Search

Omega space division architectures for photonic switching are investigated, with focus on optical crosstalk reduction and conflict avoidance. In such networks, many admissible permutations are not crosstalk free, due to the fact that different conflict-free paths are forced to share the same switching elements. To reduce those conflicts, a method operating in a multi-layer structure is proposed, that is based on a bipartite and two-colorable graph algorithm. [ABSTRACT FROM AUTHOR]

This paper considers planning and optimization of WDM networks by means of Integer Linear Programming (ILP), which is the most used exact method to perform this task. Since survivability is a very crucial issue in order to guarantee QoS in optical WDM networks, new ILP formulations are investigated here to design WDM networks under protection strategy with shared transmission resources. These formulations are applied to multifiber mesh networks with or without wavelength conversion. After presenting the formulations we discuss the results we obtained by exploiting them in the optimization of a case-study network. [ABSTRACT FROM AUTHOR]

This paper proposes a solution to support real-time multicast trafic with Quality of Service (QoS) constraints over Differentiated Services (DiffServ) IP networks. Our solution allows multicast users to dynamically join and leave the multicast tree. Moreover, it allows a multicast user which has negotiated a best-effort session to upgrade to a QoS-enabled session. Our solution is backward compatible with the Protocol Independent Multicast (PIM) scheme. It combines two ideas. First, resource availability along a new QoS path is verified via a probe-based approach. Second, QoS is maintained by marking replicated packets with a special DSCP value, before forwarding them on the QoS path. [ABSTRACT FROM AUTHOR]

The Differentiated Services architecture is a simple and scalable approach to provide Quality of Service (QoS) in IP Networks. Several studies have shown that the number of microflows in aggregates, the round trip time (RTT) or the mean packet size are key factors in the throughput of aggregates obtained using this architecture. In this paper, we examine the behaviour of one of the techniques suggested to improve fairness in a Diffserv network: the Aggregate Flow Control mechanism. We also propose two alternatives in the control overlay of this scheme and compare them with the original approach. Simulation results indicate that our proposed modifications improve throughput assurance and fairness requirements. [ABSTRACT FROM AUTHOR]

The All-IP network concept with end-to-end QoS provisioning has received particular attention in 3GPP recently. The UMTS proposals, however, have not yet solved some protocol interoperability issues. This paper analyzes the IP Multimedia Subsystem from the aspect of call control, resource reserva- tion and network policing interoperability from the viewpoint of implementa- tions. More specifically, the experiences based on a prototype implementation of the IMS based on SIP, RSVP and COPS are analyzed and conclusions are drawn to support the standardization process, as well as future implementations. The considered architecture is general and can be applied also to fixed IP net- works. [ABSTRACT FROM AUTHOR]

The main elements of a communications system that is able to provide QoS over the Internet are a scheduling policy and a signaling protocol. This paper is concerned with the design of the reservation protocol. The most widely used reservation protocol is RSVP which we take as a basis of this work. RSVP has been criticized mainly because of its complexity and poor scalability. This paper presents the first steps towards the definition of a new version of RSVP, which we call RSVPv2. The goal of RSVPv2 is to provide a more "light" approach, that can help improve handling reservations in the network by means of a simplified behaviour. [ABSTRACT FROM AUTHOR]

We consider the case of a virtual data network comprised of a set of end-to-end virtual leased lines. Our work aims at tackling the aspect of appropriate dynamic dimensioning of end-to-end paths subject to minimizing the consumed resources under the constraint of providing deterministic QoS guarantees. In particular, we propose a novel application of fuzzy logic control in order to dynamically adjust the resources assigned to one path to variations of traffic traversing that path. We design the required fuzzy sets and the respective rule base of the controller and evaluate its asymptotic stability and performance. As controller input a measure of the traffic variation shall be averaged over a variable size window. We design a second fuzzy logic controller in order to adapt this averaging window, such that significant traffic variations are captured while the measurement overhead during periods of nearly time invariant traffic characteristics is reduced. [ABSTRACT FROM AUTHOR]

