Current Issue

25 December 2013, Volume 34 Issue 12

Previous Issue    Next Issue

Papers

Based on optimally rigid graph energy efficient distributed topology control algorithm

Xiao-yuan LUO,Yan-lin YAN,Li-juan HAO,Shao-bao LI,Xin-ping GUAN

2013, 34(12):  1-10.  doi:10.3969/j.issn.1000-436X.2013.12.001

Asbtract ( 305 )   HTML ( 3)   PDF (275KB) ( 900 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

To solve the problem that most of existing algorithms cannot balance energy consumption,the energy efficient distributed topology control algorithm (OREE) based on optimally rigid graph was presented.Through the link weight function reflecting both the energy consumption and residual energy of two end nodes,OREE builds a dynamic network topology that changes with the variation of node energy,and thus balances energy consumption of nodes.The topology derived under OREE is 2-connected and sparse was proved,and the average degree of nodes in the topology converges to four approximately.Furthermore,simulation results show that OREE can balance energy consumption of nodes effi-ciently compared with other algorithms,thereby extend the lifetime of networks.

Hybrid channel allocation strategy with group switching and performance evaluation in cognitive radio network

Shun-fu JIN,Hong-ting XIE,Yuan ZHAO

2013, 34(12):  11-19.  doi:10.3969/j.issn.1000-436X.2013.12.002

Asbtract ( 205 )   HTML ( 1)   PDF (156KB) ( 476 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Considering the non-real-time traffic in cognitive radio networks and two groups of licensed channels,com-bing the methods of random back-off and control center schedule,a novel hybrid channe allocation strategy with channel group switching mechanism was proposed.Taking into account the total number of the cognitive packets and the numbers of the primary packets on each group channels,a three-dimensional Markov model was constructed,some cogniti user measures in terms of the blocking ratio,the packet loss ratio and the average delay were derived accordingly.Numerical results with analysis and simulation show that,as the back-off time changes,a tradeoff exists between different perfor-mance measures.Therefore,by minimizing the system cost,the back-off time can be optimized.

Multiple DAG off-line task scheduling for heterogeneous networked automobile electronic systems

Guo-qi XIE,Ren-fa LI,Fan YANG,Wei-hong HUANG

2013, 34(12):  20-32.  doi:10.3969/j.issn.1000-436X.2013.12.003

Asbtract ( 314 )   HTML ( 8)   PDF (286KB) ( 726 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Fairness and communication overhead have become the major bottleneck in performance of scheduling,a fair sorting criteria based on round-robin with communication overhead weight was proposed and a selection criteria based on assigning the task to the minimum selection value considering insertion was proposed in processor selection phase.The multiple DAG off-line and fairness task scheduling(MDOFTS)algorithm was proposed combining the above two phases to reduce the schedule length and communication overhead.Heterogeneous networked automobile electronic systems are typical mixed-criticality embedded systems,which must make sure to be real-time and to reduce scheduling length.The multiple DAG off-line and priority task scheduling (MDOPTS) algorithm was proposed to make sure the safety-critical DAG.The multiple DAG off-line and adaptive task scheduling(MDOATS)algorithm was proposed to improve the system performance on the basis of real-time based on MDOFTS and MDOPTS.Example analysis and experimental results show that the MDOFTS algorithm is better than other algorithms in schedule length,communication overh unfairness,worst-case response time and real-time.

CSP-based resource allocation model and algorithms for energy-efficient cloud computing

Wei-wei LIN,Bo LIU,Liang-chang ZHU,De-yu QI

2013, 34(12):  33-41.  doi:10.3969/j.issn.1000-436X.2013.12.004

Asbtract ( 319 )   HTML ( 3)   PDF (215KB) ( 797 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The resource allocation for energy-efficient cloud computing in a heterogeneous environment was modeled as a constraint satisfaction problem (CSP).By solving the constraint satisfaction problem,the optimized allocation scheme minimizing energy consumption in virtualized cloud data centers was obtained.Based on the optimized allocation scheme,an energy-efficient resource allocation algorithm,dynamicpower (DY),which takes into account the heterogene-ity of resources,was proposed.The performance of algorithm was evaluated using Choco.Experimental results show that,compared with first-fit decreasing (FFD),best-fit decreasing (BFD) and minimizing the number of physical machines (MinPM),the proposed algorithm (DY) has less energy consumption.

