Pub Date : 1992-04-01DOI: 10.1109/PCCC.1992.200542
J. Myoupo
The authors present a fully pipelined modular linear systolic algorithm for dynamic programming. They use geometric considerations to design the array. Its characteristics make it fully-pipelined. It runs in 1.5n/sup 2/+O(n) time steps on an array of n/sup 2//6+O(n) processors.<>
{"title":"Solving dynamic programming problems efficiently on modular linear pipelines","authors":"J. Myoupo","doi":"10.1109/PCCC.1992.200542","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200542","url":null,"abstract":"The authors present a fully pipelined modular linear systolic algorithm for dynamic programming. They use geometric considerations to design the array. Its characteristics make it fully-pipelined. It runs in 1.5n/sup 2/+O(n) time steps on an array of n/sup 2//6+O(n) processors.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"4 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":"121110259","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.1992.200585
Y. Hao, Jyh-Charn S. Liu, Junguk L. Kim
The authors propose a scheme for global scheduling in a distributed system using a bus network, which is shared by nodes in the system through a contention-free medium access protocol. Unlike existing solutions, the scheme does not require complete global state information to be collected for a node that has an unschedulable aperiodic task to find a node that can execute the task. The nodes are organized in the system in a logical ring fashion to synchronize the global scheduling process. To evaluate the performance of the scheme, two protocols are developed. Simulation results show that the performance of the scheme is close to an ideal distributed system, in which the centralized global scheduler takes care of aperiodic tasks without any communication and task migration overhead.<>
{"title":"Task scheduling in hard real-time distributed systems on a bus local area network","authors":"Y. Hao, Jyh-Charn S. Liu, Junguk L. Kim","doi":"10.1109/PCCC.1992.200585","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200585","url":null,"abstract":"The authors propose a scheme for global scheduling in a distributed system using a bus network, which is shared by nodes in the system through a contention-free medium access protocol. Unlike existing solutions, the scheme does not require complete global state information to be collected for a node that has an unschedulable aperiodic task to find a node that can execute the task. The nodes are organized in the system in a logical ring fashion to synchronize the global scheduling process. To evaluate the performance of the scheme, two protocols are developed. Simulation results show that the performance of the scheme is close to an ideal distributed system, in which the centralized global scheduler takes care of aperiodic tasks without any communication and task migration overhead.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"15 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":"131187986","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.1992.200573
I. Habib, T. Saadawi
The authors present a framework for a flow control algorithm and analyze an effective access congestion control scheme that is applied at the input access node to the network. The scheme is based upon a preventive control principle and uses a feedback control signal that is a function of the buffer occupancy level to control the source coder sampling rate. The scheme avoids congestion and enhances the statistical multiplexing effect, consequently providing a high bandwidth utilization at a specific class of service. The scheme can be applied as a traffic shaping and enforcement function. The resulting quasi birth death queueing process was studied with state dependent arrivals using matrix-geometric techniques, and it was found that both MMPP/ER/1/K and the Poisson process approximations yield quite close results in an asynchronous transport model multiplexer with small buffer size and feedback control.<>
{"title":"Flow control of bursty voice traffic in broadband networks","authors":"I. Habib, T. Saadawi","doi":"10.1109/PCCC.1992.200573","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200573","url":null,"abstract":"The authors present a framework for a flow control algorithm and analyze an effective access congestion control scheme that is applied at the input access node to the network. The scheme is based upon a preventive control principle and uses a feedback control signal that is a function of the buffer occupancy level to control the source coder sampling rate. The scheme avoids congestion and enhances the statistical multiplexing effect, consequently providing a high bandwidth utilization at a specific class of service. The scheme can be applied as a traffic shaping and enforcement function. The resulting quasi birth death queueing process was studied with state dependent arrivals using matrix-geometric techniques, and it was found that both MMPP/ER/1/K and the Poisson process approximations yield quite close results in an asynchronous transport model multiplexer with small buffer size and feedback control.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"42 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":"134620254","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.1992.200525
Y. Shim, C. Ramamoorthy
Monitoring consists of collecting information from the system and detecting particular events and states using the collected information. These events and states can be symptoms for performance degradations, erroneous functions, suspicious activities, etc. and are subject to further analysis. Detecting events and states requires a specification language which makes it possible to express a wide variety of events/states easily and also allows the implementation of the efficient detection algorithm. The authors introduce an event/state specification language based on classical temporal logic. They compare the language with other languages such as relational algebra, historical relational algebra, OPS5, and an extended regular expression based language. It is proved that the language is as expressive as or more expressive than the other languages. It is also shown that the language can specify a variety of events and states more conveniently than the other languages.<>