Resource reservation is a essential component in providing QoS guarantees to distributed multimedia applications that run over the internetwork. Early reservation systems, including both immediate reservation and advance reservation, have taken an "all-or-nothing" approach in which both QoS and temporal parameters of requests are in- flexible. This paper describes a new probe-based adaptive reservation approach using Probing Requests that exploits the potential flexibility in reservation requests to increase resource utilization and reduce rejection rates. Compared to other approaches that support flexibility, the probebased approach is more effcient in finding alternatives, and it causes less signaling overhead and incurs a lower computational load on the often busy resource providers. The Probing Request mechanism has been implemented as part of our inter-domain bandwidth broker, can also be applied to other resource reservation schemes. [ABSTRACT FROM AUTHOR]

In this paper, we exploit a prioritized-service architecture and apply a priority based traffic control scheme to reduce abusive Internet access, improve fairness among users and study users. behavior under a prioritized service for Internet access. The Internet access by dormitory users of National Taiwan University (NTU) serves as a conveyer problem. There are two classes of service. The regular class has a volume quota for each user, which is designed to meet majority users'. essential demands while limiting abusive usage of quality service. The custody class is lower in service priority with no volume quota limitation. Our mathematical models for design and analysis include individual and aggregate user demand models and a network performance model. The priority control scheme is implemented over an existing NTU dormitory network with additions of a QoS router, a meter reading server, an accounting server and a user interface server. The control leads to 48.9% reduction in average packet drop rate, 42.2% improvement in a fairness measure, reduction of abusive Internet access by 57.82% and 145.09% Internet access increase in majority users. Such results are quite consistent with predictions of our mathematical models. [ABSTRACT FROM AUTHOR]

Today TCP/IP congestion control implements the additive increase/multiplicative decrease (AIMD) paradigm to probe network capacity and obtain a "rough" but robust measurement of the best effort available bandwidth. Westwood TCP proposes an additive increase/adaptive decrease paradigm that adaptively sets the transmission rate at the end of the probing phase to match the bandwidth used at the time of congestion, which is the definition of best-effort available bandwidth in a connectionless packet network. This paper addresses the challenging issue of estimating the best-effort bandwidth available for a TCP/IP connection by properly counting and filtering the flow of acknowledgments packets using discrete-time filters. We show that in order to implement a low-pass filter in packet networks it is necessary to implement an anti ACK compression algorithm, which plays the role of a classic anti-aliasing filter. Moreover, a comparison of time-invariant and time- varying discrete filters to be used after the anti-aliasing algorithm is developed. [ABSTRACT FROM AUTHOR]

We propose a solution for implementing a conceptual framework for IP centric communication in the optical domain. The solution, termed Lighttrails, is a combination of node architecture and protocol for realizing efficient optical communications from IP bursts to dynamic lightpaths. It is a paradigm shift from conventional optical communication modes, in supporting amongst others, very fast optical connection set up and tear down for burst of lightpaths communication, dynamic and highly bandwidth efficient sub-lambda provisioning. Light-trails also provide a first solution to optical multicasting, a key element for many of the emerging services that motivate the need for optical capacity. Contrary to existing proposals for IP type communication in the optical domain light-trail node architecture also presents the first practically implementable solution to enable optical transport with mature technology, non stringent optical switching requirements, and presenting a much more cost effective alternative to electronics. [ABSTRACT FROM AUTHOR]

In this work, we describe the implementation of an architecture to perform admission control and traffic management for multicast sessions in Diff-Serv networks. The Bandwidth Broker functionalities are carried out by the Call Admission Multicast Protocol (CAMP). CAMP performs the set-up of a multicast RTP session. It supports dynamic changes in the group membership. The implementation has been performed in a testbed network based on the Linux platform; we discuss the measurement results obtained by performing experiments with the testbed. [ABSTRACT FROM AUTHOR]

This paper describes the design, implementation and testing of a test bed supporting flow-based classification functions for multiservice traffic. Protocol and statistical analysis of application flows is performed in the edge routers to provide EF treatment to multimedia traffic without any user signaling. These functions take advantage of the Linux Traffic Control environment and implement SLA management and traffic statistics collection. Sample measurement performed on the test bed shows the effectiveness and feasibility of the proposed solution. [ABSTRACT FROM AUTHOR]

