Current Issue

20 February 2019, Volume 35 Issue 2

Previous Issue    Next Issue

research and development

Multi-objective cloud workflow scheduling algorithm based on grid variance

Xiaoan BAO,Yundi CAO,Na ZHANG,Junyan QIAN,Jianwen CAO

2019, 35(2):  1-13.  doi:10.11959/j.issn.1000-0801.2019035

Asbtract ( 203 )   HTML ( 9)   PDF (1015KB) ( 433 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Multi-objective cloud workflow scheduling algorithm based on grid variance and the strategy of bad particles self-learning were presented.Firstly,the characteristics of task scheduling was token into consideration,and particle encoding was discredited.Secondly,the strategy of mapping Pareto optimal workflow scheduling set to self-adaptive grid coordinate system,and calculating the grid distribution value of each Pareto optimal solution was used.Thirdly,grid variance was adopted to evaluate the diversity of current Pareto front and dynamically adjust evolution strategies.Finally,the concept of being dominated times was introduced into bad particles self-learning strategy for filtering out bad particles in population.The simulation experiment shows that workflow scheduling solution set by this algorithm is better than the MOPSO algorithm on both IGD and S performance indexes,and the optimal value is superior to the ε-FDPSO and NSGA-Ⅱ algorithm.

Complexity control method of random forest based HEVC

Peng WEN,Zongju PENG,Fen CHEN,Gangyi JIANG,Mei YU

2019, 35(2):  14-26.  doi:10.11959/j.issn.1000-0801.2019034

Asbtract ( 283 )   HTML ( 5)   PDF (1024KB) ( 547 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

High efficiency video coding (HEVC) has high computational complexity,and fast algorithm cannot perform video coding under restricted coding time.Therefore,a complexity control method of HEVC based on random forest was proposed.Firstly,three random forest classifiers with different prediction accuracy were trained to provide various coding configurations for coding tree unit (CTU).Then,an average depth-complexity model was built to allocate CTU complexity.Finally,the CTU coding configuration,determined by the smoothness,average depth,bit,and CTU-level accumulated coding error,was used to complete complexity control.The experimental results show that the proposed method has better complexity control precision,and outperforms the state-of-the-art method in terms of video quality.

Data aggregation scheme based on SDN for internet of things

Bowei WANG,Jin TAN,Wenbin JIANG

2019, 35(2):  27-34.  doi:10.11959/j.issn.1000-0801.2019028

Asbtract ( 407 )   HTML ( 17)   PDF (900KB) ( 444 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

There are many redundant parts in the data generated and transmitted by IoT devices.Transmitting this data reduces network lifetime and bandwidth.Aiming at the problem,the aggregation gateway layer was proposed,and a SDN-based IoT data aggregation architecture and algorithm were designed.The simulation analysis shows that compared with the architecture of the ordinary gateway and OF-GW,the network lifetime is improved by about 32.6%,and the amount of data packets sent by the sensing layer to the network layer is reduced by 6.1% to 55.1%.Compared with the LEACH algorithm,the designed rule algorithm reduces the average delay of data aggregation by about 24.1%,while avoiding data loss and improving the accuracy of data aggregation.

A distributed CRN resource allocation algorithm based on CBR and cooperative Q-learning

Lin XU,Zhijin ZHAO

2019, 35(2):  35-42.  doi:10.11959/j.issn.1000-0801.2019005

Asbtract ( 320 )   HTML ( 7)   PDF (805KB) ( 567 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In order to solve the problem of channel and power allocation in distributed cognitive radio networks (CRN),a case-based reasoning (CBR) and cooperative Q-learning algorithm was proposed.In order to optimize the Q initialization of Q-learning algorithm,the current problem and the historical case were matched according to the similarity function,the Q value of the matching case was extracted and normalized as the initial value.Cooperative Q-learning was based on the total reward value,and each agent integrates the Q values of other agents with higher reward values with different weights to gain learning experience to reduce unnecessary exploration.Simulations show that the proposed algorithm can improve the energy efficiency of the cognitive system’s channel and power allocation,and accelerate the convergence speed of the system.

