Advanced Search

2004 Vol. 30, No. 11

  Previous Issue | Next Issue 
Display Method:
Research and design of a datalink user's safeguard strategy
Li Qianmu, You Jing, Zhang Hong, Liu Fengyu
2004, 30(11): 1029-1032.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 641
Abstract:
To solve the datalink problem of capability and service/price, a frame work model——a new datalink user's redundant strategy was proposed, based on reliability theory. The redundant strategy could be combined with the normal maintenance, which greatly improved the performance of the network system. The static-checking and policy of authentication mechanism ensured the running network without any error. The redundant equipments were independent but were capable of communication with each other when they work their actions. The model was independent of specific application environment, providing a general-purpose framework for fault diagnosis. An example was given to express the calculating method.
Study of new measure to recover and control DOS/DDOS atta ck
Fang Zhiyi, Zhang Meiwen, Wei Hua, Wang Wei
2004, 30(11): 1033-1037.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1092
Abstract:
DOS(denial\|of\|service)/DDOS(distributed denial\|of\|service) network attack no t only causes harm to attacked target, but also disturbs other flows that share the same network with attacked target. By adding computing into every bosom node (route, switch), a mechanism system to recover and control DOS/DDOS attack which based on active network was advanced. The mechanism system was composed of three mechanisms: cluster-based automatic identification and control mechanism, cluster-based active notify trace mechanism and administration domain based control cooperation mechanism. Cluster-based automatic identification and control mechanism included identification policy of attack cluster and rate-limit policy of controlling them. Cluster-based active notify trace mechanism will notify the characteristic of attack cluster to upstream active node and activate local rate-limit policy. Effective recovery and the control o f DOS/DDOS attack can be realized by using this system at lab.
Web services enabled VHE platform
Wan Neng, Chen Junliang, Su Sen
2004, 30(11): 1038-1042.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 620
Abstract:
VHE(virtual home environment) is an important concept for the 3rd gen eration mobile communication networks, which enables PSE(personal service environment) mobility across network boundaries and between terminals. A service-oriented VHE platform, WSEVHEP(Web services enabled VHE platform), was presented, which leverages Web services techniqu es for its architecture. The system requirements of WSEVHEP were discussed. A framework for service provisioning was proposed, in which the platform itself and the 3rd-party services were all published as Web services. Service work flows, such as service discovering, subscribing and executing, were all advanced. The platform realizes the concept of VHE features and satisfies the mobility requirements of users. It is useful to develop VHE platform on mobile Internet.
Analysis of model of concurrent multimedia workloads retrieving data form storage systems
Li Zhong, Wang Gang, Liu Jing
2004, 30(11): 1043-1047.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1070
Abstract:
The multimedia applications need the guarantee of the QoS(quality of servi ce) when retrieving data form the storage systems, the fraction of requests whose response time exceeds a specific delay limit must below a certain proport ion. In the concurrent multimedia workloads, many applications access a storage system at the same time, Which makes it difficult to satisfy the QoS requireme nts. The aggregate requests of concurrent multimedia applications Poisson arrive the storage system, the storage system service time is exponential distribution, su ch stochastic process was abstracted as an M/M/1 queuing model. Based on the que uing model, the relationship between the number of concurrent multimedia applica tions and the statistical distribution of the requests response time was determi ned, the relationship can be used to guarantee the QoS of concurrent multimedia applications. The experimental results validate the novel model is accurate.
Research on Web information fusion infrastructure
Li Bing, He Keqing, Liu Jin, Li Ruixuan
2004, 30(11): 1048-1051.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 819
Abstract:
Meta modeling is an effective approach to implement interoperability a mong distributed and heterogeneous information sources on Web. MMF (meta model f ramework) is a set of meta objects and meta modeling constructs to be used in th e development a metamodel in actual implementation of information registry and r epository. An architecture of Web information fusion infrastructure was presente d based on FDBS (federated database system) and MMF. A relevant process model de pended on semantic Web services and software agent was discussed. A modeling met hod, named ontology-object, was proposed in order to simplify design and develo pment of the Web information fusion infrastructure.
Resource dependency based grid resources assignment
Guo Wencai, Yang Yang, Liu Li
2004, 30(11): 1052-1056.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 740
Abstract:
As a component of a grid job, a grid service is required to select suit able resource from global wide providers so that the different resource assignme nts will have distinctive impact on the utility of the job,which results in SRA(service resource assignment) problem.Addressing it a novel method TDRA (tree decomposition based resource allocation algorithm), based on tree- decomposition of grid resource dependency graph ,was choosed to form the correct vertex elemination order on which the variables were iteratively relaxed untill the optimised result was reached. The method indicates the complexity of TDRA i s polynominal instead of exponential,which proves available to solve SRA problem s. As a conclusion the further work of this research was also discussed.