Finding a path in the network for a new incoming connection able to guarantee some quality parameters such as bandwidth and delay is the task of QoS routing techniques developed for new IP networks based on label forwarding. In this paper we focus on the routing of bandwidth guaranteed flows in a dynamic scenario where new connection requests arrive at the network edge nodes. When more than a path satisfying the bandwidth demand exists, the selection of the path is done in order to minimize the blocking probability of future requests. We propose a new routing algorithm named Virtual Flow Deviation (VFD) which exploits the information of the ingress and egress nodes of the network and the traffic statistics. We show that this new algorithm allows to reduce remarkably the blocking probability in most scenarios with respect to previously proposed schemes. [ABSTRACT FROM AUTHOR]

Voice, video and multimedia sessions are applications sensitive to the QoS provided by the underlying IP network. Therefore a lot of interest is currently devoted to the interaction of application level protocols with the QoS mechanism in IP networks. Among them SIP is currently having a lot of attention as a protocol for session signaling over the Internet. This work will describe an enhancement to SIP protocol for the interworking with a QoS enabled IP network. The proposed mechanism is simple and it fully preserves backward compatibility and interoperability with current SIP applications. Moreover the paper describes the application of this mechanism to a particular QoS enabled IP network, which implements DiffServ as transport mechanisms (the DiffServ mechanisms are obtained by means of Traffic Control functionalities with the TCAPI software libraries) and modified COPS clients for resource admission control. A test-bed implementation on Linux PCs of the proposed solutions is finally described. [ABSTRACT FROM AUTHOR]

A QoS providing distributed resource management scheme for multimedia Ad Hoc Wireless LANs(AWLANs) based on orthogonal frequency division multiplexing-code division multiple access(OFDMCDMA) is presented. This scheme implements distributed resource management with granular OFDM-CDMA channel architecture to support multimedia services with QoS provisions. The performance evaluation result for broadband wireless access(BWA)-type physical layer supporting MPEG traffic sources shows that when the number of nodes is 50 and the mean session arrival rate is less than 0.1, we can get the blocking rate of 10-2 and the QoS loss probability of 10-1 when there is no control packet loss. [ABSTRACT FROM AUTHOR]

The Internet, which has traditionally supported throughputsensitive applications such as email and file transfer, is increasingly supporting delay-sensitive multimedia applications such as interactive audio. These delay-sensitive applications would often rather sacrifice some throughput for lower delay. Unfortunately, the current Internet does not offer choices in the amount of delay or throughput an application receives, but instead provides monolithic best-effort service to all applications. This paper proposes and evaluates a new Active Queue Management (AQM) technique that employs source hints to provide service at network routers that is sensitive to the Quality of Service (QoS) expectations for a variety of applications. Applications indicate their delay or throughput sensitivity via a delay hint in their outgoing packets. The router, which we call RED-Boston, uses the delay hints to dynamically adjust the router to yield better delay performance for delay-sensitive applications and better throughput for throughput-sensitive applications. Using a new QoS metric, our simulations demonstrate that RED-Boston yields higher QoS than an adaptive version of RED for both throughputsensitive flows and delay-sensitive flows. RED-Boston operates equally well in all trafic scenarios and fits the current best-effort Internet environment without requiring trafic monitoring. [ABSTRACT FROM AUTHOR]

When a considerable amount of UDP traffic is injected into the Internet by distributed multimedia applications, the Internet easily becomes congested. Consequently, the bandwidth available to TCP connections becomes limited and performance significantly deteriorates. In order that both multimedia applications and TCP-based ones can fairly co-exist on the Internet, it is becoming increasingly important to consider inter-protocol fairness. In this paper, we first consider several issues related to TCP-friendly video transfer and propose several video-quality adjustment methods which accomplish highquality, stable, and TCP-friendly FGS (Fine Granular Scalability) video transfer. Then, with consideration on the quality degradation caused by packet loss in video quality adaptation, we extend our method so that it is applicable to a lossy environment when it is used in cooperation with the FEC (Forward Error Correction) technique. Our mechanism adjusts the video quality in accordance with the TFRC (TCP-Friendly Rate Control) rate, the packet loss probability, and the resultant video quality, while avoiding undesirable video-quality degradation. Through simulation experiments using FGS video streams, we show that our proposed method can provide high-quality, stable and TCP-friendly video transfer even in the unstable and lossy Internet. [ABSTRACT FROM AUTHOR]

