Congestion control protocols for background data are commonly conceived and designed to behave as low priority traffic, i.e., completely yield to delay sensitive flows such as web traffic. This behavior can cause starvation and hence the accumulation of large numbers of flows, leading to flow level instability. In this paper we look at the fundamental problem of designing congestion control protocols for background traffic with minimum impact on delaysensitive flows while achieving a certain desired average throughput over time. The corresponding optimal policy under various assumptions on the available information is obtained analytically. We give tight bounds for the negative impact of TCP-based background transfer protocols compared to the optimal policy, and identify the range of system parameters for which more sophisticated congestion control makes a noticeable difference. Based on these results, we propose an access control algorithm for systems where control on aggregates of background flows can be exercised, e.g., in file servers. Simulations of simple networks suggest that this type of access control performs better than protocols emulating low priority.
Trustworthiness of software and services is a key concern for their use and adoption by organizations and endusers. Trustworthiness evaluation is an important task to support making informed decisions for both providers and consumers, i.e., for selecting components from a software marketplace. An analysis of the state of the art in software evaluation technologies motivated us to develop an evidence-based approach for trustworthiness evaluation. Most of the literature evaluates trustworthiness by focusing on a single dimension (e.g., from the security perspective) while there are limited contributions towards multifaceted and end-to-end trustworthiness evaluation. Our analysis reveals that there is a lack of a comprehensive framework for comparative, multi-faceted end-to-end trustworthiness evaluation, which takes into account different layers of abstractions of both the system topology and its trustworthiness. In this paper, we provide a framework for endto-end trustworthiness evaluation using computational approaches, which is based on aggregating certified trustworthiness values for individual components. The resulted output supports in the definition of trustworthiness requirements for a software component to be locally developed and eventually integrated within a system, as well as, trustworthiness evidences for a composite system before the actual deployment. Thereby supports the designer in analyzing the end-to-end trustworthiness values. An application example illustrates the application of the framework.
Abstract. Many, if not most, well-designed Future Internet protocols fail, and some badly-designed protocols are very successful. This somewhat depressing statement illustrates starkly the critical importance of a protocol's deployability. We present a framework for considering deployment and adoption issues, and apply it to two protocols, Multipath TCP and Congestion Exposure, which we are developing in the Trilogy project. Careful consideration of such issues can increase the chances that a future Internet protocol is widely adopted.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.