IPSec interworking issues under mobile wireless network scenarios
Ye Runguo, Feng Yanjun, Zhang Fangzhou, Song Cheng
2004, 30(11): 1057-1060.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1192
Abstract:
IPSec protocol can provide end-to-end security for IP-based applicati ons in fixed wired networks; IPsec can also be used to provide communication sec urity for numerous mobile wireless terminal devices in future mobile wireless ne tworks. Interworking issues between IPSec protocol and other network pr otocols should be solved before its wide deployment.The interworking issues betw een IPSec and performance enhancing protocols and mobility management protocols under mobile wireless network scenarios were analyzed, followed by some practica l solutions. The issues related to the current IPSec architecture in mobile wire less networks were analyzed, the development trend of next generation IPSec arch itecture was pointed out.
Research of middleware technology for next generation network
Zhang Yunyong, Zhang Zhijiang, Liu Yunjie, Li Guojie
2004, 30(11): 1061-1065.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1029
Abstract:
NGN(next generation network) will be a uniform and integrated netw ork which imports distributed computing and communication technology. NGN h ave multiple kinds of applications and network environment. Middleware for NGN w as studied and practiced. Traditional middleware and mobile middleware were inte grated together to provide uniform platform. End-to-end quality of service tec hn ology, context aware technology and reflective technology were used so as to ach ieve flexible programming model (beyond C/S model) and self-adaptive ability in dynamic environment. NGN middleware can provide applications with different QoS( quality of service) and can minimize the cost for developing, deploying and mana ging network application.
Wireless video streaming oriented packet scheduling algorithm
Niu Jianwei, He Rui, Hu Jianping
2004, 30(11): 1066-1070.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1015
Abstract:
Based on end-system, an end-to-end wireless video streaming architecture was proposed. The video frame importance was defined in terms of the contribution of the VP(video packet) to the decompression video quality within a GOP(Group Of Picture). Based on this concept, a WVSPSA(wireless video streaming oriented pac ket scheduling algorithm) was proposed, which could dynamically re-order the tr ansmission schedule of the VPs according to the VPs' importance. Simulation resu lts show that the video streaming architecture is practical, and WVSPSA algorith m outperforms the prevalent EDF algorithm.
Communication model for data-intensive grid
Chen Qingkui, Na Lichun
2004, 30(11): 1071-1075.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 826
Abstract:
For resolving the conflict problem of the peak communication in data-intensive grid that composed of CC(computer clusters), a MGCM(multi-level grouping commun ica tion model) was discussed. A series of formal definitions, such as the DGDN(dat a-intensive grid data node) and the complete graphics communication wer e given. Using the mechanism of the active queue and passive queue, a GCA(group ing communication algorithm) in single computer cluster and a GCAD(grouping communication algorithm in double computer clusters) were studied. B ased on netw ork distance, GCA and GCAD,a MGCM of DGDN were designed. The analysis and experi m ent results show that MGCM effectively resolve the low efficiency during the pea k communication in grid. It can be fit for the data-intensive parallel computin g based on grid.
Monitoring grid on service level, GSLMA and its test bed
He Chuan, Gu Lei, Du Bin, Wu Jianfeng
2004, 30(11): 1076-1079.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 793
Abstract:
Grid monitoring toolkits was classified into three categories from the point the view of architecture and a novel architecture of grid monitoring, GSLMA(grid s ervice level monitoring architecture) was presented,which was based on SLA(serv ice level agreement). GSLMA was a SOA(service oriented architecture) with high flexibility and scalability. It supported third-party monitoring and cou ld be extended to monitor emerging grid services without much additional program ming. In the end, several key technologies were introduced in implementation of a test bed of GSLMA, such as: extending functionalities of the OGSA(open grid se rvice architecture) container, using WSLA(Web service level agreement) to descri be the monitoring information and etc.
Hybrid genetic algorithm for independent tasks scheduling in heterogeneous computing systems
Zhong Yiwen, Yang Jiangang
2004, 30(11): 1080-1083.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1053
Abstract:
Efficient tasks scheduling is critical for achieving high performa nce in heterogeneous computing systems. The tasks scheduling problem is NP-hard in general. In order to obtain better solutions, many scheduling heuristics were presented in the literature. Based on genetic algorithm and MCT(minimum complet ion time) algorithm, a new hybrid genetic algorithm was presented for independe nt tasks scheduling in heterogeneous computing systems. Genetic algorithm was us ed to evolve a priority queue first, and then the priority queue was mapped to a schedule using MCT algorithm. The simulation results comparing with other sched uling algorithm show that it produces better results in terms of schedule length , and it has good convergent speed.