The increasing popularity of streaming video is a cause for concern for the stability of the Internet because most streaming video content is currently delivered via UDP, without any end-to-end congestion control. Since the Internet relies on end systems implementing transmit rate regulation, there has recently been significant interest in congestion control mechanisms that are both fair to TCP and effective in delivering real-time streams. In this paper we design and implement a protocol that attempts to maximize the quality of real-time MPEG-4 video streams while simultaneously providing basic end-to-end congestion control. While several adaptive protocols have been proposed in the literature 20,27, the unique feature of our protocol, the Video Transport Protocol (VTP), is the use of receiver side bandwidth estimation. We deploy our protocol in a real network testbed and extensively study its behavior under varying link speeds and background trafic profiles using the FreeBSD Dummynet link emulator 23. Our results show that VTP delivers consistent quality video in moderately congested networks and fairly shares bandwidth with TCP in all but a few extreme cases. We also describe some of the challenges in implementing an adaptive video streaming protocol. [ABSTRACT FROM AUTHOR]

In the next future, several wireless technologies will coexist and users will require to gain wireless access at any time through the most convenient technology using the same terminal. As a result, terminals will make handover between different access solutions, which we call inter-Reference Environment (RE) handovers. This involves the concept of reconfigurability: terminals must reconfigure their internal parameters to adapt their behavior to the new RE. However, in this scenarios there are more severe problems for TCP-based data services where compared to homogeneous wireless systems. In fact, the low Mobile IP performance causes long periods of terminal unreachability. Besides, when the reference environment changes the end-to-end path changes as well and what TCP learned in terms of round trip time and available bandwidth is not valid anymore. In this paper solutions to improve TCP performance in reconfigurable networks are introduced. The proposed solutions are based on appropriate setting of the maximum segment size (MSS) and modifications to TCP algorithm which are triggered on the following of an inter-RE handover. Performance results show that the proposed algorithms dramatically increase TCP performance. [ABSTRACT FROM AUTHOR]

Reconfigurability is expected to play a critical role in the research area of Wireless/Mobile Communications, by increasing flexibility, reducing deployment as well as operation and maintenance costs, creating new business opportunities, facilitating enhancements to existing technology, service personalization, etc. On the long way leading to a complete Software Defined Radio terminal, in which a unique programmable hardware is on the fly configured via software download, the RAMON project places itself within a short/medium term time horizon. In this vision, a dual-mode terminal, enhanced with a common software module, is considered, with a massive use of Common Control Functions, limiting the use of system specific functionalities. The focus of the paper is to describe the mobility management developed in this project, providing details for both inter and intra reference environments mobility. [ABSTRACT FROM AUTHOR]

Adaptive packet scheduling for wireless data access systems, including channel state information based algorithms, is a quite well established topic. We develop here an original framework to define such an algorithm in a reconfigurable context, i.e. with the ability to exploit heterogeneous communication environments. The focus is to define platform independent algorithms that can be "adapted", by a specific software, to different environments, so that the core communication functions can be defined, modified and improved once for all. The specific communication function addressed in this work is packet scheduling. For better performance the packet scheduling design exploits the cross-layering approach, i.e. information and functions from different communication layers are used. The concept is proved with reference to the UMTS and Bluetooth technologies, as representatives of a cellular system and a local access one respectively. [ABSTRACT FROM AUTHOR]

A design study of a Re-configurable Access Module for Mobile Computing Applications is described. After a presentation of its cross-layered architecture, Control Parameters (CPs) of the module are introduced. The set of CPs both describes the functional state of the communication process in relation to the time-varying transport facilities and provides, as input of suitable Algorithms, the control information to re-configure the whole protocol stack for facing modified working conditions. The paper also presents the structure of the simulator realized to demonstrate the feasibility of the design guidelines and to evaluate reconfigurability performances. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of quality of service management in optical packet switching over DWDM. Asynchronous, variable length packets are considered and algorithms able to offer congestion resolution and quality of service differentiation are presented, with reference to both connectionless and connection oriented network scenarios. The paper aims at showing that it is possible to guarantee differentiation of the quality of service among trafic classes, with very little buffering requirements. [ABSTRACT FROM AUTHOR]

