The optical layer can restore the disrupted traffic on a percall basis in the case of link/router failures. Thus, the optical layer can provide different levels of restoration for different classes of service.
The optical layer can restore the disrupted traffic on a percall basis in the case of link/router failures. Thus, the optical layer can provide different levels of restoration for different classes of service.
Summary form only given. Today's communication networks, such as the Internet, can be seen as superfast expressways, transporting packets from sources to destinations along most efficient routes, while respecting ...
详细信息
Summary form only given. Today's communication networks, such as the Internet, can be seen as superfast expressways, transporting packets from sources to destinations along most efficient routes, while respecting priorities and quality-of-service specifications. They are heterogeneous, in terms of both the traffic they can accommodate and the services they offer. The quest for regulating traffic in such a medium, with utmost efficiency, has been and still is the driving force behind the technological advances in this area. Some of this regulation is done by the network itself (specifically, the service providers), through fast switch and router designs, protocol designs, capacity expansions, and admission policies, and some of it is done by the individual users under incentives or disincentives provided by the network (such as price structures), aimed at achieving a certain pattern of usage of the available resources (such as bandwidth). The users are faced with a multitude of issues, among which are (i) at what (flow) rate to inject packets into the network, and (ii) how to adjust these rates in response to (delayed) congestion information received from the network. Occasionally, users are also faced with the task of making decisions on routing, again based on information received from the network. The overall system is a very complex one, involving a very large number of "players" with little if any cooperation, and operating under minimum information. Several branches of engineering (particularly, control, communications, and computing) as well as economics (particularly, game theory) provide tools (conceptual and technological) necessary for architecting such networks and ensuring that the operate harmoniously in spite of their heterogeneity. control theory plays a particularly important and relevant role here, as design of robust distributed feedback loops under decentralized information is an indispensable component of an effective architecture. This plenary t
Recently, a new family of protocols has been introduced for large scale ad hoc networks that makes use of the approximate location of nodes in the network for geography-based routing. Location management plays an impo...
详细信息
Recently, a new family of protocols has been introduced for large scale ad hoc networks that makes use of the approximate location of nodes in the network for geography-based routing. Location management plays an important role in such protocols, and previous work in this area has shown that the asymptotic overhead of location management is heavily dependant on the service primitives (location registration, maintenance and discovery) supported by a location management protocol. Currently, SLALoM (C.T. Cheng et al., 2002), which is a grid-based protocol optimized for large node movements, achieves the best known upper bound on the asymptotic worst case overhead of location management. However, the location registration cost in SLALoM dominates other costs for all practical purposes, and thus novel schemes need to be designed to limit this controltraffic. In this work, we use the idea of location forwarding to devise a new scheme called ELF that limits the signalling traffic, and thus enhances the scalability of location management in large ad hoc networks. We find that, while the asymptotic overhead cost by such an improvisation matches that of SLALoM, ELF outperforms SLALoM in average case scenarios.
The paper proposes a novel ip-over-optical network interconnection model that takes the best features from both the overlay and peer models while avoiding their limitations. Specifically, the proposed model utilizes a...
详细信息
The paper proposes a novel ip-over-optical network interconnection model that takes the best features from both the overlay and peer models while avoiding their limitations. Specifically, the proposed model utilizes an optical layer-based unified control plane that manages both routers and optical switches (analogous to the peer model), while still retaining complete separation between the optical and ip layers of the overlay model. This is achieved by shifting the control plane functionalities previously associated with the ip layer to the ip/MPLS-aware, non-traffic bearing OXC controller modules located within the optical domain. In this architecture, better decisions can be made for provisioning and managing network resources, leading to their more efficient use. Based on the proposed model, an integrated dynamic routing algorithm, that takes into account the combined topology and resource usage information at both the ip and optical layers, is developed.
Composite star network is an alternative architecture for the backbone of the next generation Internet as it provides agility and means of QoS to applications. We present performance evaluation of a composite star net...