Complex wavelet packet based MIMO-OFDM wireless system
Xiao Zhengrong, Yu Zhi, Zhao Shaogang, Wu Weiling
2004, 30(11): 1084-1087.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 729
Abstract:
To provide high speed data service in frequency selective fading channels, a new MIMO-OFDM(multi-input multi-output-orthogonal frequency division multiplexi ng) system was proposed. The system used CWPT(complex wavelet packet transform) instead of FFT(fast Fourier transform) to realize OFDM. For t he property of complex wavelet packet functions, simulation of a MIMO-OFDM syst e m with two users show that, the CWPT based MIMO-OFDM system is better than the F FT based MIMO-OFDM system, but the CWPT based MIMO-OFDM scheme has some additi onal complexity.
Quantum algorithm analysis of knapsack problem
Lü Xin, Feng Denggu
2004, 30(11): 1088-1091.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1572
Abstract:
Speeding up knapsack problem, one of the NP complete problems, which could be us ed to design public-key cryptosy stems, was discussed using quantum algorithm. How to use Grover's quantum search ing algorithm to speed up the knapsack problem was talked about based on computa tional complexity theory. Comparisons of quantum searching algorithm with Shor's factoring algorithm were delivered and the factors that affected the performanc e of quantum algorithms were discussed from group theory point of view. The futu re of the quantum algorithms was also augmented in the later.
Model of network fault diagnosis
Chen Lin, Huang Jie, Gong Zhenghu
2004, 30(11): 1092-1096.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1046
Abstract:
Network diagnosis problem aims to obtain compatible fault mode which c an explain symptoms by a set of actions. Some diagnosis models have been propose d, but their descriptions of the problem with dependent actions were not accurat e enough and the results are not very optimal. A DBN(diagnosis Bayesian network) model was presented that consisted of symptoms nodes, fault hypothesis nodes, d ia gnosis action nodes and observation nodes. It combined the general Bayesian netw ork and the requirements of fault diagnosis. Under the assumption of independent diagnosis process, a fault diagnosis algorithm based on DBN model was proposed. The algorithm took dependent actions into account. Observation nodes were introduced to achieve lower diagnosis cost. Experiments show that the fault diagnosis method based on DBN can reduce the diagnostic cost effectively and sol ve diagnosis problem under dependent actions condition preferably.
Job scheduling algorithm based on credit model in cluster environment
Zhang Shudong, Cao Yu, a, Liao Lejian
2004, 30(11): 1097-1100.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 881
Abstract:
A novel job scheduling algorithm based on credit model was presented for cluster system. In the credit model, every computing node was looked as a c redit entity. The scheduling system gave each entity a credit value according it s power ability, its status and its history records. The scheduling system also gave a credit estimate to the entire cluster, and according this estimate and en tering job request decided to reject or accept the job. It was realized based on OpenPBS framework. It was also simulated using GridSim toolkits to proceed a co ntrastive test with FIFO(first input first output) and Libra. The result shows that it works well.
New evolutionary neural networks
Gao Wei
2004, 30(11): 1101-1105.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1782
Abstract:
Nowadays, the genetic algorithm is a main evolutionary algorithm in ev olutionary neural network study. But the previous researches show that, this kin d of evolutionary neural network model have many shortcomings, such as genetic c ode problem, genetic operation problem and restriction on structure of neural ne twork, \%et al\%. And the evolutionary programming is a good method. To overcome the shortcomings of traditional evolutionary programming, combining the immunized e volutionary programming proposed by author and BP neural network, a new evolutio nary neural network model whose architecture and connection weights evolve simul taneously was proposed. Through the typical XOR problem, the new model was compa red and analyzed with BP neural network and traditional evolutionary neural netw ork. The computing results show that the precision and efficiency of the new mod el are all good.
Method of cooperative detecting DDoS attacks across multiple domains
Su Heng, Ju Jiubin
2004, 30(11): 1106-1110.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 688
Abstract:
To prevent the DDoS(distributed denial of service) attacks effectivel y, a cooperative detection model was proposed based on the cooperation among the IDSs(intrusion detection systems) distributed in multiple administrative domai ns. Surrounding some valuable network assets, the enclosed defense ring was set up that consists of 〈IDS, Router〉 pairs with the IDS monitoring specific route r traffic. The IDSs reside in the ring were allotted to a cooperation group. With the information exchanging and alert correlating within the group, the signature s of DDoS attacks aimed at the network assets could be captured timely before th e overwhelming attack flooding aggregates. The construction method of cooperatio n rings, the information exchange mode, alerts correlation method and infrastruc ture of cooperative IDS entity were proposed. Some experiments were conducted wi th the MDCI(multiple domains cooperative intrusion-detection) system, a protot ype system. Results show that the prototype improves detection performance effec tively.
