Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200533
Mike A. Csoppenszky, Arun Kumar Somani
The authors discuss the development and performance of a series of distributed routing algorithms for a synchronous circuit-switched hypercube and enhanced hypercube. Several centralized routing algorithms were developed. All the routing algorithms developed were derived from a single high-level concept of how routing can be accomplished in a circuit-switched hypercube. The routing algorithms were derived from the corresponding packet switching routing algorithm. The routing algorithms have different strengths and weaknesses when compared with one another, and these are considered. It is shown that the highest performing algorithm has an extremely low probability of blocking for the enhanced hypercube and the standard hypercube.<>
{"title":"Efficient distributed routing algorithms for a synchronous circuit-switched hypercube","authors":"Mike A. Csoppenszky, Arun Kumar Somani","doi":"10.1109/PCCC.1992.200533","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200533","url":null,"abstract":"The authors discuss the development and performance of a series of distributed routing algorithms for a synchronous circuit-switched hypercube and enhanced hypercube. Several centralized routing algorithms were developed. All the routing algorithms developed were derived from a single high-level concept of how routing can be accomplished in a circuit-switched hypercube. The routing algorithms were derived from the corresponding packet switching routing algorithm. The routing algorithms have different strengths and weaknesses when compared with one another, and these are considered. It is shown that the highest performing algorithm has an extremely low probability of blocking for the enhanced hypercube and the standard hypercube.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121522020","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200587
A. Amin
A communication network N in which links are subject to failures is typically modeled by a probabilistic graph G in which each edge exists independently with a fixed probability p, 0>
{"title":"On a class of reliable networks","authors":"A. Amin","doi":"10.1109/PCCC.1992.200587","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200587","url":null,"abstract":"A communication network N in which links are subject to failures is typically modeled by a probabilistic graph G in which each edge exists independently with a fixed probability p, 0<p<1. The pair-connected reliability of G is the expected number of connected pairs of vertices in G. The network N when link are overloaded and hence unavailable may be modeled by G with p approximately 0, while N subject to only link failures may be modeled by G with p approximately 1. Therefore, of interest are graphs which maximize the expected number of connected pairs of vertices for both p approximately 0 and p approximately 1. However, it is known that such graphs do not exist. Therefore, the author considers graphs which maximize the expected number of connected pairs of vertices for p approximately 0, among all graphs with specified edge connectivity and hence desired reliability for p approximately 1.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133520328","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200584
C. Han, K.-J. Lin, P. Tu
In many applications, a real-time job allows part of its computation to be finished after its primary deadline but before its extended deadline. The authors study the scheduling issues for real-time computations with extended deadlines. They call the part of a computation which must be completed before the primary deadline its hard part, and the rest its soft part. Two variations of this problem are studied. In the first model, the system receives a penalty for each late soft part. In the second model, there is an extra computation overhead for each late soft part. For each model, algorithms to produce a feasible schedule are presented.<>
{"title":"Scheduling real-time computations with extended deadlines","authors":"C. Han, K.-J. Lin, P. Tu","doi":"10.1109/PCCC.1992.200584","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200584","url":null,"abstract":"In many applications, a real-time job allows part of its computation to be finished after its primary deadline but before its extended deadline. The authors study the scheduling issues for real-time computations with extended deadlines. They call the part of a computation which must be completed before the primary deadline its hard part, and the rest its soft part. Two variations of this problem are studied. In the first model, the system receives a penalty for each late soft part. In the second model, there is an extra computation overhead for each late soft part. For each model, algorithms to produce a feasible schedule are presented.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131994953","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200586
H. Sholl, P. J. Pia, L. Lipsky
Through the use of a simulation study, the authors show how the effectiveness of a scheduling algorithm varies depending on the metric of interest and the execution environment under which it runs. They describe the timing requirements of real-time jobs. Three measures of response time controls are described. The model of a scheduling algorithm's execution environment is described, followed by a discussion on the variations in execution environment necessary to investigate adequately the behavior of a given scheduling algorithm. A simulation case study is described. A description of the modeling and simulation of a single processor running M job classes is included. Two scheduling algorithms for controlling response times are described, the results of running three scheduling algorithms under varying execution environments are presented.<>