In the present Internet, due to the popularity of AQM techniques, great effort is being devoting to analyzing their performance and optimizing their parameters. In this perspective the target of the paper is to provide an analytical framework to evaluate the performance of an AQM router loaded by any kind of traffic which can be modeled with a Markov arrival process. In order for complexity of the model to be independent of the queue size, the model uses a fluid-flow approach. In the paper, just to provide an example of application, we will consider RED routers. After the model definition, with the aim of demonstrating the correctness of the proposed methodology, the case of a RED buffer loaded by a constant bit-rate (CBR) source has been studied. In this particular case, where the classical theory of probability can be applied, the solution of the problem is derived, verifying that obtained analytical equations coincide with those obtained in the general case. [ABSTRACT FROM AUTHOR]

Wavelength routed optical networks allow to design a logical topology, comprising lightpaths and routers, which is overlayed on the physical topology, comprising optical fibers and optical cross-connects, by solving a Routing and Wavelength Assignment (RWA) problem. In this paper we extend the concept of lightpath to the one of Super-LightTree, which uses a simple bit level Time Division Multiplexing that can be directly implemented in the optical domain, to split the wavelength bandwidth over a tree among more than one traffic flow. This allows to design logical topologies with an increased number of logical links, reducing the average distance among nodes, i.e., the number of electro-optic and opto-electronic conversions, and the traffic congestion on logical links. This also reduces the number of wavelengths required to solve the RWA problem. Being the Super-LightTree RWA problem computationally intractable, we propose two heuristics for which we show that the number of wavelengths required to overlay a given logical topology on a given physical topology is reduced by more that 70% using Super-LightTrees. [ABSTRACT FROM AUTHOR]

Protocol design and modification was traditionally based on heuristic processes and decisions. In particular several new versions of TCP have been proposed which not always stand up to their expectations when they are deployed or simply simulated in a real scenario. The closed-loop, nonlinear characteristics of TCP makes it very difficult to predict its performance in even simple scenarios. This papers analyses the process of modeling TCP-SACK with an Open Multiclass Queueing Network (OMQN) model, showing that sometimes simple implementation choices can have a non-marginal impact on the performance and discussing how an analytical model can be used to design and study TCP (or other protocols) modifications and improvements. As an example we present a modified, more robust implementation of TCP-SACK, named R-SACK, that significantly reduces the protocol timeout probability in presence of bursty losses. [ABSTRACT FROM AUTHOR]

Live streaming media is delay sensitive with small allowable delay. Current conventional multicast protocols do not provide a loss retransmission mechanism. Even there are some researches on a reliable multicast providing a loss retransmission mechanism but a long delay and a high packet loss rate make them inefficient for live streaming. This paper proposes a new multicast protocol based on a protocol relay concept. The proposed protocol focusing on allowable delay provides Quality of Service (QoS) for live streaming. Relay nodes are placed along the multicast tree. Data recovery is done between relay nodes. We propose the methods that enable protocol relays to request retransmission immediately and reduce the number of duplicate packets. Finally, we make a mathematical analysis of the proposed protocol and compare it with other multicast protocols. The results show that the proposed protocol is more efficient for live streaming than the conventional protocols. [ABSTRACT FROM AUTHOR]

We present an analytical model, based on a Fixed Point Approximation (FPA) solution, that can be used to derive the performance of different sets of TCP connections that share, and compete for, a common resource, typically a link and its associated buffer. A set of TCP connections is a group of connections that can be considered homogeneous, e.g., they have similar RTTs and all have a wireless access. TCP connections are modeled through the OMQN (Open Multiclass Queueing Network) paradigm. The conditions that define the feasibility of the solution and allow the convergence of the model are discussed and an application example with a RED buffer where wired and wireless connections converge is presented. [ABSTRACT FROM AUTHOR]

This paper proposes the Stochastic Petri Nets (SPN) formalism to model IP networks loaded with traffic resulting from a set of ON-OFF finite TCP-Reno connections. The approach is based on separate model descriptions of the TCP connection latency to transfer a finite number of packets and the IP network links; the two models are iteratively solved using a fixed point method. The overall model parameters are the primitive network characteristics; the model solution yields an estimation of the packet loss probability, the average completion time to transfer a finite number of packets over a TCP connection as well as the distribution of the completion time. The validation of the proposed approach and the future extensions to this work will also be discussed. [ABSTRACT FROM AUTHOR]