Sensing-then-transmission slot scheduling optimization for opportu-nistic spectrum access under the interference constraint of primary us-ers

Jing ZHANG,Yin LU,Xi-qi GAO,Fu-chun ZHENG

2013, 34(12):  42-48.  doi:10.3969/j.issn.1000-436X.2013.12.005

Asbtract ( 180 )   HTML ( 1)   PDF (273KB) ( 230 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

A sensing-then-transmission slot scheduling scheme for opportunistic spectrum access under the interference constraint of primary users (PU) was proposed.First,the throughput of the cognitive radio system (CRS) with channel handoff mechanism and the interference ratio to PU were deduced.An optimization model combined the sensing time and sensing period was also set up.Then,the optimal sensing time and sensing period were derived by maximizing the throughput of the CRS subject to both interference ratio and sensing qua ity constraints.Based on the analysis,a time slot scheduling scheme was finally proposed.The achievable maximum throughput of the CRS was derived too.Simulation results indicate that the proposed slot scheduling scheme can enhance the throughput of the CRS,and fit the changes of interference ratio and sensing quality constraints better.

Personalized requirements oriented trustworthy services recommendation based on social network

Pei-yun ZHANG,En-hong CHEN,Bo HUANG

2013, 34(12):  49-59.  doi:10.3969/j.issn.1000-436X.2013.12.006

Asbtract ( 264 )   HTML ( 3)   PDF (219KB) ( 581 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In current background of service computing and society computing,it is difficult to obtain the trustworthy Web services to meet users' personalized requirements.To olve the problem,the trustworthy services recommendation model was proposed which orients personalized requirements based on socia twork.The decomposing and matching algorithm was proposed for users' personalized requirements,and the algorithm utilizes the semantic dictionary to improve the se-mantic match of users' requirements.Based on personal zed functional requirements and social trust values of network nodes and the services' direct trust values and indirect trust values,a trustworthy service recommendation algorithm was designed which meets the personalized requirements.By analyzing the correlation among the social network and services' trust values,the method can improve the performance of trustworthy services collaborative recommendation.The analysis of algorithms and experiments' results show that the approach is effective and feasible.

VS-split load balancing algorithm in DHT-based P2P systems

Lei YANG,Ren-fa LI,Shi LIU,Zhi-xing CHEN,Ken-li LI

2013, 34(12):  60-70.  doi:10.3969/j.issn.1000-436X.2013.12.007

Asbtract ( 280 )   HTML ( 3)   PDF (216KB) ( 405 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The method using virtual servers for balancing the load in DHT-Based P2P systems was studied.The load distribution in DHT-based overlay network using virtual servers were modeled,the occurrence probability of single vir-tual server problem (SVSP) was analyzed and computed in details,and a novel VS-split load balancing algorithm (VSSLBA) was proposed to deal with the SVSP by splitting virtual server into small ones.Simulations verify the cor-rectness of probability analysis and the performance of VSSLBA.

Ternary polynomial based key management scheme for wireless sensor network

Zhi-tao GUAN,Yue XU,Jun WU

2013, 34(12):  71-78.  doi:10.3969/j.issn.1000-436X.2013.12.008

Asbtract ( 274 )   HTML ( 2)   PDF (129KB) ( 508 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

A ternary polynomial based key management (KMTP) scheme was proposed,which is effective in cluster based wireless sensor networks.Firstly,the base station will give each node one unique identi ier to ensure the validity of the node.Then,algorithm of the inner-cluster and inter-cluster key pre-distribution based on the ternary polynomial of the same order was stated,which can ensure the value of the cracking threshold is bigger than the number of nodes of a clus-ter and all clusters separately,which means it's very hard to be cracked even all nodes of a cluster or all clusters are com-promised.To assure the communication security,inter-cluster multi-hop routing mechanism was designed based on con-structing secure conjunct neighbor table.Finally,the updating parameter and the updating authentication number were in-troduced in rekeying phase.The analysis shows that the proposed scheme can meet the security requirement of key man-agement,and it also has less computation cost and storage cost than the existing schemes.

