Current Issue

20 May 2019, Volume 35 Issue 5

Previous Issue    Next Issue

Topics:intelligent communication technologies and applications

Application of artificial intelligence in NFV

Zengyi LIU, Bo LEI, Mingchuan YANG

2019, 35(5):  1-8.  doi:10.11959/j.issn.1000-0801.2019094

Asbtract ( 446 )   HTML ( 56)   PDF (2047KB) ( 626 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

How to deal with the complicated management problem after network function virtualization,how to flexibly schedule the network,all these problems need to be solved in order to develop the NFV.The rise of artificial intelligence provides a new solution for the management and orchestration in the NFV system.The present situations of NFV and artificial intelligence were summarily introduced,and use-cases of artificial intelligence in NFV were discussed.In addition,a solution for the management of VNF lifecycle management based on machine learning was proposed to provide reference for artificial intelligence technology in the application of NFV.

Application of artificial intelligence in network orchestration system

Tianjiao CHEN,Jiang LIU,Tao HUANG

2019, 35(5):  9-16.  doi:10.11959/j.issn.1000-0801.2019095

Asbtract ( 701 )   HTML ( 59)   PDF (675KB) ( 606 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Based on software defined networking and network virtualization,network orchestration can deploy differentiated services quickly and efficiently.With the wide application of artificial intelligence technology,the combination of network orchestration system and artificial intelligence has become the research focus of major operators and manufacturers.Firstly,the concept of network orchestration system and the development status of major open source organizations were introduced.Then,based on the design time and the run time,the network orchestration architecture using artificial intelligence was introduced to realize the transformation from “automation” to “intelligence”.Finally,the application of artificial intelligence in SDN,NFV and operation and maintenance in the orchestration system was summarized,and the future development direction was proposed.

Smart prediction of the complaint hotspot problem in mobile network

Lin ZHU,Juan ZHAO,Yiting WANG,Junlan FENG,Gchao DEN

2019, 35(5):  17-24.  doi:10.11959/j.issn.1000-0801.2019099

Asbtract ( 466 )   HTML ( 36)   PDF (811KB) ( 198 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In telecom communication network,a hot customer complaint often affects hundreds even thousands of users’ service and leads to significant economic losses and bulk complaints.An approach was proposed to predict a customer complaint based on real-time user signaling data.Through analyzing the network business layer logic,30 key segments related to the user experience in the S1 interface data were selected.Further,one-hot features,statistical derived features,and differential features were extracted to classify user perceptions in detail.Considering the problems of noise data and unbalanced training samples,LightGBM was chosen to train the prediction model.Experiments are conducted to prove the effectiveness and efficiency of this proposal.As of today,this approach has been deployed in our daily business to locate the hot complaint problem scope as well as to report affected users and area.

Research on compatibility of high-performance network stack

Huiyou JIANG,Junfeng LI,Dan LI

2019, 35(5):  25-31.  doi:10.11959/j.issn.1000-0801.2019096

Asbtract ( 356 )   HTML ( 19)   PDF (670KB) ( 431 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The massive data processing in artificial intelligence system requires the support of high-throughput network.As a bridge between the upper artificial intelligence applications and the underlying high-speed hardware network adapter,the network stack plays a crucial role in the performance of the whole network system.Through detailed analysis and comparison of the compatibility of mainstream network stacks,it is found that the compatibility of various network stacks is generally low,which leads to difficulties in transplantation of artificial intelligence applications.The idea of achieving both high-compatibility and high-performance of user-level network stack were proposed.In this way,legacy applications could directly attain the improvement of performance without modification of the source code.Path for the subsequent development of network stack was lit up.

Neural network based image and video coding technologies

Chuanmin JIA, Zhenghui ZHAO, Shanshe WANG, Siwei MA

2019, 35(5):  32-42.  doi:10.11959/j.issn.1000-0801.2019142

Asbtract ( 739 )   HTML ( 51)   PDF (919KB) ( 433 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Deep neural networks have achieved tremendous success in artificial intelligence,which makes the broad and in-depth research of neural network resurge in recent years.Recently,the neural network based image and video coding has become one of the front-edge topics.A systematic and comprehensive review of neural network based image and video coding approaches based on network structure and coding modules were provided.The development of neural network based image compression,e.g.multi-layer perceptron,random neural network,convolutional neural network,recurrent neural network and generative adversarial network based image compression methods and neural network based video compression tools were introduced respectively.Moreover,the future trends in neural network based compression were also envisioned and discussed.