Loop unrolling optimization for software pipelining
Li Wenlong, Liu Li, Tang Zhizhong
2004, 30(11): 1111-1115.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 2658
Abstract:
Loop unrolling can m ake software pipelining achieve fractional initiation interval and improve resou rce utilization. Optimizations based on unrolling can reduce resource requiremen ts and the heights of critical paths. An algorithm named UTBPC(unrolling times based program characteristics) for determining unrolling factors and unrol ling b ased optimization for software data prefetching were proposed. These optimizatio ns were implemented in ORC(open research compiler), and SPEC CPU2000 benchmarks were tested in Itanium processor. The average perf ormance was improv ed by 2.6%. The results show that UTBPC algorithm and unrolling base d optimization for software data prefetching can improve the overall performance of compilers.
Flow control algorithm based on ECN probabilistic marking
Yang Zhen, Ding Wei
2004, 30(11): 1116-1120.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 1095
Abstract:
A flow model upon users' utility function was proposed, with which the optimization-based congestion control could transform optimization-based flow c ontrol. A dual gradient projection algorithm was given to optimize network flow control. In this algorithm, sources selected transmission rates that maximize th eir own benefits, utility minus bandwidth cost, and network links adjusted conge stion prices to coordinate the sources' decision. To implement this network flow control algorithm, the routers set ECN(explicit congestion notification) bit p robabilistically to transmit congestion price information. And the necessary lin k position information of packet path was estimated by exploiting the existing s emantics of the TTL(time-to-live) field in IP header. The computer simulation verified the convergence and validity of this flow control algorithm.
Load balance for network parallel computing
Hu Kai
2004, 30(11): 1121-1124.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 903
Abstract:
The principle of the dynamic load balance for a distributed system was used to design a preemptive dynamic load balancing architecture to meet require ments of the network distributed and parallel computing environment. A centraliz ed information system was introduced to collect all process's information. A hyb rid transmission control method of combining centralized and distributed logic w as described based on the process migrating. A new load index was particularly d efined for the case to make sure a processor's load level. The corresponding the oretic analysis for the important problems in implementation was discussed. The critical strategies and algorithms were presented. The results of the simulation show that the architecture and related strategies can provide effectively suppo rt to the network parallel computing.
Modeling trust with uncertainty for open networks
He Rui, Niu Jianwei, Hu Jianping
2004, 30(11): 1125-1128.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 774
Abstract:
Since trust relationships are uncertain in open networks, it is necess ary to consider uncertainty when modeling trust.A novel trust model was proposed based on the cloud model, which was a useful tool to describe uncertain concept . Modeling trust with cloud, the proposed model could present trust degree and u ncertainty degree in a uniform form. Based on such a definition of trust, algori thms to compute propagated trust clouds and aggregated trust clouds were propose d, which were needed in the case of trust reasoning. With simulation experiment s, the proposed model were compared with the other two trust models. The results show the cloud-based trust model performs much better in open networks, especi ally in helping entities to cooperate with each other.
HPI based multi-processor system for neural network image matching
Shi Zhenghao, Feng Yaning, Zhang Suinan, Huang Shitan
2004, 30(11): 1129-1132.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 870
Abstract:
Aim at the characteristics of embedded image matching, TMS320C6X family processo rs were used for parallel neural computing units, and a multi-processo r system structure based on TMS320C6X HPI(host parasite interface) interlinkin g for neural network image matching was presented. In this parallel computing sy stem, one host computer and three DSP neural units are contained, the master com puter is directly linked with each neural computing unit by HPI, memory of each neural unit can be directly accessed by the master computer through their HPI, t he real time image datum and the inter-results datum of each neural unit can be immediately directed transmit to each other between neural unit and the master c omputer. As a result, the real time performance of image matching is effectively improved according to theoretical analysis.
Research on self-adapted security cooperation framework based on peer-to-peer network
Chen Haitao, Pei Jinze, Hu Huaping, Gong Zhenghu
2004, 30(11): 1133-1136.
Abstract References Relative Articles Resource File Access Statistics iconViews $!{v.htmlCount} iconDownload 747
Abstract:
A new decentralized security cooperation framework based on P2P(peer-to -peer) network was proposed which supported many kinds of security service such as se cu rity management, the storage and sharing of security related data and distributi on of security information.A security management method which supports the robot icized construction of management structure and self-configuration of security c omponents under large-scale and dynamic network environment by the mixture of h i berarchy and peer-to-peer structure was presented. The dynamic change of secur it y components and requirements of security cooperation is self-adapted with the h elp of dynamic property of peer-to-peer architecture. Its decentralized archit ec ture eliminates the performance and bandwidth bottleneck of centralized manageme nt and has good survivability under node failure and network attack.

Copyright © Journal of Beijing University of Aeronautics and Astronautics

Address: Editorial Department of Journal of Beijing University of Aeronautics and Astronautics, 37 Xueyuan Road, Haidian District, Beijing Post Code: 100191 Email: jbuaa@buaa.edu.cn

Supported by: Beijing Renhe Information Technology Co., Ltd.