QUIC is a secure, general-purpose transport protocol using congestion and retransmission controls on User Datagram Protocol (UDP). QUIC can use the same congestion control algorithms as Transmission Control Protocol (TCP). Thus, when communication between multiple terminals using QUIC competes, the performance of each terminal depends on the congestion control algorithm adopted by the terminal. In related works, the communication performance was evaluated by simulation when both CUBIC and BBR flows competed in TCP, and the results showed that each communication performance became unfair with the buffer size of the shared bottleneck link. In this study, we verify the communication performance at different bottleneck link buffer sizes when BBR and CUBIC compete through actual experiments.

QUIC is a transport protocol that adds congestion control, retransmission control, and TLS to UDP. QUIC can use the same congestion control algorithms as TCP. In previous work, when both CUBIC and BBR flows compete within QUIC, we have shown that communication performance becomes unfair with respect to the buffer size of the shared bottleneck link. In this study, we improve the communication performance at different bottleneck link buffer sizes using Round Trip Time (RTT) and Explicit Congestion Notification (ECN) when CUBIC and BBR compete within QUIC through actual experiments.

This paper analyses the problem of efficiently routing XML documents on a network whose nodes are capable of distributed XML processing. The goal of our study is to find network paths for which XML documents’ transmission will result in high likelihood that a large portion of the documents be processed within the network, decreasing the amount of XML processing at documents arrival at the destination site. We propose several routing algorithms for single route and multipath routing and evaluate them on a distributed XML network simulation environment. We show the benefits of the proposed XML routing algorithms as compared with widespread minimum hop routing strategy of the Internet.

Abstract In this paper, we introduce a Mean Value Analysis (MVA)‐based methodology for the performance evaluation of transactions executed in a multi‐server distributed system. We first present probabilistic arguments to investigate under which conditions MVA models provide worst‐case response times, without large overshoots. Then, we show how to characterize the distributed server system so as to construct a model to predict response times as well as to estimate system capacity. Finally, we exemplify the methodology usage via transactions implemented in two distributed Linux systems. Copyright © 2010 John Wiley & Sons, Ltd.

In this paper, we introduce a mean value analysis based methodology for performance evaluation of transactions executed in a multi server distributed system. We first present probabilistic arguments to investigate under which conditions MVA models provide worst case response times, without large overshoots. Then, we show how to characterize the distributed server system so as to construct a model to predict response times as well as estimate system capacity. Finally, we exemplify the methodology usage via transactions implemented in two distributed Linux systems.

In this paper, we address the problem of how TCP sessions ramp up their transmission windows in a controlled way. We introduce a conservative slow start scheme that reduces to the regular slow start when session path queues are empty, but slows down the control window increase speed as buffers start to build up. We show via event driven simulation and open source based high speed experimental testbed the effectiveness of our conservative slow start mechanism in reducing packet losses and consequent retransmissions.

Basic message processing tasks, such as well-formedness checking and grammar validation, can be off-loaded from the service providers' own infrastructures. To enable effective off-loading of processing tasks, we introduce the Prefix Automata SyStem - PASS, a middleware architecture which distributively processes XML payloads of Web service SOAP messages during their routing towards Web servers. PASS is based on a network of automata, where PASS-nodes independently but cooperatively process parts of the SOAP message XML payload.

This article evaluates various networking technologies to satisfy A/V requirements in the home, and advocates Ethernet as the backbone network technology in the home, identifying areas of improvement so that Ethernet can accommodate consumer electronics applications in addition to computer devices.

The information service is a critical component of a grid infrastructure for resource discovery. Although P2P computing paradigm could address some of the scalability issues that plagued grid resource discovery, most existing distributed hash table (DHT) based P2P overlays have difficulty in treating attribute range queries that are common in resource discovery lookups due to the inherent randomness of hash functions. Recently, there have been various attempts to solve the range search problem over DHT networks [Aberer, et al. (2003), Gao and Steenkiste (2004), Ratnasamy, et al. (2003), and Tanin, et al. (2005)]. Central to all of these is a mapping scheme which maps the tree-structured logical index space to some DHT-based physical node space. In this paper, we propose a general framework to put all these under the same umbrella based on how mapping of the tree-structured index that identifies a physical node responsible of a particular range is done through replication. We identify three schemes which should cover the spectrum of all meaningful replication schemes: the tree replication scheme (TRS) replicates the entire tree; the path caching scheme (PCS) replicates paths from root to leaves; and the node replication scheme (NRS) replicates individual logical nodes.

A resource directory is a critical component of a grid architecture. P2P computing paradigm could address some of the scalability issues that make distributed resource discovery services challenging. Unfortunately, most existing distributed hash table (DHT) based P2P overlays have difficulty in treating attribute range queries that are common in resource discovery lookups. This paper proposes a general framework for range-based resource discovery. In particular, the proposed framework maps tree-structured logical data (i.e., range indexing) onto a DHT-based physical node space (i.e., resource brokers). In this paper, we consider three mapping schemes from the logical space onto the physical space. Each mapping scheme uses a different replication mechanism to reduce range search time and to achieve load balance. We analytically and experimentally compare the performance characteristics (query/update costs and workload distributions) of these schemes and discuss their applicability under different resource discovery service scenarios.