{"title":"Performance evaluation of local real-time schedulers","authors":"H. Sholl, P. J. Pia, L. Lipsky","doi":"10.1109/PCCC.1992.200586","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200586","url":null,"abstract":"Through the use of a simulation study, the authors show how the effectiveness of a scheduling algorithm varies depending on the metric of interest and the execution environment under which it runs. They describe the timing requirements of real-time jobs. Three measures of response time controls are described. The model of a scheduling algorithm's execution environment is described, followed by a discussion on the variations in execution environment necessary to investigate adequately the behavior of a given scheduling algorithm. A simulation case study is described. A description of the modeling and simulation of a single processor running M job classes is included. Two scheduling algorithms for controlling response times are described, the results of running three scheduling algorithms under varying execution environments are presented.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130080560","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200538
S. Majumdar
Based on analytic models this research concentrates on basic issues that are important in the context of processor scheduling in multiprogrammed parallel systems. Local scheduling policies for fork and join jobs are analyzed. Demonstration of the appropriate multiprogramming level, and the computation of the number of processors to be allocated to a job are investigated in the context of global scheduling. The global scheduling approach is static in the sense that a fixed set of processors is allocated to a job and dynamic processor switching from one application to another is not allowed. Important insights into system behavior that are useful in the context of operating systems for parallel systems are obtained as a result of this analysis. A number of basic issues that concern the relationship between job characteristics, system performance, and effective scheduling are discussed.<>
{"title":"The performance of local and global scheduling strategies in multiprogrammed parallel systems","authors":"S. Majumdar","doi":"10.1109/PCCC.1992.200538","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200538","url":null,"abstract":"Based on analytic models this research concentrates on basic issues that are important in the context of processor scheduling in multiprogrammed parallel systems. Local scheduling policies for fork and join jobs are analyzed. Demonstration of the appropriate multiprogramming level, and the computation of the number of processors to be allocated to a job are investigated in the context of global scheduling. The global scheduling approach is static in the sense that a fixed set of processors is allocated to a job and dynamic processor switching from one application to another is not allowed. Important insights into system behavior that are useful in the context of operating systems for parallel systems are obtained as a result of this analysis. A number of basic issues that concern the relationship between job characteristics, system performance, and effective scheduling are discussed.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134369268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200521
M. Ooki, Y. Fukagawa, S. Murakami, S. Yoshida
The authors present an approximate analysis of an asymmetric token ring network with nonpreemptive priority under a new discipline named the go-back-to-higher-priority discipline. This discipline is described and the token ring network is analysed on the basis of the independent hypothesis for the distribution of the cycle time. Then, probabilities for the number of messages at the token's arrival, a mean cycle time, a mean intervisit time of the token and a mean waiting time of messages for each priority level are obtained. To calculate the mean waiting time for a lower priority level, a root of the denominator of the generating function plays an important role. In some numerical examples with simulation results, the mean waiting time for the go-back-to-higher-priority discipline is compared with that for an asymmetric token ring network with nonpreemptive priority under a one-limited discipline.<>
{"title":"An approximate analysis of a token ring network with nonpreemptive priority under go-back-to-higher-priority discipline","authors":"M. Ooki, Y. Fukagawa, S. Murakami, S. Yoshida","doi":"10.1109/PCCC.1992.200521","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200521","url":null,"abstract":"The authors present an approximate analysis of an asymmetric token ring network with nonpreemptive priority under a new discipline named the go-back-to-higher-priority discipline. This discipline is described and the token ring network is analysed on the basis of the independent hypothesis for the distribution of the cycle time. Then, probabilities for the number of messages at the token's arrival, a mean cycle time, a mean intervisit time of the token and a mean waiting time of messages for each priority level are obtained. To calculate the mean waiting time for a lower priority level, a root of the denominator of the generating function plays an important role. In some numerical examples with simulation results, the mean waiting time for the go-back-to-higher-priority discipline is compared with that for an asymmetric token ring network with nonpreemptive priority under a one-limited discipline.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133867963","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200512
E. Swartzlander, R. Jones
The authors provide a comparison of implementation approaches for digital neural networks. Digital neural networks of large size are feasible if the inputs and outputs are single-bit binary signals. A key component for this application is the parallel counter, which counts the number of inputs that are ONEs. Progress is reported toward the implementation of parallel counters with up to 1022 inputs, as required to realize multilayer neural networks with up to 1000 neurons per layer.<>