{"title":"Specification of events and states in complex systems","authors":"Y. Shim, C. Ramamoorthy","doi":"10.1109/PCCC.1992.200525","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200525","url":null,"abstract":"Monitoring consists of collecting information from the system and detecting particular events and states using the collected information. These events and states can be symptoms for performance degradations, erroneous functions, suspicious activities, etc. and are subject to further analysis. Detecting events and states requires a specification language which makes it possible to express a wide variety of events/states easily and also allows the implementation of the efficient detection algorithm. The authors introduce an event/state specification language based on classical temporal logic. They compare the language with other languages such as relational algebra, historical relational algebra, OPS5, and an extended regular expression based language. It is proved that the language is as expressive as or more expressive than the other languages. It is also shown that the language can specify a variety of events and states more conveniently than the other languages.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"63 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":"132602365","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.1992.200524
K. Saleh
The author proposes a new approach for designing easily testable protocols. Testability requirements are identified and specified at the service definition stage, which, by synthesis or any other refinement or protocol design process, can be mapped onto the protocol design. The advantage of this approach is that it is much simpler to specify testability requirements at the service level than at the protocol design level. A simple example illustrating the approach is presented.<>
{"title":"Testability-directed service definitions and their synthesis","authors":"K. Saleh","doi":"10.1109/PCCC.1992.200524","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200524","url":null,"abstract":"The author proposes a new approach for designing easily testable protocols. Testability requirements are identified and specified at the service definition stage, which, by synthesis or any other refinement or protocol design process, can be mapped onto the protocol design. The advantage of this approach is that it is much simpler to specify testability requirements at the service level than at the protocol design level. A simple example illustrating the approach is presented.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"82 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131452551","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.1992.200592
E. Sha, L. Chao
The Diogenes model is a fault-tolerant model for many interconnection networks. The authors analyze a distributed reconfiguration algorithm for binary-tree structures based on this model. After a nonleaf node n in a binary tree has failed, the damage effect is minimized by substituting a leaf for n. The number of processing elements and the communication cost involved in the reconfiguration process are analyzed in the worst case and the average case. Several measures are defined for analyzing the performance of reconfigurable algorithms. It is shown that the reconfiguration algorithm can make the binary-tree structure become expected-case locally reconfigurable but not locally reconfigurable. By modifying the original Diogenes construction, the reconfiguration algorithm can achieve local reconfigurability. The performance analysis is presented.<>
{"title":"Efficient distributed reconfiguration for binary trees on Diogenes model","authors":"E. Sha, L. Chao","doi":"10.1109/PCCC.1992.200592","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200592","url":null,"abstract":"The Diogenes model is a fault-tolerant model for many interconnection networks. The authors analyze a distributed reconfiguration algorithm for binary-tree structures based on this model. After a nonleaf node n in a binary tree has failed, the damage effect is minimized by substituting a leaf for n. The number of processing elements and the communication cost involved in the reconfiguration process are analyzed in the worst case and the average case. Several measures are defined for analyzing the performance of reconfigurable algorithms. It is shown that the reconfiguration algorithm can make the binary-tree structure become expected-case locally reconfigurable but not locally reconfigurable. By modifying the original Diogenes construction, the reconfiguration algorithm can achieve local reconfigurability. The performance analysis is presented.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"76 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":"122940737","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.1992.200572
W. D. Shambroom
The TELNET and LAT protocols offer two alternatives for connecting terminals to UNIX hosts over an Ethernet network. The author presents an analysis of the characteristics of typical user input/output (I/O) traffic and of the two protocols, and suggests appropriate metrics for performance evaluation. The results of comparative measurements of the performance of specific implementations of these two protocols show that LAT can provide a significant performance advantage over TELNET.<>
{"title":"Connecting terminals to UNIX hosts over an Ethernet network: a performance comparison of the TELNET and LAT protocol options","authors":"W. D. Shambroom","doi":"10.1109/PCCC.1992.200572","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200572","url":null,"abstract":"The TELNET and LAT protocols offer two alternatives for connecting terminals to UNIX hosts over an Ethernet network. The author presents an analysis of the characteristics of typical user input/output (I/O) traffic and of the two protocols, and suggests appropriate metrics for performance evaluation. The results of comparative measurements of the performance of specific implementations of these two protocols show that LAT can provide a significant performance advantage over TELNET.<<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":"124894227","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.1992.200580
