Kilnam Chon

Testing methods for SDL systems

The aim of this tutorial is to present an overview of conformance testing methods for SDL systems. These methods can be classified into two main groups: those whose intent is the totally automated test generation from the SDL system specification and those that provide interactive test generation methods. We present some of the more representative methods illustrating their application by a common example. This example, the Inres protocol and service, will facilitate the comparison of the methods with respect to SDL and test notations, restrictions on the language, results of the application to the protocol and the tools supporting the methods.

Proactive Web caching with cumulative prefetching for large multimedia data

Abstract Web proxy caching provides an effective way to reduce access latency and bandwidth consumption, and thus improve performance of the World Wide Web. Due to the simplicity of incorporating various types of data into the Web, non-negligible access for large multimedia data is observed. Large multimedia data transfer, requiring a long-lived connection with consecutive data transmissions, often fails over wide area networks. Repetitive attempts to fetch large objects through proxy caches would waste network resources with no success guaranteed. In this paper, we propose on-demand cumulative prefetching to enhance proxy caches for better data availability of large multimedia files. Our analysis shows that on-demand cumulative prefetching can be performed within a few trials with minimal additional bandwidth requirement. The experimental results demonstrate that the proposed scheme substantially reduces error rate and increases byte hit rate for large multimedia data.

Bio-Mirror project for public bio-data distribution

Timely worldwide distribution of biosequence and bioinformatics data depends on high performance networking and advances in Internet transport methods. The Bio-Mirror project focuses on providing up-to-date distribution of this rapidly growing and changing data. It offers FTP, Web and Rsync access to many high-volume databanks from several sites around the world. Experiments with data grids and other methods offer future improvements in biology data distribution.

The future of the internet digital divide

< Technologically, Internet capacity will continue to increase into the foreseeable future. The available bandwidth, storage capacity, and processing capacity will grow from the current gigabits, gigabytes, and billions of instructions per second to tera-, peta-, exa-, zeta-measured amounts. The bandwidth available with deployment of dense wavelength wide division multiplexing (DWDM) and other emerging technologies promises transmission rates approaching terabits per second and beyond for the Internet backbone. User access rates to the Internet will grow to gigabits per second with deployment of Gigabit Ethernet, followed by 10-Gigabit Ethernet and other technologies to come. Sociologically, the Internet will be used by practically all human beings—several billion human users during the next century, a large increase from the current hundreds of millions of users. China will become the largest Internet user community in the coming decade, followed by India, which will become the second largest Internet user community toward the end of the decade. Many interesting developments are anticipated for the global Internet community when currently developing countries dominate the global Internet community, in contrast with the present-day Internet, which is dominated by the U.S. and other Western civilizations. This shift in the Internet user community will lead to greater internationalization of the Internet and will facilitate better representation of various languages and cultures. Although there will be a fundamental change in the human user demographics, another significant development is that there will be more machines connected to the Internet than human users within the next decade. Examples of machines that will be connected to the Internet include home appliances, automobiles, and furniture. The Internet is forcing merger of computer communications, telecommunications , broadcasting, and publishing, with the Internet as the primary infrastructure. Through this merger process, the Internet will become ubiquitous like electricity. The telecommunications and broadcasting industries will continue to experience radical changes caused by the Internet revolution. The telecommunications industry, with the telephone as its primary application, contributed to widening the economic gap between the rich and the poor. The rich, developed countries benefited through deployment of the telecommunications enormously, but the emerging countries did not benefit equally. The Internet may follow a pattern similar to the telecommunications revolution of the 19th and 20th centuries. The last decade has seen the developed countries benefit from the Internet far more than the emerging countries. Countries such as China and India can narrow this Internet digital divide by effectively exploiting …

Internet inroads

T he Internet continues to grow at a rapid pace. The growth is global, encompassing traffic volume, population, as well as the variety of applications being used. In 1995, the Internet reached 148 out of 185 United Nations member countries (86%), compared with 73 out of 159 United Nations member countries (46%) in 1991. The backbone speeds are currently 45Mbps to 622Mbps in advanced countries, moving toward Gbps rates and beyond from the past speeds of 64Kbps and 1.5Mbps. The number of users and computers is also increasing rapidly, with an annual growth rate of 50% to 100%. User applications cover an enormous variety of areas—technical, commercial, cultural, and educational.

Scalable and topologically-aware application-layer multicast

We present a scalable and topologically-aware application-layer multicast approach, specially designed for large-scale distributed applications. The proposed approach constructs topologically-aware data paths which are based on topological clustering of multicast group members. The approach does not require any exact network topology information, but instead requires the relative location information of members using landmarks. We partition the members into topologically-aware clusters based on the ordering of their close landmarks. We hierarchically arrange the clusters and separate data paths into two types (i.e., inside-cluster path and outside-cluster path) to exclude outsider nodes, not belonging to the same cluster, from the inside-cluster paths. Our results on performance evaluation show that constructing topologically-aware data paths can reduce unnecessary high latency and redundant network resource usage with low overhead over existing scalable approaches.

Dynamic rate control mechanism for large scale sessions

The network environments of the Internet Multicast session performance are various. The size of Internet community gets bigger and bigger and the variety gets deepening. This paper suggests to build a barrier on every link to restrict the amount of data flowing into the network. Each link has threshold value to specify the requirement for the packets. Multimedia stream is encoded into multiple layers and each layer has its own TTL value. The lowest layer has the largest TTL and the highest layer has the smallest TTL. The threshold is modified according to the traffic of the network and that of the link. It helps the network and the links avoid congestion and make full use of available network capacity.

On Optimal Regulation Policies for Certain Multi-Reservoir Systems

Optimal regulation policies for two reservoirs connected in series, with seasonal and possibly cross-correlated inflows, are analytically characterized when the immediate reward function is a piecewise linear concave function of the release and the objective is to maximize the expected discounted rewards over finite and infinite time horizons. The optimal policies are expressed in terms of certain critical numbers for which upper and lower bounds are obtained. We propose a heuristic approach for multi-reservoir systems under less restrictive assumptions and apply it to two reservoirs connected in parallel.

A Measurement Study of Storage Resource and Multimedia Contents on a High-Performance Research and Education Network

With the rapid advent and proliferation of peer-to-peer applications, research and education communities who have high performance networks have started to search for the possibilities of being benefited from the peer-to-peer architecture, in designing and supporting supercomputing applications as well as data-intensive applications. However, surprisingly, although they are mainly concerned with ways to take advantage of edge resources on their high performance network infrastructure, there has not been much work to measure and characterize the amount and usage pattern of the resources. In this paper, we remedy this situation by presenting a measurement study performed over a set of machines in a high performance research and education network, focusing on storage resources, multimedia contents, and availability of machines. We believe that this report will be a good reference to peer-to-peer and/or Grid system designers who have had troubles in getting an estimate for the amount of storage resource and multimedia contents on a high performance network, prior to planning and offering a distributed data service or storage service to re-search and education communities.

Overview of a Parallel Object-Oriented Language CLIX

CLIX is a parallel object oriented language that embodies communicating process model of object oriented programming. It incorporates the notion of objects with communications in distributed systems. The objects in CLIX are encapsulation of information and communication handlers that operate on the information in response to messages. The underlying communication system is modeled as a mail system. In addition to presenting an overview of CLIX, it is reviewed in terms of relevant issues in parallel object oriented programming.