Survivable virtual network mapping with differentiated services QoP

Guang-yuan LIU,Kai SHUANG,Sen SU

2013, 34(12):  79-83.  doi:10.3969/j.issn.1000-436X.2013.12.009

Asbtract ( 240 )   HTML ( 1)   PDF (76KB) ( 878 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The protection needs of virtual network are more and more diversified.So how to realize survival virtual net-work mapping of different needs is the important issue.The mathematical model of differentiated services QoP issue was formulated.Then a heuristic algorithm was proposed to solve it.The simulation results demonstrate the algorithm can sa-tisfy various virtual networks protection for users and decline the substrate network bandwidth cost compared with other algorithms.

Research and design of hybrid scheduling system of node operating system for cyber physical systems

Xiao-zhou DU,Chen-hong CAO,Jian-zhong QIAO,Shu-kuan LIN

2013, 34(12):  84-93.  doi:10.3969/j.issn.1000-436X.2013.12.010

Asbtract ( 295 )   HTML ( 3)   PDF (237KB) ( 693 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In accordance with the new characteristics of physical,internet and intelligence of CPS,a hybrid scheduling system was designed for CPS node operating system with real-time performance,flexibility,and adaptability.Based on the in-depth study of the intrinsic differences between event-driven scheduling mechanism and multithreaded one,a hybrid programming model was designed that combines collaborative multithread and hybrid stack management using adapter mechanism.A hybrid scheduling model was also designed that supports both event and thread task types.Reasonable real-time algorithms were implemented for verification.Experimental results show that the real-time per-formance of the scheduling system is obviously better than that of TinyOS,thus it is better suited for CPS than TinyOS does.

Super group peer selection algorithm in hybrid layered P2P streaming system based on evolutionary game

Jing CHEN,Bing ZHOU,Chen-shan REN,Xiao-she DONG

2013, 34(12):  94-103.  doi:10.3969/j.issn.1000-436X.2013.12.011

Asbtract ( 229 )   HTML ( 2)   PDF (191KB) ( 394 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Due to high efficiency and good scalability,hybrid layered P2P architecture receives more and more impor-tance in P2P research and P2P streaming application fields recently.The blems about super peers selection are highly challenging because super peers must be selected from huge and dynamically changing network.A distributed super group peer(SGP) selection algorithm based on evolutionary game is proposed in this paper,in hich the SGP selection procedure is modeled based on evolutionary game framework firstly and its evolutionarily Moreover,a distributed Q-Learning algorithm (ESS-SGP) according to the mixed strategies by analysis is proposed for the peers to converge to the ESSs based on their own payoff history.Compared to the traditional random selected SGP P2P schemes,experiments results show that the proposed ESS-SGP algorithm achieves much better performance in terms of social welfare,average upload rate of SGPs,and keeps the upload capacity of the P2P streaming system increasing steadily with the number of peers increasing.

Multi-cell multi-user downlink beam forming design considering transceiver impairment

Can-feng XIA,Shi-wen HE,Yong-ming HUANG,Hai-ming WANG,Lv-xi YANG

2013, 34(12):  104-112.  doi:10.3969/j.issn.1000-436X.2013.12.012

Asbtract ( 224 )   HTML ( 2)   PDF (271KB) ( 467 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Multi-cell multi-user downlink beamforming schemes considering transceiver impairments were developed,with the objective of minimizing the maximum mean square error (MSE) and the sum of MSE of all users.The optimiza-tion problem was transformed into a second-order cone programming(SOCP) standard form,and a hierarchical iterative algorithm was developed to solve the original problem.Compared with the traditional beamforming methods developed for ideal transceivers,numerical results show that our proposed optimization algorithms greatly reduce the impact of im-pairments and improve the system performance.