P. Green, L. Becker, R. J. Duckworth
The Activation Framework (AF) is a new software tool for the development of artificial intelligence (AI) applications which can execute on a geographically distributed heterogeneous network of communicating processors. The AF consists of tools for compiling groups of expert-system if. . .then. . .else. . . rules into executable code modules, for automatically generating C code modules from high-level system configuration descriptions, and for automatically generating makefiles for program compilation and linking on each processor. It also includes an operating system environment which integrates the estimated 70% of code which is common from one real-time AI application to the next. AF also includes support libraries which provide extensive functionality which is common from application to application.<>
{"title":"A software tool for developing distributed networked real-time artificial intelligence applications","authors":"P. Green, L. Becker, R. J. Duckworth","doi":"10.1109/PCCC.1992.200580","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200580","url":null,"abstract":"The Activation Framework (AF) is a new software tool for the development of artificial intelligence (AI) applications which can execute on a geographically distributed heterogeneous network of communicating processors. The AF consists of tools for compiling groups of expert-system if. . .then. . .else. . . rules into executable code modules, for automatically generating C code modules from high-level system configuration descriptions, and for automatically generating makefiles for program compilation and linking on each processor. It also includes an operating system environment which integrates the estimated 70% of code which is common from one real-time AI application to the next. AF also includes support libraries which provide extensive functionality which is common from application to application.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"30 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":"123576286","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.1992.200529
D. Saha, P. Dhar
A technique to improve the formal protocol conversion method suggested by K. Okumura (1986) is presented. The algorithm first constructs the reduced state transition graphs (STGs) of the protocols for which the converter is to be designed, and then follows the Okumura algorithm to produce a reduced communicating finite state machine (CFSM) model of the converter. This reduced machine is expanded to obtain the complete single CFSM converter. The reduction of STGs lowers the computational complexity of Okumura's algorithm by one or two orders of magnitude but does not disturb the desired functionality of the converter. The approach has been applied to design a converter between two example protocols.<>
{"title":"A fast protocol conversion technique using reduction of state transition graphs","authors":"D. Saha, P. Dhar","doi":"10.1109/PCCC.1992.200529","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200529","url":null,"abstract":"A technique to improve the formal protocol conversion method suggested by K. Okumura (1986) is presented. The algorithm first constructs the reduced state transition graphs (STGs) of the protocols for which the converter is to be designed, and then follows the Okumura algorithm to produce a reduced communicating finite state machine (CFSM) model of the converter. This reduced machine is expanded to obtain the complete single CFSM converter. The reduction of STGs lowers the computational complexity of Okumura's algorithm by one or two orders of magnitude but does not disturb the desired functionality of the converter. The approach has been applied to design a converter between two example protocols.<<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":"131498791","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.1992.200575
J. Vergeest, R. Vroom, T.W.T. De Waal
An approach to enhance the flexibility of communication between dissimilar CAD/CAM systems is presented. The proposed method involves techniques that have recently become available from STEP, the emerging ISO standard for the exchange of product data. The key principle of the approach is the specification of the entity domains of the interconnect CAD/CAM systems in the EXPRESS language. If a database system is applied as a central storage facility for the involved systems, then the required database schemes as well as the access software can be automatically derived from the EXPRESS definitions. Results from experiments with a data link between a commercial CAD/CAM system and a commercial database system are presented.<>
{"title":"Flexible communication between dissimilar CAD/CAM systems","authors":"J. Vergeest, R. Vroom, T.W.T. De Waal","doi":"10.1109/PCCC.1992.200575","DOIUrl":"https://doi.org/10.1109/PCCC.1992.200575","url":null,"abstract":"An approach to enhance the flexibility of communication between dissimilar CAD/CAM systems is presented. The proposed method involves techniques that have recently become available from STEP, the emerging ISO standard for the exchange of product data. The key principle of the approach is the specification of the entity domains of the interconnect CAD/CAM systems in the EXPRESS language. If a database system is applied as a central storage facility for the involved systems, then the required database schemes as well as the access software can be automatically derived from the EXPRESS definitions. Results from experiments with a data link between a commercial CAD/CAM system and a commercial database system are presented.<<ETX>>","PeriodicalId":250212,"journal":{"name":"Eleventh Annual International Phoenix Conference on Computers and Communication [1992 Conference Proceedings]","volume":"12 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":"130213426","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：