Saravut Charcranoon

A preemptive scheduling technique for OBS networks with service differentiation

Existing burst scheduling techniques could be improved in terms of bandwidth utilization and QoS support. In this paper, we introduced a new partially preemptive scheduling technique with QoS support. The technique is capable of handling data bursts in parts and may use preemption due to the priorities of data bursts in a multi-service OBS network environment. Simulation studies suggest that more than 50% reduction in dropping probability and approximately 40% improvement in channel utilization is reachable at 0.8 load. The studies also reveal that the new scheduling technique has the ability to predict and control service performance differentiation among defined service classes.

Parallel processor configuration design with processing/transmission costs

A computer configuration design problem where the objective is to configure a parallel processor to do processing in a cost effective manner is examined. The application envisioned is a specialized on-line service that rents time on its machine. The combinatorial optimization problem involved is examined analytically and a heuristic algorithm for its solution is provided. Lessons learned in this work appear in the conclusion.

Group-scheduling for optical burst switched (OBS) networks

To date, all of the proposals for burst-scheduling techniques have considered scheduling individual bursts. We introduce a new scheme called OBS group scheduling. In this approach, a burst is represented by an interval of time. The process of scheduling a number of bursts, thus, turns to be a process of fitting a set of the corresponding time intervals on a channel time line that represents a channel-time resource. By doing so, we can formulate the scheduling process as a combinatorial optimization problem. Then, graph theory is applied to schedule as many non-overlapping intervals as possible onto the channel time line. The underlying concept of the group scheduling is that of briefly delaying the scheduling of a burst so that a much better decision can be made about a number of bursts all-together. This scheme is shown, through simulations, to improve performance in terms of burst loss probability and channel utilization over existing schemes.

Load sequencing for a parallel processing utility

The monetary cost optimization of a computer utility load distribution problem is examined. The problem is to find the sequence in which to distribute divisible computing load from a root processor to its children processors which achieves the lowest monetary distribution cost. The convergence performance of a heuristic greedy algorithm is studied. This problem is directly relevant to computer utilities which offer computing and software hosting to organizations for a monetary charge.

Operation and maintenance issues in optical burst-switched networks

We discuss a number of operation and maintenance (OAM) issues in optical burst-switched (OBS) networks and propose an OAM framework in this respect. Four OBS functions are identified as bases for the proposed framework. These are the burst termination function, the burst transmission function, the switching function, and the routing and switching control function. The study focuses on the unique aspects of OBS networks and how they can be addressed from an OAM point of view. We propose an OAM architecture and investigate the functional, network, and node aspects of it. Novel OAM-capable structures for core and edge OBS nodes are introduced for the first time.

Procedures and functions for operation and maintenance in optical burst-switching networks

Very little work has been carried out so far in the area of operation and maintenance (OAM) schemes for optical bursts switched (OBS) networks. We continue to discuss OAM issues for OBS networks. The spectral and temporal de-coupling of data and control in an OBS network calls for special OAM attention. In this paper, OAM procedures, functions and sessions based on specific network and node architectures are discussed. An OAM information model, encompassing OAM-specific packet/burst formats is also presented.

Group-scheduling for multi-service optical burst switching (OBS) networks

We propose a new burst scheduling mechanism for Optical Burst-Switched (OBS) networks. The new approach is made possible by gathering data bursts into groups and performing the scheduling decision for each group collectively. In OBS group-scheduling, bursts will not be considered for scheduling until a pre-defined time period elapses, during which the group of burst header packets would be gathered. By transforming a set of data bursts into a set of corresponding time intervals, the problem of scheduling these bursts is transformed into a combinatorial optimization problem. Graph algorithms are applied to obtain the maximum number of non-overlapping bursts. The proposed OBS group-scheduling scheme is shown to improve the performance of OBS networks over existing scheduling schemes in terms of burst loss probability and channel utilization. With an extension through a sequential optimization, using a Branch-and-Bound technique, the proposed scheme can support multiple classes of service. It is shown that the new scheduling approach has several desired characteristics including fairness and service differentiability among classes in terms of burst loss probability and channel utilization.

Framework for Operation and Maintenance (OAM) in Optical Burst Switched Networks

The problem of operation and maintenance (OAM) in optical burst switched (OBS) networks is not addressed by the optical-networking community even though OBS has been gaining research interest in recent years. In this paper, we introduce a functional model of OBS networks and an OAM architecture that is designed to meet the operational requirements based on this model. We present the first framework to realize OAM in OBS networks, including OAM activities, OAM-aware OBS nodes, OAM information and communication models and protocols. A number of practical examples of possible network failures and the corresponding reactions are presented.

Performance of group scheduling with look-ahead algorithm for optical burst switching networks

An algorithm to reduce data burst processing delay in group scheduling in core nodes of optical burst switching networks has been proposed. Since, in this algorithm, look-up tables containing all the void time information in scheduling windows are generated as soon as the primary group scheduling session terminates, it becomes faster to reassign dropped data bursts to proper voids, if any, by referring to the tables. The group scheduling with this algorithm showed almost the same performance as the previous group scheduling in regard to channel utilization and wavelength conversion rate. On the other hand, per-burst processing time has been reduced dramatically in the load region of higher than 0.8 and burst loss probability has been improved more under lighter traffic condition.