Layered multicast has been shown as a promising technique for dis-tributing a video program to a potentially large number of heterogeneous re-ceivers. While several layered multicast approaches have been proposed, prior work has almost exclusively focused on improving the perceived video quality in terms of PSNR value or/and the fairness among different receivers. In this paper, we first argue that stability is another major index for the video playback quality; in an extreme case even with the best possible achievable PSNR value, the frequent switching between different layers still can lead to significant qual-ity degradation. We then introduce an end-to-end adaptation framework that improves the stability and fairness for receivers by employing a dynamic layer rate allocation scheme on the sender's side. Specifically, we propose a new metric, called Stability-aware Fairness Index (SFI), which can capture both the stability as well as fairness. We then formulate the optimal layer rate allocation problem with the objective of minimizing the expected SFI for all the receivers in a multicast session, and derive an efficient algorithm to solve the problem. Simulation results demonstrate that our scheme can significantly improve the degree of stability and fairness, thus leading to better video playback quality. [ABSTRACT FROM AUTHOR]

Traditional multicast routing protocols use RPF (Reverse Path Forwarding) concept to build multicast trees. This concept is based upon the idea that an actual delivery path to a node is the reverse of the path from this node to the source. This concept fits well in symmetric environments, but in a routing environment where Quality of Service is considered the guarantee that a symmetrical path will exist between two network addresses is broken. Available network resources impose specific Quality of Service asymmetries, therefore reverse path routing may not be used. In this paper a new multicast routing strategy is proposed, enabling directed trees establishment, instead of reverse path ones. This new strategy, DTMP- Directed Trees Multicast Routing, is then implemented and simulated using Network Simulator. Simulation results, driven from several scenarios are presented, analyzed and compared with PIM-SM. [ABSTRACT FROM AUTHOR]

In this paper we propose a new scheme named DQM (Differentiated QoS Multicast) based on the Source Specific Multicast (SSM) 7 model in order to provide limited and qualitative QoS channels to support heterogeneous end users. Similar to the DiffServ paradigm, in DQM the network is configured to provide finite QoS service levels to both content provider and receivers. Based on the Service Level Agreement, both the content provider and group members should select a specific QoS channel available from the network for data transmission, and in this case arbitrarily quantitative QoS states are eliminated. Moreover, we use the group address G contained in the (S, G) tuple in SSM service model to encode QoS channels, and data packets that belong to the same QoS channel identified by a common class D address can be treated aggregately, and this can be regarded as an overlay solution to Differentiated Services, specifically for source specific multicast applications. [ABSTRACT FROM AUTHOR]

The core aspect of most multimedia applications is the transmission of Variable Bit Rate (VBR) video streams requiring a sustained relatively high bandwidth with stringent Quality of Service (QoS) guarantee. In such systems, a statistical bandwidth estimation is especially relevant as it is indispensable for implementing an efficient admission control. In this paper a novel measurement-based admission control algorithm for video distribution systems has been developed. Such an algorithm estimates the bandwidth occupied by the actual video streams, observing their temporal evolution in a chosen measurement window. In particular, for each video stream, the estimation procedure needs an online measurement of the bandwidth values assumed successively and their persistence time. An exponential weighted moving average filter is used for smoothing the estimated values. Finally, some measurements of video traffic are shown comparing the results with simulation and other analytical results, discussing the effectiveness of the considered solution. [ABSTRACT FROM AUTHOR]

Pricing for the use of telecommunication services has received in the last years a growing attention. The main objective is to establish various fairness criteria in the bandwidth allocation among different traffic classes. In the context of Quality of Service (QoS) networks (e.g. ATM, DiffServ, IntServ) the pricing scheme influences the Call Admission Control (CAC) rules. While, for Best Effort (BE) services, users accept a variable bandwidth allocation and are not subject to CAC and the pricing policies, according to the Proportional Fairnes Pricing, are integrated within the flow control. In this paper we consider both BE traffic and traffic explicitly requiring QoS (Guaranteed Performance, GP) and we propose three Call Admission Control rules for the GP traffic. The aim is to maximize the Internet Service Provider.s overall revenue and to establish a bound over the GP traffic prices. Numerical results are presented to show the good performance of the proposed techniques. [ABSTRACT FROM AUTHOR]