Spectrum allocation algorithm of elastic optical networks in power distribution system

Biyou ZHAO,Shanhui ZHANG,Jinshuai WANG

2019, 35(2):  43-50.  doi:10.11959/j.issn.1000-0801.2019027

Asbtract ( 220 )   HTML ( 2)   PDF (795KB) ( 285 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In order to effectively improve the resource utilization of information transmission network in reality,the spectrum allocation problem in flexible optical network in smart grid environment was analyzed,and a spectrum allocation method was proposed.Combining the advantages of genetic algorithm and ant colony algorithm,the genetic algorithm was used to generate the initial solution.Finally,the optimal solution of the spectrum allocation problem was obtained by the characteristics of positive feedback of ant colony algorithm and efficient convergence.The performance of the optimization algorithm mentioned was verified by software simulation.The proposed algorithm can optimize the use of idle spectrum resources to meet the needs of mass data transmission in distribution network,and improve the reliability of distribution network communication system,which is of great significance to the safe and reliable operation of smart grid.

Polarization-based puncturing pattern selection of shortened Polar code

Shunlan LIU,Aiming HU,Jianrong BAO

2019, 35(2):  51-58.  doi:10.11959/j.issn.1000-0801.2019006

Asbtract ( 345 )   HTML ( 7)   PDF (852KB) ( 411 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In order to solve the problem of inflexible Polar code length,a new puncturing pattern selection scheme based on feedback cost puncturing was proposed to generate shortened Polar codes.In this scheme,the Polarization effect was reserved,the sub-optimal subchannel was selected as the cost bit,the output bit corresponding to the cost bit was used as the puncturing pattern,and then the value of the cost bit is adjusted by feedback to obtain the shortened Polar code.Simulation results show that the performance of the proposed shortened Polar code is better than that of the shortened Polar code generated by bit-reversal permutation method,and the shortened Polar codes generated by direct puncturing method at the end under different channel and code rate conditions,and slightly weaker than that of the shortened Polar code proposed by Wang R and Liu R et al.In the AGWN channel,when the block error rate is 10^-3,the proposed shortened Polar code with a code rate of 0.5 can obtain a gain of 0.5 dB compared with the shortened Polar code generated by bit-reversal permutation method,and the coding performance is greatly improved.

Omni directional video coding rate distortion optimization method based on weighted-to-spherically-uniform structural similarity

Yufeng ZHOU,Mei YU,Baozhen DU,Yang SONG,Gangyi JIANG

2019, 35(2):  59-69.  doi:10.11959/j.issn.1000-0801.2019025

Asbtract ( 216 )   HTML ( 7)   PDF (3081KB) ( 492 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Aiming at the shortcomings of the coding rate distortion model in traditional video coding,considering the characteristics of equal rectangular omnidirectional video,a method of omnidirectional video coding rate distortion optimization based on spherical weighted structural similarity (WS-SSIM) was proposed.By considering the distortion of the internal structural similarity of the projection plane and the relationship between the spherical distortion and the projection plane distortion,the “spherical weighted structural similarity” was proposed to describe the degree of distortion of the planar omnidirectional image local block relative to the viewing sphere,which was applied to the rate-distortion optimization process of omnidirectional video coding and adaptive selection of quantization parameters to improve vision-based coding efficiency.The experimental results show that compared with the HEVC video coding standard HM16.9 test platform,the proposed method can save an average of 24.48% code rate under the same visual quality,which indicates that the method has significant performance for improving the omnidirectional video coding rate distortion performance.

Overview

Research status and development trend of optical communication technology

Xiangdong LV,Xuerui LIANG,Qianchen YU,Weidong MA

2019, 35(2):  70-78.  doi:10.11959/j.issn.1000-0801.2019009

Asbtract ( 678 )   HTML ( 77)   PDF (870KB) ( 1751 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

With the rapid growth of network bandwidth demand and the development of new applications and network services,the development of optical networks is facing a huge bottleneck of transmission capacity,switching granularity and network control.In order to solve these problems,researchers around the world have put forward new solutions.The technological innovation in optical networks in recent years was discussed,the development of dynamic reconfigurable optical network,integrated optoelectronics and optical fiber transmission technology was mainly discussed,and the future development trend of optical networks was analyzed in order to provide reference for related research.

