Search

Abstract: The exponential increase in the data center network traffic has posed new challenges for achieving high throughput, low latency and energy consumption. To address these challenges, this paper presents a novel scheme that combines optical circuit and packet switching technologies together in a single optical cross-connect for data center networks. By utilizing fast tunable linecards and parallel array waveguide grating routers, the optical cross-connect can offer low latency, large scalability and high throughput in datacenter networks. A scheduling technique is designed to simultaneously accommodate circuit switching and packet switching in the optical cross-connect. The physical performance of this optical cross-connect in circuit and packet switching modes is investigated by experiments. In addition, the network performance of the optical cross-connect is evaluated through simulations under a mixed circuit/packet-switched traffic pattern. Results show that the network performance of the optical cross-connect does not degrade significantly with an increase in the number of racks, but it is affected by the circuit-switched traffic proportion. [Copyright &y& Elsevier]

Abstract: The basic standard protocol for supporting network mobility (NEMO) (i.e., NEMO-BSP specified by IETF) introduces several performance problems, such as multiple tunneling overhead and packet delivery latency, because it exploits mobile IPv6 (MIPv6), which was proposed for host mobility. To improve the basic NEMO solution, two network-based NEMO approaches, rNEMO and N-PMIPv6, have been proposed. The rNEMO is able to reduce significant packet tunneling overhead. The N-PMIPv6, on the other hand, reduces location update cost, but it leads to packet tunneling overhead. Thus, they commonly waste network resources in both wired and especially wireless network. No efficient and practical solutions for minimizing both the location update overhead and packet tunneling overhead have been presented until now. This situation motivated us to propose a resource-efficient network mobility scheme (RENEMO), reducing resource utilization required for network mobility support. We show that the proposed RENEMO outperforms both rNEMO and N-PMIPv6 in terms of network resource. [Copyright &y& Elsevier]

Abstract: This paper presents PI-OPS (Parallel-Iterative Optical Packet Scheduler) a parallel-iterative scheduler for asynchronous Optical Packet Switching nodes with optical buffering. Optical packets are assembled by aggregating IP packets, and attaching an optical packet header. Conventional schemes process optical packet headers one by one, in a sequential form. Then, the worst case algorithm response time is tightly coupled to switch size. In contrast, in PI-OPS all the optical packets received during a given time window are jointly processed to optimize the delay and output wavelength allocation, applying void filling techniques. The scheduler has a deterministic response time, independent of the traffic arrival’s pattern. In addition, PI-OPS has been specifically designed to allow a parallel electronic implementation similar to the ones found in VOQ schedulers. In this respect, we evaluate the traffic loss performance of the scheduler in different settings, to dimension a set of hardware related parameters. Finally, we conduct an emulation of an FPGA implementation of a large-scale version of the scheduler. Results support the feasibility of its implementation. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, an optically controlled loop buffer is proposed and experimentally demonstrated. It utilizes a semiconductor optical amplifier (SOA) based loop mirror as a fast switch to realize the writing and reading of packets within this buffer and a circulator based optical reflector to achieve internal re-circulations. By adding fibers with different lengths between the loop mirror and the circulator, variable delay time per circulations can be obtained. We also demonstrate the performance of the 2.5 and 10Gb/s data packets in this buffer structure, as well as the influence of the two different control signal formats on the packet signal degradation during buffering. This loop buffer has the merits of low cost and improved performance compared to other similar loop buffer structures. According to the obtained results, the achieved number of circulations is greatly increased. [Copyright &y& Elsevier]

Abstract: In mobile ad hoc networks (MANETs), flooding is a required message dissemination technique for network-wide broadcast. The conventional blind flooding algorithm causes broadcast storm problem, a high number of unnecessary packet rebroadcasts thus resulting in high contention and packet collisions. This paper proposes a new probabilistic approach that dynamically fine-tunes the rebroadcasting probability of a node for routing request packets (RREQs) according to the number of neighbour nodes. We evaluate the performance of the proposed approach for the ad hoc on demand distance vector (AODV) routing protocol and compared against the blind flooding, fixed probabilistic and adjusted probabilistic flooding [L.M.M.M. Bani-Yassein, M. Ould-Khaoua et al., Performance analysis of adjusted probabilistic broadcasting in mobile ad hoc networks, International Journal of Wireless Information Networks 13(2) (2006) 127–140; M.B. Yassein, M.O. Khaoua et al., Improving route discovery in on-demand routing protocols using local topology information in MANETs, Proceedings of the ACM international workshop on Performance Monitoring, Measurement, and Evaluation of Heterogeneous Wireless and Wired Networks, Terromolinos, Spain, ACM Press, 2006, pp. 95–99.] approaches. The simulation results show that our proposed approach demonstrates better performance than blind flooding, fixed probabilistic and adjusted flooding approaches. [Copyright &y& Elsevier]

Abstract: Metropolitan Area Networks (MANs) are considered one of the best arenas for an early introduction of optical packet switching technologies. However, there is still no clear definition of how these technologies should be exploited in the design of innovative MANs. In this paper we consider different architectures that have been studied in the framework of the European Network of Excellence e-Photon/ONe and analyze their advantages and limitations. The main contribution of this work is the proposal of an analytical framework to compare these architectures in terms of resource demand when dimensioning them for a given traffic matrix and, from this analysis, discuss issues and trade-offs arising in the design of optical packet MANs. [Copyright &y& Elsevier]

Abstract: System performance of Gigabit network hosts can be severely degraded due to interrupt overhead caused by heavy incoming traffic. One of the most popular solutions to mitigate such overhead is interrupt coalescing in which a single interrupt is generated for multiple incoming packets. This is opposed to normal interruption in which an interrupt is generated for every incoming packet. In this paper we investigate the performance of interrupt coalescing analytically and compare it with that of normal interruption. We consider two types of coalescing (viz. count-based and time-based). The system performance is studied in terms of throughput, CPU availability for user applications, latency and packet loss. [Copyright &y& Elsevier]

Abstract: The performance of a Differential-Phase-Shift-Keying Optical Code-Division Multiplexing (DPSK-OCDM) system with balanced detection is investigated in the case in which Gold codes are used. Allowing more optical codes to be supported, the performance in Packet Loss Probability of a DPSK-OCDM Optical Packet Switch, using both code and wavelength domains to solve output packet contentions, may be increased. In the proposed case study the increase is of three orders of magnitude when the offered traffic is 0.7. [Copyright &y& Elsevier]