{"title":"Digital neural network implementation","authors":"E. Swartzlander, R. Jones","doi":"10.1109/PCCC.1992.200512","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200512","url":null,"abstract":"The authors provide a comparison of implementation approaches for digital neural networks. Digital neural networks of large size are feasible if the inputs and outputs are single-bit binary signals. A key component for this application is the parallel counter, which counts the number of inputs that are ONEs. Progress is reported toward the implementation of parallel counters with up to 1022 inputs, as required to realize multilayer neural networks with up to 1000 neurons per layer.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133983480","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200568
W. Zaumen, J. J. Garcia-Luna-Aceves
The steady-state response of link-state and loop-free distance-vector routing algorithms to multiple changes in the costs of links is investigated. A quantitative comparison of an ideal link-state algorithm similar to the one used in the open shortest path first (OSPF) and in the OSI intradomain routing protocol, and a new loop-free distance-vector algorithm, is made for several computer network topologies. A variety of quantities, including the length of messages and the average number of paths affected by routing loops, are computed as a function of time after a link or node change. Probabilities of various conditions, including the existence of loops, are also obtained as a function of time. The results show that in steady state, a loop-free distance-vector algorithm operates with essentially the same communication overhead as the ideal link state algorithm, and requires substantially fewer CPU cycles. The results also suggest that it may be possible to correlate the performance of the routing algorithms with various parameters that can be used to characterize networks.<>
{"title":"Steady-state response of shortest-path routing algorithms","authors":"W. Zaumen, J. J. Garcia-Luna-Aceves","doi":"10.1109/PCCC.1992.200568","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200568","url":null,"abstract":"The steady-state response of link-state and loop-free distance-vector routing algorithms to multiple changes in the costs of links is investigated. A quantitative comparison of an ideal link-state algorithm similar to the one used in the open shortest path first (OSPF) and in the OSI intradomain routing protocol, and a new loop-free distance-vector algorithm, is made for several computer network topologies. A variety of quantities, including the length of messages and the average number of paths affected by routing loops, are computed as a function of time after a link or node change. Probabilities of various conditions, including the existence of loops, are also obtained as a function of time. The results show that in steady state, a loop-free distance-vector algorithm operates with essentially the same communication overhead as the ideal link state algorithm, and requires substantially fewer CPU cycles. The results also suggest that it may be possible to correlate the performance of the routing algorithms with various parameters that can be used to characterize networks.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"132 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134117279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200556
H. Ellis, R. Ammar, S. Demurjian
The authors investigate the role of propagation in database support for performance-modeling environments (PMEs) and elaborate on the use of the active data model (ADAM) object-oriented design tool to design performance modeling applications that include propagation. ADAM was used to design a particular performance model including the complex interdependencies between the components of the model, thereby maintaining the consistency of data throughout the model. The analysis contains background on propagation and the performance models of interest, and a discussion of the benefits of propagation for PMEs is discussed by presenting a list of goals for the database support for PMEs and outlining two categories of propagations for performance-modeling data.<>
{"title":"The role of propagation in database support for performance-modeling environments","authors":"H. Ellis, R. Ammar, S. Demurjian","doi":"10.1109/PCCC.1992.200556","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200556","url":null,"abstract":"The authors investigate the role of propagation in database support for performance-modeling environments (PMEs) and elaborate on the use of the active data model (ADAM) object-oriented design tool to design performance modeling applications that include propagation. ADAM was used to design a particular performance model including the complex interdependencies between the components of the model, thereby maintaining the consistency of data throughout the model. The analysis contains background on propagation and the performance models of interest, and a discussion of the benefits of propagation for PMEs is discussed by presenting a list of goals for the database support for PMEs and outlining two categories of propagations for performance-modeling data.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130532031","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200569
Chinchen Chang, Shin-Jia Hwang
The authors propose a new finite ideal threshold scheme which provides Shannon perfect secrecy and can be used to handle the changeable master keys system in finite times without affecting any secret shadow. Since the idea was inspired by A. Shamir's (1979) threshold scheme, this threshold scheme is introduced. The definition of the finite ideal threshold scheme is given. The dynamic secret is described. The security analysis of the method is considered.<>
{"title":"Sharing a dynamic secret","authors":"Chinchen Chang, Shin-Jia Hwang","doi":"10.1109/PCCC.1992.200569","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200569","url":null,"abstract":"The authors propose a new finite ideal threshold scheme which provides Shannon perfect secrecy and can be used to handle the changeable master keys system in finite times without affecting any secret shadow. Since the idea was inspired by A. Shamir's (1979) threshold scheme, this threshold scheme is introduced. The definition of the finite ideal threshold scheme is given. The dynamic secret is described. The security analysis of the method is considered.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133179015","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：