Operational Technology Wide Angle

Architecture evolution and bandwidth analysis of 5G bearer networks

Chunhua ZHAO

2019, 35(2):  79-83.  doi:10.11959/j.issn.1000-0801.2019037

Asbtract ( 588 )   HTML ( 37)   PDF (688KB) ( 958 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The influence of 5G mobile network architecture evolution on bearer network was demonstrated.A comparison of fiber requirement between 5G front-haul network and fixed broadband access network was proposed.The various bearer schemes in fiber shortage area were introduced.The characteristics of various device networking solution were summarized.The influence on bearer network architecture and bandwidth was analyzed when 5G core network moved down to the metro core DC and metro edge DC.The bandwidth calculation model on mobile metro mid-haul and back-haul network was presented.Bandwidth calculations on each layer of the networks were provided.Calculation results show that compared with 4G network,bandwidth requirements on bearer back-haul network are greatly increased as the bandwidth of the wireless side of the 5G network increases.The growth trend of the metro back-haul and export bandwidth is slowed down with the core network moving down and the deployment of CDN and MEC.

Implementation strategy of telecom operators’ data lake technology

Junjun HU,Xiaojun XIE,Yanbin SHI,Qi YU

2019, 35(2):  84-94.  doi:10.11959/j.issn.1000-0801.2019029

Asbtract ( 457 )   HTML ( 36)   PDF (996KB) ( 937 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Firstly,the traditional data architecture of telecom operators was analyzed and the problems such as low quality,lack of real-time,lack of flexibility and mutual constraints of storage and application were identified.Then,the feasibility of introducing data lake technology was discussed,the storage scale and low cost of data lake,the originality of data and ease of use,and the application of on-demand modeling were clarified,which could provide great benefits for telecom operators’ data architecture optimization.Finally,the data lake construction plan of unified storage,unified standards,near-source acquisition,and application separation was proposed,and key points such as data lake data partitioning,deployment,lake entry and application data dynamic loading,and unified data management were clearly stated.

Data center energy efficiency case and energy saving operation

Jieliang LOU

2019, 35(2):  95-104.  doi:10.11959/j.issn.1000-0801.2019036

Asbtract ( 274 )   HTML ( 21)   PDF (679KB) ( 331 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

PUE value is the key indicator of energy efficiency in data center.Through the comparative analysis of refrigeration efficiency,cooling efficiency,energy conversion efficiency and energy transmission efficiency,the reasons for difference in PUE between two data center cases were found.Three measures on the strategy of energy-saving operation of data center were put forward,as to control the number of refrigerators,to reduce the temperature of cooling water and to raise the temperature of chilled water.Some inspirations on energy saving in the construction and operation of the data center were raised.

Practical exploration on the construction of data asset management system

Guohe LI,Zheng FENG,Zhuoyu WANG,Yong SUN,Yang GUO,Qiguo SAN

2019, 35(2):  105-112.  doi:10.11959/j.issn.1000-0801.2019039

Asbtract ( 662 )   HTML ( 31)   PDF (761KB) ( 566 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

With the change of policy environment and the upgrading of information technology,the big data strategy has officially risen to the national strategy.Digital China construction is accelerating in all directions,and data asset management has received unprecedented attention.The construction of data asset management system has become an inevitable choice and inevitable trend for enterprises to cope with the requirements of the era of big data.Data value mining needs to provide protection through data asset management system.Focusing on the problems existing in the big data management and control of State Grid Xinyuan Co.,Ltd.,based on the advanced and mature data asset management knowledge system at home and abroad,combined with the company’s business characteristics and business needs,research on data asset management system was carried out from three aspects:organizational system,management and control system and system platform.The research of data asset management system laid foundation for the construction practice of the later data asset management system,and provided effective data for improving corporate governance capabilities,optimizing core resource allocation,promoting the birth of big data in the pumped storage industry and the sustainable and healthy development of corporate data assets.

Application of DBN deep learning algorithm in anti stealing electricity system

Dandan LI,Bingyu GE,Wenwen HUANG,Lei XIE,Shengqi QIAN

2019, 35(2):  113-117.  doi:10.11959/j.issn.1000-0801.2019038

Asbtract ( 283 )   HTML ( 6)   PDF (829KB) ( 523 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

With the development of economy,the electric power demand increases gradually,but because of the relative backwardness in the automation of electricity,the phenomenon of electric stealing is common.But the traditional anti electric stealing means generally centered around how to strengthen the technical transformation of the electric energy metering device,and the management efficiency is low.The purpose of deep learning is to use the method of constructing the multi-layer neural network model.To learn the potential features of image,text,voice and other data,it also has good effect on the classification problem.The successful application of the deep learning algorithm in many complex fields provides a new effective way to solve the problem of anti stealing electricity.The structure and learning algorithm of DBN and the anti-stealing model based on DBN algorithm was mainly introduced.Finally,experiments were carried out and the results were analyzed.

