Conference Paper

Failure Aware Diverse Routing: A Novel Algorithm to Improve Availability in ASON/GMPLS Networks

Opt. Commun. Group, Univ. Politecnica de Catalunya, Barcelona
DOI: 10.1109/ICTON.2006.248434 Conference: Transparent Optical Networks, 2006 International Conference on, Volume: 3
Source: IEEE Xplore

ABSTRACT The average increment and the increasing variability of the traffic to be carried by ASON/GMPLS transport networks will require on-demand bandwidth provisioning in the near future. In this scenario dynamic routing based on traffic engineering techniques is a key issue. This paper deals with dynamic routing in ASON/GMPLS networks based on using network failures information. We define three routing strategies which differ in the use of failure information. Their performance is evaluated by simulation in terms of request blocking probability and availability


Available from: Jaume Comellas, Jun 03, 2015
  • [Show abstract] [Hide abstract]
    ABSTRACT: Extended use of MPLS technology in transport networks which improves routing performance introduced some more sophisticated requirements like Advanced Network Management, Quality of Service (QoS) and Resource Optimization. Generalized Multiprotocol Label Switching (GMPLS) architecture mostly meet these requirements however, connection and network availability needs to be considered. In this paper, we propose a new survivable connection provisioning scheme based on Spare Capacity Allocation (SCA) to maximize resource utilization while guaranteeing availability requirement defined in Service Level Agreement (SLA). We also present two additional class-based heuristic methods, namely Least Reliable Path (LRP) and Exchange Method (EM) in order to minimize blocking probability of incoming connection requests. We compare the performance of the optimization-based model and the proposed heuristics by simulation. The simulation results show that the LRP method leads to the least blocking probability while the SCA-based connection provisioning and the EM demonstrates the same blocking probability level. Furthermore, it is also shown that the heuristics do not introduce additional resource overbuild to the network when compared to the SCA-based model for different load levels.
    Proceedings of the 15th IEEE Symposium on Computers and Communications, ISCC 2010, Riccione, Italy, June 22-25, 2010; 01/2010
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we propose an availability analysis method for overlapping shared segment protection. We validate the proposed analysis method for different traffic demand matrices and different failure rate values. We show that the proposed analysis method achieves estimating the availability of a connection with a negligible error ratio. We then present two connection provisioning schemes, namely availability constrained generalized segment protection (AC-GSP), and share ability driven availability constrained generalized segment protection (SDAC-GSP) that are availability-aware adaptation of a conventional segment selection algorithm, namely the generalized segment protection (GSP). We evaluate the performance of these proposed schemes, and we show that as an enhancement to GSP, SDAC-GSP enhances the performance of AC-GSP in terms of the availability per connection, blocking probability and availability satisfaction ratio. However, it utilizes more wavelengths in comparison to AC-GSP since the more segments protects the links of the working path, the higher availability.
    Computer and Information Sciences, 2008. ISCIS '08. 23rd International Symposium on; 11/2008
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we present two adaptive schemes, namely, global differentiated availability-aware provisioning (G-DAP) and link-by-link differentiated availability-aware provisioning (LBL-DAP) for connection provisioning in optical transport networks under dual failure assumption. G-DAP attempts to determine a feasible sharing degree for each availability class for the links in the network globally. As an enhancement to G-DAP, LBL-DAP tries to determine a feasible sharing degree per availability class on each link. Connections belong to one of the five availability classes with the availability requirements of 0.98, 0.99, 0.999, 0.9999, and 0.99999, which are specified in their service level agreement (SLA). We compare the performance of the proposed schemes with a conventional service provisioning scheme, compute a feasible solution (CAFES) under the NSFNET and European optical network (EON) topologies. The simulation results show that the proposed schemes introduce better availability per connection and lower blocking probabilities. LBL-DAP performs the best by monitoring the shareability status of each link periodically with respect to the metrics studied, and decreases resource overbuild.
    Journal of Lightwave Technology 11/2009; DOI:10.1109/JLT.2009.2025246 · 2.86 Impact Factor