Showing papers by "Helen C. Leligou published in 2006"

Efficient medium arbitration of FSAN-compliant GPONs: Research Articles

Abstract: The steadily rising demand for multimedia and data services, the falling cost and omnipresence of Ethernet and the maturity of passive optical networks (PON) technology, promise to radically change the landscape in the local loop. The heart of a gigabit PON system (recently standardized by FSAN/ITU) is the medium access controller (MAC), which arbitrates access to the upstream link among users with fluctuating traffic demands and effects the multiplexing and concentration policy. At the same time, it has to safeguard the service quality and enforce the parameters agreed in the service level agreements (SLAs) between the users and the service provider. In this paper, a MAC protocol designed to serve any mix of services according to their quality of service (QoS) needs, employing four priority levels along with a high number of logically separate data queues is presented. The architecture and implementation in hardware of a MAC algorithm capable of allocating bandwidth down to a resolution of a byte with QoS differentiation is the focus of this paper. It employs the bandwidth arbitration tools of the FSAN/ITU G.984.3 standard and maps SLA parameters to GPON service parameters to create an efficient, fair and flexible residential access system. Copyright © 2005 John Wiley & Sons, Ltd.

Slotted optical switching with pipelined two-way reservations

Abstract: Optical burst switching is a core architecture designed to reconcile the available optical technology with the increasing burstiness of traffic. However, disappointing performance in terms of high packet loss and/or low system utilization discouraged broader experimental implementations. A method to avoid these losses by first sending over the control channel a short scout packet that simulates the events that the actual burst will experience is proposed in this paper. Once the scout message detects a drop at any intermediate node, it returns back to the source to avert the payload emission and repeat the process. The way the control works results in essential service quality features, i.e., no loss of bursts, no out-of-order emissions, increased efficiency, much reduced delay variation, and graceful throttling of the load respecting the contracted rates

Hybrid burst/packet switching architectures from IP NOBEL

Abstract: In spite of its long term promise, all-optical switching is still plagued by high cost, low efficiency when handling bursty data traffic, immature management and protection and poor output port contention resolution leading to heavy loss. Given the current situation, hybrid approaches that keep the best features of optics, reverting to the electrical plane when expedient, constitute sensible interim steps that can offer cost-effective solutions along the road to an eventual all-optical core. Two such approaches developed in the framework of the European IP project NOBEL are presented in this work. The first is a quite mature solution that extends present day concepts to achieve multiplexing gain while keeping all the management and restoration benefits of SDH. The other mimics early LANs in executing a distributed switching via its electrical control plane using two-way reservations, thus restricting its applicability to smaller domains. Combining the two leads to a system fulfilling most of today's requirements for Tb/s core networks.

Clustering core nodes allows efficient all-optical payload switching

Abstract: Clustering neighboring nodes of an all-optical core network into medium-sized rings featuring reservation-based control, allows lossless aggregation of bursts and a reduction of possible source-destination pairs. Both act towards reducing the system cost and burst loss creating an architecture that obviates most of the expensive and loss-prone all-optical switches by delegating the switching function to a medium access control protocol coordinating the distributed laser transmitters. This is in principle the approach taken in early LANs, but with the optical switches the incentive is not limited to avoiding their high cost but also the high burst loss arising from the very limited optical buffering. The end result is a core architecture with much lower losses than one-way systems, much lower delay than solutions employing two-way reservations end-to-end, still achieving multiplexing gain.

Probe-and-go: a novel approach to OBS

Abstract: Optical burst switching achieves multiplexing gain in the optical domain but cannot reach reasonable utilization before burst losses become unacceptable. A method to avoid these losses by first sending over the control channel a short scout packet that simulates the events that the actual packet will experience, is proposed in this paper. Once the scout message detects a drop at any of the intermediate nodes, the actual packet is not sent but the process repeated.