Power Informatization Column

Security signaling optimization in optical network for smart substation

Yuhao ZHAO,Xiaoguang HAO,Shaobo GENG,Peng LUO,Kun ZHOU,Jialin ZHEN,Yurong MAO

2019, 35(2):  118-124.  doi:10.11959/j.issn.1000-0801.2019026

Asbtract ( 269 )   HTML ( 4)   PDF (805KB) ( 401 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The communication security issue brought by smart substation is of great importance and should not be ignored in backbone optical networks.To solve this problem,firstly,the communication security challenges faced by optical networks in smart substation environment were analyzed,a signaling security optimization technique for optical networks for smart substation services was proposed.A so-called light-weighted security signaling scheme was designed to establish the end-to-end trusted connection.Under the environment of multi-domain communication services in smart substation,evaluation simulation for the signaling interaction was conducted.Simulation results show that the proposed approach can greatly enhance the security ability of multi-domain optical network with better performance in term of connection success rate.

Application of big data method in forecasting the risk of tariff recovery

Yadi ZHAO,Zhao WU,Qingbing LI,Xiaofeng CHEN,Baoting WANG

2019, 35(2):  125-133.  doi:10.11959/j.issn.1000-0801.2019040

Asbtract ( 489 )   HTML ( 17)   PDF (1016KB) ( 703 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Based on the historical data of electricity customers,the model index system was determined according to the customers’ basic attributes,the electricity consumption and the payment behavior,the customers’ credit,the industry prospects’ information and so on.Through the correlation coefficient matrix and the information value of the index,the index variables that enter the model were selected.At the same time,the best grouping method was used to group variables and WOE (weight of evidence) transformation was carried out.Based on the processed data,the logic regression algorithm were used to construct the electricity cost risk forecasting model of the electric customers,and output variable standard score card was quantified according to the model results.Thus the customers were divided into high,middle and low risk users that could provide the basis for taking differential marketing measures to the different customers.

Application of BeiDou RNSS timing technology in smart grid lightning locating system

Hao ZHANG,Hongzhen YANG,Weiping SHAO,Jiming YAO

2019, 35(2):  134-141.  doi:10.11959/j.issn.1000-0801.2019032

Asbtract ( 357 )   HTML ( 14)   PDF (957KB) ( 611 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The smart grid lightning location system is an important power service,and its lightning location accuracy largely depends on the accuracy of time synchronization.Three kinds of BeiDou timing methods were studied and compared,and the RNSS timing method was found to be most suitable for lightning location systems.Furthermore,the principle and characteristics of RNSS timing were studied,and a low-cost user terminal timing hardware and software architecture and algorithm were proposed.The Kalman filter-based BeiDou second pulse signal was used to calibrate the user clock and adjust the voltage through the PID algorithm.The temperature-controlled crystal oscillator made the deviation between the user terminal time and the GNSS time below 100 ns at any time to meet the accuracy requirements of the lightning positioning system.

Power substitution potential user identification model based on big data

Huaijian XIA,Wen ZHAGN,Sujin JIA,Haiying LIN,Xinfang HE

2019, 35(2):  142-146.  doi:10.11959/j.issn.1000-0801.2019030

Asbtract ( 369 )   HTML ( 10)   PDF (797KB) ( 477 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The implementation of electric energy substitution plays an important role in promoting the development of energy conservation and emission reduction in China.Firstly,the user clustering model of the same industry was introduced,then the user identification model of electric energy substitution potential was studied.Finally,an industry user was selected for instance verification.The results show that the model can effectively predict the user’s energy substitution potential.

Design and application of content distribution engine for the nuclear power enterprise

Peng WANG,Hongbing TU,Yunfu WANG

2019, 35(2):  147-156.  doi:10.11959/j.issn.1000-0801.2019033

Asbtract ( 248 )   HTML ( 2)   PDF (1052KB) ( 290 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Content distribution is an important business of nuclear power enterprises,and is responsible for transferring structured/unstructured content media between companies,departments,project groups and other levels.The complexity of the content media structure,the participation in the distribution hierarchy,and the variety of distribution business scenarios pose great challenges on the design of the distribution engine.After sorting out the company’s distribution business,a dynamically configurable content distribution engine was designed,which supported business logic configuration and reduced frequent engine changes caused by changes in distribution business.Combined with the database trigger technology,a pre-engine content crawling mechanism was designed,which could greatly reduce the crawling time complexity and space complexity of the content to be distributed,and ensure the immediacy of the task.