In this paper we propose an enhancement to the TCP protocol, called TCP Smart Framing, or TCP-SF for short, that enables the Fast Retransmit/Recovery algorithm even when the congestion window is small. TCP-SF is particularly effective for short-lived flows, as most of the current Internet traffic is. Without modifying the TCP congestion control based on the additive-increase/ multiplicative-decrease paradigm, TCP-SF adopts a novel segmentation algorithm: while Classic TCP starts sending one segment, a TCP-SF source is allowed to send an initial window of 4 smaller segments, whose aggregate payload is equal to the connection's MSS. This key idea can be implemented on top of any TCP flavor, from Tahoe to SACK, and requires modifications to the server behavior only. Analytical results, simulation results, as well as testbed implementation measurements show that TCP-SF sources outperforms Classic TCP in terms of completion time. [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel bandwidth estimation algorithm for TCP connections and its possible application to congestion control mechanism. The estimation algorithm relies upon an analytic relation which expresses the connections' available bandwidth as a function of the inter-departure time of packets and of the inter-arrival time of ACKs. It is worth noticing that this approach can be extended to protocols other than TCP, as long as they support an acknowledgment mechanism. The bandwidth estimation performance is assessed through discrete event simulations under various network topologies, traffic scenarios and link error conditions. Bandwidth estimation is then applied to TCP congestion control to select the value of the congestion window after a packet loss episode. Performance of this modified version of TCP is validated by means of simulations and compared to the one achieved by TCP NewReno. Finally, the possible coexistence of the modified version of TCP and TCP NewReno is proved through a detailed analysis of fairness and friendliness of the new protocol. [ABSTRACT FROM AUTHOR]

Admission Control and routing are key aspects in next generation networks supporting advanced Quality of Service and Traffic Engineering functionalities. However, traditional Traffic Engineering solutions usually rely exclusively on routing algorithms to achieve optimization in network resource usage, while Admission Control is simply limited to a local check of resource availability, without any optimization purpose. This paper proposes a novel approach for Admission Control in IP/MPLS networks, which applies at network edges by means of dynamic thresholds evaluated on the basis of network status. The proposed solution allows the achievement of more efficient usage of network resources, especially at medium/high load, with an increased robustness of the network and an overall performance improvement. [ABSTRACT FROM AUTHOR]

The Multi-Protocol Lambda Switching (MPλS) has been recently applied in the optical network control plane to provide fast lightpath provisioning. With the growth of traffic in optical network, the survivability of the network becomes extremely important. Therefore, an efficient protection mechanism is needed. Shared backup tree lightpath protection is a promising paradigm in MPλS networks because of its fast recovery from single failure and its efficiency in terms of consumed capacity. In this paradigm, a shared backup tree is used to protect a group of working lightpaths. In this paper, we propose a heuristic that yields an efficient placement of working paths and backup trees in polynomial time. [ABSTRACT FROM AUTHOR]

MPLS can be used to provide network robustness to faults through path protection techniques. In this paper we present a dynamic model supporting different classes of end-to-end protection, including protection against Single Fault and Dual Fault, with and without sharing of backup bandwidth. Beyond link and node failures we also consider protection against Shared Risk Link Group (SLRG) failure. An interesting feature of the proposed scheme is the ability to offer service differentiation with respect to the recovery probability, by coupling the differentiation on the number of backup paths with bandwidth assignment policy. In this paper we describe the underlying algorithms for route selection and backup bandwidth sharing. The route selection is based on explicit load-dependent routing of service and backup paths. We show by simulation that the proposed route selection algorithm is effective in improving the network utilization. We discuss two alternative implementations of our model: distributed and partially centralized. The primary concern with the distributed approach is the message overhead implied by link-load dissemination, e.g. by flooding. However we show by simulation that message overhead can be taken under control by adopting a well-tuned adaptive overhead reduction algorithm. Our conclusion is that both distributed and partially-centralized implementation are feasible. [ABSTRACT FROM AUTHOR]