A survey of neural architecture search

Mingjie HE,Jie ZHANG,Shiguang SHAN

2019, 35(5):  43-50.  doi:10.11959/j.issn.1000-0801.2019097

Asbtract ( 364 )   HTML ( 17)   PDF (846KB) ( 228 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Recently,deep learning has achieved impressive success on various computer vision tasks.The neural architecture is usually a key factor which directly determines the performance of the deep learning algorithm.The automated neural architecture search methods have attracted more and more attentions in recent years.The neural architecture search is the automated process of seeking the optimal neural architecture for specific tasks.Currently,the neural architecture search methods have shown great potential in exploring high-performance and high-efficiency neural architectures.In this paper,a survey in this research field and categorize existing methods based on their performance estimation methods,search spaces and architecture search strategies were presented.Specifically,there were four performance estimation methods for computation cost reduction,two typical neural architecture search spaces and two types of search strategies based on discrete and continuous spaces respectively.Neural architecture search methods based on continuous space are becoming the trend of researches on neural architecture search.

Key technologies and application of artificial intelligence in telecom real-name system

Hui YAO,Siyan MA

2019, 35(5):  51-58.  doi:10.11959/j.issn.1000-0801.2019093

Asbtract ( 396 )   HTML ( 17)   PDF (789KB) ( 237 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

According to the real-name system requirements of the State Council and the Ministry of Industry and Information Technology, combined with the needs of telecom operators’ business transformation, an online identity authentication solution based on artificial intelligence technologies was proposed. Firstly, a real-name authentication system architecture which is based on artificial intelligence technologies was introduced. Then the key modules of the system were expounded in detail. Secondly, the key technologies applied in real-name authentication system were explained, such as human authentication and identification, OCR, face liveness detection. The present situations of applying identity authentication solution were presented. Besides, the means of real-name authentication business scenarios against attacks from the perspective of security were introduced and the effectiveness and efficiency of the proposed solution were demonstrated. Finally, the application problems of artificial intelligence technologies in the field of operator business were analyzed and the future works on real-name authentication products were discussed.

research and development

Intrusion detection model based on fuzzy theory and association rules

Jianwu ZHANG,Jiasen HUANG,Di ZHOU

2019, 35(5):  59-69.  doi:10.11959/j.issn.1000-0801.2019077

Asbtract ( 306 )   HTML ( 20)   PDF (790KB) ( 207 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

An intrusion detection model based on fuzzy theory and improved Apriori algorithm was proposed.The BV-Apriori algorithm was used to generate the matching rule base,and the problem of excessive boundary in the continuous data partitioning process was solved by fuzzy set technology.The real-time analysis of the relationship between features and the update of the rule base were completed,and the intrusion detection model BVA-IDS (Boolean vector Apriori-intrusion detection system) was built.The results show that the mining efficiency of the BV-Apriori algorithm is significantly improved when compared with the existing Apriori-BR algorithm,in addition,the BVA-IDS model also performs well on intrusion detection indicators with high detection accuracy,and low false positive rate and false negative rate.

Communication performance analysis of wireless sensor network for railway environmental monitoring

Ruifeng CHEN,Ming NI,Chunjie XU,Hao HU

2019, 35(5):  70-77.  doi:10.11959/j.issn.1000-0801.2019091

Asbtract ( 308 )   HTML ( 7)   PDF (1382KB) ( 234 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Wireless sensor network (WSN) deployed along the railway can be commonly applied to environmental monitoring.In order to lengthen the lifetime of WSN,it is essential to introduce the node sleeping mechanism.Therefore,firstly the stochastic process theory was adopted and an analytical framework was proposed to study the communication performance of WSN in railway scenarios based on the node sleeping mechanism.Furthermore,the impacts of different system factors on the reception performance of WSN were analyzed under railway scenarios,and then the relation between the optimal ratio of active nodes and other system factors could be obtained.A simulation platform was established to verify the reception performance of WSN in safety areas and hazards-prone sections respectively.Finally,practical case was presented to verify the deployment scheme based on node sleeping mechanism.The analytical results could provide the guidelines for practical deployment of WSN under railway scenarios.

