Traffic grooming refers to techniques used to combine low-speed traffic streams onto highspeed wavelengths in order to minimize the networkwide cost in terms of line terminating equipment and/or electronic switching. Such techniques become increasingly important for emerging network technologies, including SONET/WDM rings and MPLS/MPλS backbones, for which traffic grooming is essential. In this article we formally define the traffic grooming problem, and we provide a general formulation that captures the features of a wide range of problem variants. We then present a comprehensive comparative survey of the literature that unveils the significant amount of research on this subject (the traffic grooming past). We also offer a broad set of ambitious research directions (the traffic grooming future) that are motivated by the exciting new challenges arising with the advent of MPλS technology.

No abstract

No abstract

We present an architecture for a core dWDM network which utilizes the concept of Optical Burst Switching (OBS) coupled with a Just-In-Time (JIT) signaling scheme. It is a reservation based architecture whose distinguishing characteristics are its relative simplicity, its amenability to hardware implementation, support for quality of service and multicast natively. Another important feature is data transparency-the network infrastructure is independent of the format of the data being transmitted on individual wavelengths. In this article we present a brief overview of the architecture and outline the most salient features. This research effort is being supported through a contract with ARDA (Advanced Research and Development Activity, http://www.ic-arda.org).

We study the problem of constructing multicast trees to meet the quality of service requirements of real-time, interactive applications operating in high-speed packet-switched environments. In particular, we assume that multicast communication depends on (a) bounded delay along the paths from the source to each destination, and (b) bounded variation among the delays along these paths. We rst establish that the problem of determining such a constrained tree is NP-complete. We then derive heuristics that demonstrate good average case behavior in terms of the maximum inter-destination delay variation of the nal tree. In addition, our heuristics achieve their best performance under conditions typical of multicast scenarios in high-speed networks. We also show that it is possible to dynamically reorganize the initial tree in response to changes in the destination set, in a way that is minimally disruptive to the multicast session.

We present a detailed analysis of the JIT, JET, and Horizon wavelength reservation schemes for optical burst switched (OBS) networks. Our analysis accounts for several important parameters, including the burst offset length, and the optical switching and hardware processing overheads associated with bursts as they travel across the network. The contributions of our work include: (i) analytical models of JET and Horizon (on a single OBS node) that are more accurate than previously published ones, and which are valid for general burst length and offset length distributions; (ii) the determination of the regions of parameter values in which a more complex reservation scheme reduces to a simpler one; and (iii) a new reservation scheme, JIT + , which is as simple to implement as JIT, but whose performance tracks that of Horizon and JET. We compare the performance of the four wavelength reservation schemes on a single OBS node, as well as on a path of OBS nodes with cross traffic, under various sets of parameter values. Our major finding is that, under reasonable assumptions regarding the current and future state-of-the-art in optical switch and electronic hardware technologies, the simplicity of JIT and JIT + seem to outweigh any performance benefits of Horizon and JET.

No abstract

No abstract