Technical Report

Downlink hierarchical scheduling method for relay-enhanced cellular network based on TG-PF

Hua QU,Ya-lan SONG,Ji-hong ZHAO,Yi-wei LI

2013, 34(12):  113-119.  doi:10.3969/j.issn.1000-436X.2013.12.013

Asbtract ( 202 )   HTML ( 2)   PDF (162KB) ( 310 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Hierarchical scheduling model for relay-enhanced cellular network was designed to divide the w e process of resource scheduling in relay-enhanced cell into two parts,which are high-level scheduling accomplished by eNode B and low-level scheduling accomplished by Relay Node.Based on this,a throughput guaranteed proportional fair (TG-PF) scheduling algorithm was also proposed.The algorithm fully takes the advantages of Max C/I algorithm and PF algorithm consideration,so that it can effectively improve cell lay-enhanced cellular network based on TG-PF,PF algorithm and TG-PF algorithm are separately used in high-level and low-level scheduling process to satisfy the system throughput and scheduling fairness requirements of relay-enhanced cell.Simulation results suggest that the proposed method can tremendously improve cell edge throughput with a small size sa-crifice in scheduling fairness.

Improved information hiding algorithm based on twice positioning in coding channel

Li-quan CHEN,Miao LU,Ai-qun HU,Xiao-yan SUN

2013, 34(12):  120-130.  doi:10.3969/j.issn.1000-436X.2013.12.014

Asbtract ( 338 )   HTML ( 7)   PDF (274KB) ( 536 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Based on analysis of existing information hiding algorithms and capacities in coding channel,the defects of those existing algorithms were pointed out,and an improved information hiding algorithm with capacity upper limit analysis and based on twice positioning was proposed.the proposed algorithm,the information hiding capacity upper limit is firstly computed before embedding,and used for comparison with the size of hiding secret data.Then,twice posi-tioning scheme and cyclic shift mechanism for choosing the embedding position were used in the proposed algorithm.Simulation results show that,compared with those existing algorithms,the proposed algorithm is able to avoid the risk of exposure of the secret channel caused by capacity exceeding of hiding data.It can not only improve the randomness of the embedding position,but also avoid the risk of overlap among different secret data.The hiding performances of the coding channel information hiding system are improved.

Cloud computing node based on Loongson multi-core CPU

Li RUAN,Guang-jun QIN,Li-min XIAO,Ming-fa ZHU

2013, 34(12):  131-141.  doi:10.3969/j.issn.1000-436X.2013.12.015

Asbtract ( 263 )   HTML ( 3)   PDF (209KB) ( 521 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

A cloud computing node based on Loongson multi-core CPUs was introduced.Both the design and implemen-tation methods of its hardware and software were introduced.Moreover,the prototype was successfully implemented.Experimental results show that one computing node of this system can achieve the performance of 0.256×10＜sup＞12＜/sup＞TFlops per second,42 1U ranks and 672 CPUs with 2688 CPU cores (672×4) in one cabinet.i.e.it is competing in metrics like Loongson multi-core CPU based,high performance with low power,high integration,etc.and lays a solid foundation for Loongson multi-core CPUs based cloud computing systems.

End hopping based on message tampering

Kai LIN,Chun-fu JIA

2013, 34(12):  142-148.  doi:10.3969/j.issn.1000-436X.2013.12.016

Asbtract ( 259 )   HTML ( 7)   PDF (120KB) ( 749 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The main theoretical and technical problems were studied in the application of end hopping,and the idea of message tampering was proposed and the model of end hopping stack was built upon it.Three feasible implementations for the end hopping and discuss their advantages and disadvantages were respectively provided.And the experiment re-sult shows the potential application of the message tampering and the end hopping stack model in practice.

