Abstract IQ‐Services are application‐specific, resource‐aware code modules executed by data transport middleware. They constitute a ‘thin’ layer between application components and the underlying computational and communication resources. This layer implements the data manipulations necessary to permit wide‐area collaborations to proceed smoothly in the presence of dynamic resource variations. IQ‐Services interact with the application and resource layers via dynamic performance attributes, and end‐to‐end implementations of such attributes also permit clients to interact with data providers. The joint middleware/resource and provider/consumer interactions implemented with performance attributes may be used to realize effective methods for managing the data flows in the large‐data, distributed Grid applications targeted by our research. Experimental results in this paper demonstrate substantial performance improvements. These are attained by coordinating network‐level with service‐level adaptations of the data being transported and by permitting end users to dynamically deploy and use application‐specific services for manipulating data in ways suitable for their current needs. Copyright © 2005 John Wiley & Sons, Ltd.

Predicting the throughput of large TCP transfers is important for a broad class of applications. This paper focuses on the design, empirical evaluation, and analysis of TCP throughput predictors. We first classify TCP throughput prediction techniques into two categories: Formula-Based (FB) and History-Based (HB). Within each class, we develop representative prediction algorithms, which we then evaluate empirically over the RON testbed. FB prediction relies on mathematical models that express the TCP throughput as a function of the characteristics of the underlying network path. It does not rely on previous TCP transfers in the given path, and it can be performed with non-intrusive network measurements. We show, however, that the FB method is accurate only if the TCP transfer is window-limited to the point that it does not saturate the underlying path, and explain the main causes of the prediction errors. HB techniques predict the throughput of TCP flows from a time series of previous TCP throughput measurements on the same path, when such a history is available. We show that even simple HB predictors, such as Moving Average and Holt-Winters, using a history of few and sporadic samples, can be quite accurate. On the negative side, HB predictors are highly path-dependent. We explain the cause of such path dependencies based on two key factors: the load on the path and the degree of statistical multiplexing.

No abstract available.

Multicast routing research efforts have mostly focused on the host-group model in which multicast packets are addressed to a host group. Another multicast routing approach uses multidestination addressing, where a multicast packet carries a list of the unique (unicast) addresses of all the group members. This form of routing uses limited or no additional state beyond the existing unicast routing tables. It, therefore, scales well with the number of multicast sessions but does not scale well with the multicast group size and, in fact, requires the size of the group to be below a certain threshold. In this paper, we envision a future scenario in which both host-group and multidestination addressing routing approaches coexist within the Internet. We develop a dynamic routing context for this scenario wherein a multicast session can adapt among different routing configurations depending on the multicast group size and how this size changes over time. We consider three routing options: 1) a single multi-destination addressed flow - suitable for small-group sessions, 2) multiple multi-destination addressed flows - suitable for medium-group sessions and 3) a single host-group addressed flow - suitable for large-group sessions. Our work is concerned with the development and evaluation of protocols that allow a multicast session to dynamically switch among these three routing options as the size of the session changes.

Our research addresses the efficient transfer of large data across wide-area networks, focusing on applications like remote visualization and real-time collaboration. To attain high performance in the real-time exchange of data across collaborating machines and end users, we are developing and evaluating methods and techniques for coordinating application-level with network transport-level adaptations of data communication. Specifically, complementing previous work on TCP-friendly communication and on adaptive transport protocols, our approach is to strongly coordinate application-level with transport-level changes in communication behavior, so as to best meet application needs without violating fairness in network resource usage. The approach is evaluated with the IQ-ECho middleware, which implements the distribution of scientific data to remote collaborators. Using IQ-ECho, application-level adaptations like selective data down-sampling are triggered by transport-level information provided by the instrumented IQ-RUDP protocol underlying IQ-ECho's communications. The application- to network-layer exchange of information necessary for such coordinated adaptations is implemented with ECho attributes, which provide a lightweight way for an application to provide quality of service information and to describe its adaptation to the transport layer.