详细信息
Composite star network is an alternative architecture for the backbone of the next generation Internet as it provides agility and means of QoS to applications. We present performance evaluation of a composite star network based on a crossbar, a Benes or a log/sub 2/(N, m, p) architecture. Crosstalk property of these architectures are analyzed and compared. Based on the analysis of switch properties, crosstalk property of the composite star network is analyzed. As path length is short in such network, photonic switches with large size can be adopted. Our results indicate that crossbar switch is not suitable for this network as its crosstalk performance is quite poor. The results can be used as guides to designing of photonic switches.
A new terabit level switching fabric: multi-ary hypercube switching fabric (MHSF) and the related Hamming random routing algorithm (HRRA) is proposed in this paper. Theoretical analysis shows that when each switching ...
详细信息
ISBN:
(纸本)7563506861
A new terabit level switching fabric: multi-ary hypercube switching fabric (MHSF) and the related Hamming random routing algorithm (HRRA) is proposed in this paper. Theoretical analysis shows that when each switching node evenly routes traffic to the other nodes, MHSF is stable if and only if the speedup factor of links between them is larger than M/k. In order to improve scalability for practical applications, this paper also proposes a small-granularity MHSF expansion method, keeping its stability characteristics.
In this paper, we show that significant simplicities can arise in the analysis of a network when link capacities are large enough to carry many flows. In particular, we prove that, when an upstream queue serves a larg...
详细信息
In this paper, we show that significant simplicities can arise in the analysis of a network when link capacities are large enough to carry many flows. In particular, we prove that, when an upstream queue serves a large number of regulated traffic sources, the queue-length of the downstream queue converges almost surely to the queue-length of a simplified queueing system (single queue) obtained by removing the upstream queue. We provide similar results (convergence of the queue-length in distribution) for general (including nonregulated) traffic arrivals. In both cases, the convergence of the overflow probability is uniform and at least exponentially fast. Through an extensive numerical investigation, we demonstrate several aspects and implications of our results in simplifying network analysis.
Inter vehicle communication (IVC) has become a major topic during the last few years. Within the FleetNet project a novel mobile ad hoc network will be developed - based on the UMTS terrestrial radio access time divis...
详细信息
Inter vehicle communication (IVC) has become a major topic during the last few years. Within the FleetNet project a novel mobile ad hoc network will be developed - based on the UMTS terrestrial radio access time division duplex (UTRA TDD) air interface - to interconnect vehicles and vehicles with roadside gateways via a mobile Internet. In this paper we focus on the impact of vehicular traffic dynamics on protocols for ad hoc networks. Based on analytical treatment of vehicular traffic and on realistic traffic scenarios, we deduce requirements and their interdependencies for the developed ad hoc networking protocols, and verify them by simulation. We show that the proposed MAC and RRM protocols are suitable for inter vehicle communication with its high dynamics in freeway environments.
Future communications networks are expected to support applications with quality of service (QOS) requirements. Supporting QoS poses major challenges due to the large size and complex structure of networks. Several st...
详细信息
Future communications networks are expected to support applications with quality of service (QOS) requirements. Supporting QoS poses major challenges due to the large size and complex structure of networks. Several studies have addressed different aspects of QoS routing. In this paper, we present a new approach of QoS routing based on the concept of multiple routing tables. traffic with QoS requirements will have their own routing table. This scheme can prevent such traffic from concentrating on some link, so it can improve their performances. From our extensive simulation results, we have found that both packet loss of QoS traffic and throughput of best-effort traffic can be improved by using our scheme.
The paper presents a solution for providing differentiated capacity allocation in an access network. The system is based on CLAMP (curtailing the large TCP advertised window to maximize performance), an algorithm that...
详细信息
The paper presents a solution for providing differentiated capacity allocation in an access network. The system is based on CLAMP (curtailing the large TCP advertised window to maximize performance), an algorithm that can differentiate between flows sharing the same FIFO queue. The system is suitable for access networks, such as those based on DSL and HFC modems and wireless LAN access points. The deployment of CLAMP is completely contained within the access network; no changes to the remainder of the network are required. CLAMP provides the opportunity to enforce local policies on TCP flows that originate from sources distributed globally. The performance of CLAMP is verified by both simulation and analysis.
暂无评论