Experimental research on a direct-detection optical OFDM system based on cascaded variable coefficient training sequences and pre-emphasis technique

Wei WU,Jiang-nan XIAO,Ming CHEN,Jian-jun YU,Lin CHEN

2013, 34(12):  149-157.  doi:10.3969/j.issn.1000-436X.2013.12.017

Asbtract ( 276 )   HTML ( 1)   PDF (248KB) ( 876 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Frequency selective fading (FF)caused by digital to analog converter and optical fibe dispersion in optical fiber link was analysed,and signal-signal mixing interference (SSMI) when the orthogonal frequency division multiplex-ing (OFDM) signal was detected by PIN after transmitting over the fiber in direct-detection optical OFDM system.The cascaded variable coefficient training sequences and pre-emphasized configuration of O-OFDM signal were proposed to decrease the influence of the SSMI and FF.The theoretical analysis and experimental results show that the receiving sen-sitivity of the cascaded variable coefficient training sequences and pre-emphasized configuration of OFDM signal is higher than the interleaved configuration of training ed that the fiber transmission performance of the cascaded variable coefficient training sequences and pre-emphasized configuration of O-OFDM signal is better than the interleaved configuration of training sequence of OFDM signal in ro-bust tolerance against SSMI and FF.

Overview

Overview of the data security and privacy-preserving of mobile cloud services

Rui-xuan LI,Xin-hua DONG,Xi-wu GU,Wan-wan ZHOU,Cong WANG

2013, 34(12):  158-166.  doi:10.3969/j.issn.1000-436X.2013.12.018

Asbtract ( 447 )   HTML ( 25)   PDF (118KB) ( 1825 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Compared with the traditional cloud services,the mobi cloud services have the characteristics of the mobile interconnected,flexible end-user applications and convenient data access.However,the rich applications of mobile cloud services bring more security and privacy-leaking problems.The basic concepts,applications and security issues of mobile cloud services were shaded light on,and then describes the architecture of security and privacy-preserving was described,some key technologies were studied in this domain,mainly focusing on the research progress of the security protocols and authentication,access control,integrity verification,mobile trusted computing and the privacy-preserving based on encryption,anonymous and confusion,and the advantages and disadvantages were pointed out respectively.Finally,some future research direction were given at the end.

Academic communication

Building mechanism of trusted virtual domain via the VPE

Li-na WANG,Hao ZHANG,Rong-wei YU,Han-jun GAO,Ning GAN

2013, 34(12):  167-177.  doi:10.3969/j.issn.1000-436X.2013.12.019

Asbtract ( 213 )   HTML ( 1)   PDF (244KB) ( 739 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Due to lack of flexible networking control,most exiting trusted virtual domain deployment approaches fail to provide elastic and secure interconnection.A trusted virtual domain architecture TVD-VPE was proposed in cloud com-puting enterprises which greatly enhances sensitive data protection.TVD-VPE constructs a virtual private ethernet based on separate device driver,VPE captures network packets at the backend driver and checks whether the packets comply with border security strategy,and data frames are encrypted among trusted virtual domains to ensure the security of sensi-tive data.Simultaneously,four protocols were proposed,TVDJOP/TVDEXP protocol for any new VM joining in or exit-ing TVD securely,TVDMP protocol for deploying border cross-domain access.Finally,the prototype system and tests of its functionality and performance were implemented.The experiment results reveal that the architecture can effectively prevent unauthorized access between these trusted virtual domains,while introduces little overhead to Xen network performance.

Low-complexity and QoS-awared scheduling algorithm for LTE

Fei-long ZHAO

2013, 34(12):  178-184.  doi:10.3969/j.issn.1000-436X.2013.12.020

Asbtract ( 202 )   HTML ( 2)   PDF (134KB) ( 424 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Based on the relaxation based linearization principle and the greedy algorithm,a scheduling algorithm named QPF was designed with low complexity and full QoS-awared.It is divided into two parts:time and frequency domain,which can allocate all resources at once according to two service types:GBR and Non-GBR service.By this way the computation complexity decreases from O(M^N)to O(M lb N).The proposed algorithm has a balanced performance in si-mulation,which can get 10% improvement in some dimensions compared with the references with high load and provide a nice trade-off between the performance and complexity.