A ranking hashing algorithm based on listwise supervision

Anbang YANG,Jiangbo QIAN,Yihong DONG,Huahui CHEN

2019, 35(5):  78-85.  doi:10.11959/j.issn.1000-0801.2019072

Asbtract ( 202 )   HTML ( 10)   PDF (806KB) ( 180 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Recently,learning to hash technology has been used for the similarity search of large-scale data.It can simultaneous increase the search speed and reduce the storage cost through transforming the data into binary codes.At present,most ranking hashing algorithms compare the consistency of data in the Euclidean space and the Hamming space to construct the loss function.However,because the Hamming distance is a discrete integer value,there may be many data points sharing the same Hamming distance result in the exact ranking cannot be performed.To address this challenging issue,the encoded data was divided into several subspaces with the same length.Each subspace was set with different weights.The Hamming distance was calculated according to different subspace weights.The experimental results show that this algorithm can effectively sort the data in the Hamming space and improve the accuracy of the query compared with other learning to hash algorithms.

Provident garbage collection algorithm for SSD storage system

Xuezhen TU,Zhenjiang Huang,Zhengguang CHEN

2019, 35(5):  86-96.  doi:10.11959/j.issn.1000-0801.2019075

Asbtract ( 312 )   HTML ( 11)   PDF (887KB) ( 268 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

A predictive based proactive garbage collection algorithm was proposed.Firstly,the data was separated according to different heat factors,then the upper and lower predictions were performed on the number of different types of page allocation requests (PAR) that would be reached in the future.While satisfying the page allocation request lower prediction,the PAR upper prediction requirement was maximally satisfied,the WA problem was optimized,and invalid effective data migration was reduced,thereby maximizing the garbage collection utility.A mathematical model was defined for this problem,and an algorithm for obtaining the approximate optimal solution was given.The applicable scenario of the model was analyzed.The practical results show that the algorithm can obtain the maximum benefit and can significantly improve the performance of SSD and reduce the cost.

Enterprises:

Telecom complaint hot topic detection method based on density peaks clustering

Jun JIANG, Hua HUANG, Tiaojuan REN, Denghui ZHANG

2019, 35(5):  97-103.  doi:10.11959/j.issn.1000-0801.2019076

Asbtract ( 269 )   HTML ( 22)   PDF (484KB) ( 159 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In view of the lack of effective detection methods for hot topics in telecom industry,a method of complaint hotspots detection based on density peaks clustering algorithm was proposed.Firstly,a special vocabulary for telecommunication industry was established to segment the complaint samples.The vector space model was used to represent the text segmentation.Then,the similarity and density of the text segmentation were calculated,and the clustering analysis of the words was carried out by using the density peaks clustering algorithm.Finally,keywords were selected and sorted by clustering.This method was applied to the complaint hotspots detection tasks within a telecom company.The results show that this method is effective and has practical application value.

research and development

Quality assessment of synthetic viewpoint stereo image with multi-feature fusion

Shuainan CUI,Zongju PENG,Wenhui ZOU,Fen CHEN,Hua CHEN

2019, 35(5):  104-112.  doi:10.11959/j.issn.1000-0801.2019070

Asbtract ( 268 )   HTML ( 9)   PDF (1380KB) ( 224 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

A multi-feature fusion method was proposed,the features of the distortion areas,the distortion edges and the singular value were fused.Firstly,the distortion areas corresponding to the human eyes were obtained by using the parallax and the threshold,and then the average structural similarity in these areas was calculated.Secondly,the distorted edges in the synthetic viewpoint edge image were extracted,and then the average structural similarity was calculated.Thirdly,the difference between the original image and the distorted image singular value was calculated.Finally,the three features were fused to obtain the final objective quality score.The experiment was carried out on the synthetic viewpoint stereo image library.The experimental results show that the Pearson linear correlation coefficient and the Spearman correlation coefficient are both higher than 0.86.Compared with the existing assessment methods,it can better reflect the quality of synthetic viewpoint stereo images.

summarize

Developing background,service characteristics and challenges of LEO IoT

Jun SHEN,Weibin GAO,Gengxin ZHANG

2019, 35(5):  113-119.  doi:10.11959/j.issn.1000-0801.2019089

Asbtract ( 778 )   HTML ( 109)   PDF (682KB) ( 1272 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The low earth orbit (LEO) internet of things (IoT) is an effective solution to solve the coverage inefficiency problem of terrestrial IoT in ocean,mountain and desert areas.On the basis of analyzing the developing background and situation of LEO IoT,the main application and service characteristics were given,finally the technical challenges and key technologies need to be solved for LEO IoT were discussed.

