 186 )  
Related Articles | Metrics
Attribute-based encryption has great advantages in achieving fine-grained secure sharing for cloud data. Due to the dynamic changes of user access rights in cloud storage, data re-encryption is an effective method to ensure the forward security of ciphertext when the attribute or user private key is revoked, but the corresponding computation overhead and communication overhead of data uploading and downloading are too large. To address these issues, an updatable attribute-based encryption scheme is proposed to support dynamic changes of user rights (SDCUR-UABE). By constructing the attribute version key and user version key in ciphertext-policy attribute-based encryption, only the corresponding components of transformation key in user’s private key need to be updated when the user attribute is revoked. Similarly, when a system attribute is revoked, the corresponding attribute version key needs to be updated to implement replaceable update of part components for the ciphertext and key. Next, only the user version key needs to be updated when the user private key is revoked. Therefore the expensive computation and communication overhead caused by ciphertext update based on data re-encryption can be avoided. Besides, key segmentation is used to realize data decryption outsourcing to reduce the user’s decryption overhead in the construction of the scheme. Theoretical analysis and experimental verification show that the proposed scheme can effectively solve the computing efficiency and communication overhead of ciphertext update when the user rights are dynamically changed in the cloud storage system, and greatly reduce the computational complexity of user decryption under the premise of guaranteeing forward security for ciphertext.
An Anonymous Agent Tracking Privacy Preserving Scheme in Mobile Healthcare System
Luo Entao, Duan Guoyun, Zhou Lei, Zhu Xiaoyu
2020, 57(5):  1070-1079.  doi:10.7544/issn1000-1239.2020.20190307
Asbtract ( 536 )   HTML ( 16)   PDF (3938KB) ( 234 )  
Related Articles | Metrics
To solve doctor’s service bottlenecks under the peak condition of mobile medical, the paper proposes utilizing authorized agents to provide service, so as to reduce the burden of doctors and enhance the flexibility of service as well. We use attribute encryption technology to encrypt medical data. The ciphertext can be decrypted correctly when the proxy attribute satisfies the policy set by the authorized user, so as to obtain proxy signature to provide medical service for medical users. The scheme not only can alleviate the burden on the mobile medical service providers in the peak of the bottleneck, but also can solve the authorization problem during the absence of doctors. At the same time, we can achieve proxy traceback to preform the agency accountability and avoid unauthorized agency audit. This scheme can be used to reduce the computing cost of authorized doctors by using trusted authorization center. The simulation results reveal that the performance can be improved significantly than other scheme. At the same time, the candidate number prediction model and the supply computational resources prediction model are proposed. They can effectively estimate the number of candidate and collaborative computing power ability, which can be used for reference to alleviate the bottleneck problem in peak computing.
Model of Trusted Cooperative Service for Edge Computing
Yue Guangxue, Dai Yasheng, Yang Xiaohui, Liu Jianhua, You Zhenxu, Zhu Youkang
2020, 57(5):  1080-1102.  doi:10.7544/issn1000-1239.2020.20190077
Asbtract ( 611 )   HTML ( 16)   PDF (7223KB) ( 237 )  
Related Articles | Metrics
With the development and widespread application of the Internet of things and 4G/5G wireless network technology, we have entered into the Internet of everything era. It is easier to connect the edge computing devices, such as mobile phones, PAD, etc., to the Internet. Thus, the number of data generated by edge computing devices is increasing significantly. However, the current network services cannot provide such demand, posed by edge computing, on high throughput, frequently connection, sensitive to location and latency. It is an efficient way to improve the quality of service by 1)considering the characteristics of the intelligence, diversity and flexibility of node at the edge of network, 2)locally aggregating computing, storage and network service resources, and 3)adaptively building trusted cooperative service system. The key to efficiently build a cooperative service system is quickly looking for and then dynamically organizing the edge computing nodes. In this paper, we propose a leaderbased trusted cooperative service for edge computing (TCSEC). The main idea is the leader node selects its cooperative service node set with a selfadjustable clustering, which takes into consideration the features of a node, e.g. trust degree, influence, volume, bandwidth, and quality of the link, and realizes the rapid resource aggregation and computing migration. Based on our approach, it is fast to respond to the computing request and provide reliable service. The simulation shows TCSEC can efficiently speed up the ability to construct a cooperative service system and improve the quality of service.
Review on the Development of Microservice Architecture
Feng Zhiyong, Xu Yanwei, Xue Xiao, Chen Shizhan
2020, 57(5):  1103-1122.  doi:10.7544/issn1000-1239.2020.20190460
Asbtract ( 3116 )   HTML ( 170)   PDF (3960KB) ( 1726 )  
Related Articles | Metrics
With the rapid development of cloud computing and Internet of things, users’ demand for software systems tends to be diversified. Service oriented architecture (SOA) needs to strike a balance between stable service integration and flexible adaptation of requirements. Based on this situation, the microservice technology, which goes with independent process as well as independent deployment capability, emerges as the times require. It has a slew of advantages, such as distributed storage, high availability, scalability, and intelligent operation maintenance, which can make up for the shortcomings of the traditional SOA architecture. From the perspective of system integration, the paper firstly describes the application background of microservice, which include the core components of microservice, software technology development and architecture evolution to ensure the availability of microservice infrastructure. Secondly, in view of problems existing in practical applications, the paper analyzes the key technologies utilized in the specific application of the microservice architecture through the aspects of distributed communication, distributed data storage, distributed call chain, and testing complexity; then, a specific application case is given to confirm the technical feasibility of microservice. Finally, this paper intends to explore the challenges by microservice through the aspects of infrastructure, information exchange, data security, and network security. Meanwhile, the future development trend is analyzed so as to provide valuable theoretical and technical reference for the future innovation and development of microservice.