Quality, Reliability, Security and Robustness in Heterogeneous Networks. 7th International Conference on Heterogeneous Networking for Quality, Reliability, Security and Robustness, QShine 2010, and Dedicated Short Range Communications Workshop, DSRC 2010, Houston, TX, USA, November 17-19, 2010, Revised Selected Papers

Research Article

Distributed Scheduling for Advance Bandwidth Reservation in High-Performance Networks

Download145 downloads
  • @INPROCEEDINGS{10.1007/978-3-642-29222-4_27,
        author={Qishi Wu and Yunyue Lin},
        title={Distributed Scheduling for Advance Bandwidth Reservation in High-Performance Networks},
        proceedings={Quality, Reliability, Security and Robustness in Heterogeneous Networks. 7th International Conference on Heterogeneous Networking for Quality, Reliability, Security and Robustness, QShine 2010, and Dedicated Short Range Communications Workshop, DSRC 2010, Houston, TX, USA, November 17-19, 2010, Revised Selected Papers},
        proceedings_a={QSHINE},
        year={2012},
        month={10},
        keywords={distributed scheduling bandwidth reservation high- performance networks},
        doi={10.1007/978-3-642-29222-4_27}
    }
    
  • Qishi Wu
    Yunyue Lin
    Year: 2012
    Distributed Scheduling for Advance Bandwidth Reservation in High-Performance Networks
    QSHINE
    Springer
    DOI: 10.1007/978-3-642-29222-4_27
Qishi Wu1,*, Yunyue Lin1,*
  • 1: University of Memphis
*Contact email: qishiwu@memphis.edu, ylin1@memphis.edu

Abstract

High-performance networks are capable of provisioning dedicated channels through circuit/lambda-switching or MPLS/GMPLS techniques to support large-scale data transfer. These dedicated links are typically shared by multiple users through advance resource reservations, resulting in varying bandwidth availability in future time periods. Most previous efforts were focused on centralized bandwidth scheduling to improve the utilization of network resources and meet the transport requirements of application users. These centralized scheduling schemes imply the use of a central control plane, posing significant reliability and scalability challenges as the network size rapidly grows. We propose distributed algorithms for path computation and bandwidth scheduling in response to four basic bandwidth reservation requests: (i) fixed bandwidth in a fixed slot, (ii) highest bandwidth in a fixed slot, (iii) first slot with fixed bandwidth and duration, and (iv) all slots with fixed bandwidth and duration. These algorithms are developed through a rigorous extension of the classical breadth first search and Bellman-Ford algorithms to a completely distributed manner, and their performances are evaluated and analyzed through extensive simulations.