Operation technology wide Angle

Application of digital twin model based industrial internet

Bin LIU,Yunyong ZHANG

2019, 35(5):  120-128.  doi:10.11959/j.issn.1000-0801.2019088

Asbtract ( 1004 )   HTML ( 82)   PDF (896KB) ( 2838 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Industrial internet is an important engine for manufacturing enterprises to provide intelligent products and services in China.From manufacturing enterprises’ point of view,the definition and data sources of industrial internet were introduced,and the five-layer function architecture of industrial internet was proposed.The industrial data analysis model represented by digital twin played a crucial role in developing industrial internet for manufacturing enterprises,whose definition,characteristics,key technologies and applications in the whole product lifecycle were elaborated.Finally,the problems and solutions of carrying forward industrial internet in China were addressed.

DNS log analysis system based on big data fusion algorithm

Ming LIAO,Ming CHEN,Ji ZHOU,Xiaohua XIANG,Fang LI,Yefen JIAO

2019, 35(5):  129-139.  doi:10.11959/j.issn.1000-0801.2019065

Asbtract ( 546 )   HTML ( 35)   PDF (3284KB) ( 282 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The existing problems of big data analysis of DNS logs in the current industry were summarized,and a DNS log analysis system based on big data fusion algorithm was designed.The system integrates six information databases,such as record information database,website database,domain name database,IP address database,CNAME domain name database and security information database.It has powerful log analysis performance,comprehensive content analysis,and classification.With high analysis depth and flexible report customization function,it can provide decision-making basis for telecom operators to improve the management efficiency,optimize network security and enhance the network rate.

Solution and key technology of integrating vFW in cloud platform

Zhilan HUANG,Yongbing FAN,Ning FAN,Nan CHEN,Linze WU,Baohong LIN

2019, 35(5):  140-148.  doi:10.11959/j.issn.1000-0801.2019047

Asbtract ( 547 )   HTML ( 25)   PDF (1357KB) ( 372 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

FWaaS is one of the key cloud network services,which requires ability of auto provisioning and flexibility of on-demand adjustment.It is suitable for cloud platforms to provide FWaaS based on virtual firewalls (vFW) appliance.vFW has its own particularity comparing to other cloud components.It is not only a network service but also a network element,which should be orchestrated by business systems and should be automatically configured by network systems.In the environment of cloud resource pool,vFW faces the integration problem with other cloud components like SDN and cloud management platform.The main method of integrating vFW in cloud resource pool was studied,the existing integration problems were analyzed,and an integration method based on extensible model driven framework was proposed.

Optimization method of wireless indoor coverage based on big data analysis

Yuanfeng GONG,Yi HUANG

2019, 35(5):  149-154.  doi:10.11959/j.issn.1000-0801.2019140

Asbtract ( 362 )   HTML ( 13)   PDF (1723KB) ( 363 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The indoor prediction method of indoor depth coverage includes user complaints,CQT and propagation model correction,and the prediction error of these methods is large.Firstly,the indoor coverage optimization method was introduced.Then,the indoor deep coverage optimization method based on big data was introduced,which significantly improved the network quality and user perception.Finally,the application case was introduced,which provided a very important reference and guiding significance for the construction,optimization and maintenance of indoor deep coverage in the later period.

Application of blockchain technology in the demand response management of power grid

Haowen REN,Yaqi YANG

2019, 35(5):  155-160.  doi:10.11959/j.issn.1000-0801.2019066

Asbtract ( 657 )   HTML ( 38)   PDF (733KB) ( 243 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The theory of decentralized blockchain mechanism was studied,and a method of applying distributed block chain technology to manage demand response planning in smart grid environment was proposed.This method was based on the blockchain distributed ledger accounting tamper-proof technology,which stored the energy consumption information collected from the intelligent metering devices of the IoT,and defined the self-executing intelligent contract at each consumer level in a programmatic way to ensure the flexibility of transactions.Blockchain distributed technologies provided incentives or penalties to balance energy demand with grid energy production rules.Consensus-based validation would be used to validate demand response plans and provide appropriate financial settlement for flexible suppliers.The results show that the power demand side management based on blockchain distributed technology can be used to match energy demand and production under the condition of smart grid,with high accuracy of demand response signal,but less flexibility required for convergence.