We study routing and wavelength assignment (RWA) tasks in WDM mesh networks with wavelength conversion capabilities. For each lightpath request, we solve the routing and wavelength assignment sub-problems as a single problem, rather than separating them into two separate problems. We use four different optimization criteria: (1) minimum number of wavelength conversions; (2) minimum number of wavelengths used; (3) minimum hop count; (4) minimum use of scarce resources (trunking). Each of these problems is solved by a Dijkstra type of algorithm, executed sequentially on one request at a time. We also develop an integer linear programming (ILP) model with the objective of minimizing the number of wavelength conversions. The optimal ILP solution can only be obtained for small networks. We conduct a comprehensive simulation study of the performance of the algorithms for different network topologies and traffic scenarios.

In this paper, we study the routing and wavelength assignment (RWA) tasks in WDM mesh networks with wavelength conversion capabilities. We solve the routing and wavelength assignment sub-problems as a single problem, rather than separating them into two separate problems. We use two different optimization criteria: (1) minimize the number of wavelength conversions used; and (2) minimize the number of wavelengths used. Each of these problems is solved by using a method based on the Dijkstra algorithm, sequentially on one connection at a time. We study and comment on the performance of the algorithms for different network topologies and traffic scenarios.

In this paper, we introduce a control theoretical analysis of the closed-loop congestion control problem in packet networks. The control theoretical approach is used in a proportional rate controller, where packets are admitted into the network in accordance with network buffer occupancy. A Smith Predictor is used to deal with large propagation delays, common to high speed backbone networks. The analytical approach leads to accurate predictions regarding both transients as well as steady-state behavior of buffers and input rates. Moreover, it exposes tradeoffs regarding buffer dimensioning, packet loss, and throughput.

Packet switches are required to support quality of service (QoS), that is, the capability to differentiate packet servicing on a flow basis, ensuring switching objectives such as guaranteed throughput and bounded switching delays. The core of a packet switch is a switch fabric, which is typically available as a single chip or chip set. Currently, these chip sets do not support QoS features, such as required packet service rates. We develop a framework based on performance indicators for evaluating a packet switch with respect to fairness performance. In conjunction with a robust simulation environment, we analyze the performance of two well known core packet schedulers.

The fast growth of the Internet has made popular applications with diverse and stringent transmission requirements. In order to meet this demand, packet switches today are designed to include the support of quality of service (QoS) as a powerful product differentiator. Customized packet handling on a flow by flow basis, ensuring switching objectives such as guaranteed throughput and bounded switching delays, is a goal likely to affect the entire switch architecture. The core of a packet switch, the switch fabric, is typically available as a single chip or chip set. These chip sets currently do not support QoS features, such as minimum packet service rates. This paper addresses an important feature of an input queue packet switch in the support of QoS: minimum service rates. We develop a framework for evaluating a switch fabric to support minimum service rate guarantees, and apply it to two well know switch fabric schedulers. In conjunction with a robust simulation environment, we verify that specific algorithms to ensure minimum service rates are needed. We develop several algorithms for this purpose and evaluate tradeoffs between service rate guarantees and throughput.

In this paper, we study the routing and wavelength assignment (RWA) tasks in WDM rings with heterogeneous wavelength conversion capabilities. By heterogeneous, we mean that different nodes of the WDM ring might have different wavelength conversion capabilities. An integer linear programming (ILP) formulation is used to optimally solve small examples of the problem. Because many RWA problems have exponential complexity, efficient heuristics are needed. We use the results from the ILP to evaluate the performance of several fast RWA algorithms, which are shown to give accurate solutions under various traffic scenarios.

To protect against possible network node or link failure and achieve high reliability of communications, pre-planned failure recovery schemes are needed in modern high-speed communication networks. A couple of schemes have been previously reported for multicast communications. We present a scheme based on a "dual-tree" structure in which a secondary tree for fault-tolerance purpose is built as a complement to a primary multicast tree. The secondary tree provides alternative delivery paths that can be activated when link or node failure is detected in the primary multicast tree. Simulation experiments show that this scheme has shorter restoration time and cause less multicast tree cost increase after restoration than some schemes proposed previously.

We propose the use of QoS routing to enhance the support of IP telephony. Our proposed scheme is based on QoS intradomain OSPF routing, an extension of the conventional OSPF routing protocol. A DiffServ model is used (no per flow signaling, nor per flow accounting at intermediate nodes). Processing O/H is shifted from core to edge routers, which compute routes, monitor QoS path quality and enforce call acceptance control (CAC) using the link state information advertised by OSPF. Via simulation, we show significant delay and throughput improvement over IP telephony strategies currently used in the Internet. In particular, hot spots and focussed congestion points are easily avoided. Moreover, the ability to control voice via QoS routing and CAC permits us to adjust the capacity sharing between voice traffic and TCP traffic, by reserving a fraction of link bandwidth to TCP data traffic. We also show that the added control and processing overhead is quite manageable, even in fairly large networks.

Presents a feedback control algorithm for ATM congestion control in which source rates are adjusted according to VC queue lengths at intermediate nodes along the path. The goal is to "fill in" the residual bandwidth, without exceeding a specified queue threshold. In order to obtain this, we propose a simple and classical proportional controller, plus a Smith predictor to overcome instabilities due to large propagation delays, as well as to avoid cell loss. We propose an effective EPRCA implementation in which each source computes its input rate based on the maximum VC queue length along the path. Theoretical and experimental results show that high throughput is achieved even with queue sizes independent of the round trip delay.