Skip to main content
SpringerLink
Log in

Path Computation for Dynamic Provisioning in Multi-Technology Multi-Layer Transport Networks

  • Original Research
  • Published:
SN Computer Science Aims and scope Submit manuscript

Abstract

Service providers employ different transport technologies such as PDH, SDH/SONET, OTN, DWDM, Ethernet, MPLS-TP to support different types of traffic and service requirements. A typical transport network element supports adaptation of multiple technologies and multiple layers of those technologies to carry the input traffic. Further, transport networks are deployed such that they follow different topologies like linear, ring, mesh, protected linear, dual homing etc. in different layers. Dynamic service provisioning requires the use of on-line algorithms that automatically compute the path to be taken to satisfy the given service request. Path computation algorithms can be implemented in Path Computation Element (PCE) which can be invoked from Transport SDN controller to automate service provisioning. This paper studies automated path computation for service requests considering the above factors where, a new mechanism for building an auxiliary graph that models each layer as a node within each network element and creates adaptation edges between them and also supports creation of special edges to represent different types of topologies, is proposed. Logical links that represent multiplexing or adaptation are also created in the auxiliary graph. An initial weight assignment scheme for non-adaptation edges that consider both link distance and link capacity is introduced along with three dynamic weight assignment functions that consider the current link utilization. Path computation algorithms considering adaptation and topologies are proposed over the auxiliary graph structure. The performance of the algorithms is evaluated and it is found that the weighted number of requests accepted is higher and the weighted capacity provisioned is lesser for one of the dynamic weight function and certain combination of values proposed as part of the weight assignment. It is found that the proposed approach results in better overall network utilization (improvement of up to 30 Gbps for a scenario with 50,000 service requests) and fragmentation compared to the traditional layered path computation approach for a representative large-scale service provider transport network (Network 1) with 2955 network elements, 5753 physical links and 480 hub nodes. It is also found that the proposed approach results in better overall network utilization (3–4 times lesser utilization for a scenario up to 50,000 service requests) and fragmentation compared to the traditional layered path computation approach for another representative large-scale service provider transport network (Network 2) generated randomly with 2040 network elements and more than 7000 physical links.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12

Similar content being viewed by others

References

  1. Generic functional architecture of transport networks. ITU-T Recommendation G.805. 2000.

  2. Bhide NM, Sivalingam KM, Fabry-Asztalos T. Routing mechanisms employing adaptive weight functions for shortest path routing in optical WDM networks. Photonic Netw Commun. 2001;3(3):227–36.

    Article  Google Scholar 

  3. Dijkstra F, Andree B, Koymans K, Van Der Ham J, Grosso P, de Laat C A multi-layer network model based on ITU-T G. 805. Computer Networks 2008; 52(10):1927–1937

  4. Dijkstra F, van der Ham J, Grosso P, de Laat C. A path finding implementation for multi-layer networks. Future Gen Comput Syst. 2009;25(2):142–6.

    Article  Google Scholar 

  5. Gerstel O, Jinno M, Lord A, Yoo SB. Elastic optical networking: a new dawn for the optical layer? IEEE Commun Mag. 2012;50(2):s12–20.

    Article  Google Scholar 

  6. Gong L, Zhou X, Liu X, Zhao W, Lu W, Zhu Z. Efficient resource allocation for all-optical multicasting over spectrum-sliced elastic optical networks. J Opt Commun Netw. 2013;5(8):836–47.

    Article  Google Scholar 

  7. Iqbal F, van der Ham J, Kuipers F. Technology-aware multi-domain multi-layer routing. Comput Commun. 2015;62:85–96.

    Article  Google Scholar 

  8. Janz C, Ong L, Sethuraman K, Shukla V. Emerging transport SDN architecture and use cases. IEEE Commun Mag. 2016;54(10):116–21.

    Article  Google Scholar 

  9. Kuipers F, Dijkstra F. Path selection in multi-layer networks. Comput Commun. 2009;32(1):78–85.

    Article  Google Scholar 

  10. Lamali ML, Fergani N, Coheny J, Pouyllau H Path computation in multi-layer networks: Complexity and algorithms. In: Proceedings of IEEE INFOCOM, pp 2016; 1–9

  11. Lamali ML, Fergani N, Cohen J. Algorithmic and complexity aspects of path computation in multi-layer networks. IEEE/ACM Trans Netw. 2018;26(6):2787–800.

    Article  Google Scholar 

  12. Liu S, Lu W, Zhu Z. On the cross-layer orchestration to address ip router outages with cost-efficient multilayer restoration in ip-over-eons. IEEE/OSA J Opt Commun Netw. 2018;10(1):A122–32.

    Article  Google Scholar 

  13. Lopez-Pajares D, Alvarez-Horcajo J, Rojas E, Carral JA, Ibanez G Iterative discovery of multiple disjoint paths in switched networks with multicast frames. In: 2018 IEEE 43rd Conference on Local Computer Networks (LCN), IEEE, 2018; pp 409–412

  14. Lu P, Zhang L, Liu X, Yao J, Zhu Z. Highly efficient data migration and backup for big data applications in elastic optical inter-data-center networks. IEEE Netw. 2015;29(5):36–42.

    Article  Google Scholar 

  15. Lu W, Yin X, Cheng X, Zhu Z. On cost-efficient integrated multilayer protection planning in ip-over-eons. J Lightwave Technol. 2018;36(10):2037–48.

    Article  Google Scholar 

  16. Madanagopal R, Rani NU, Gonsalves TA Path computation algorithms for dynamic service provisioning in SDH networks. In: \(10^{th}\) IFIP/IEEE Symposium on Integrated Management (IM), 2007; pp 206–215

  17. Madanagopal R, Rani NU, Gonsalves TA. Path computation algorithms for dynamic service provisioning with protection and inverse multiplexing in SDH/SONET networks. IEEE/ACM Trans Netw. 2010;18(5):1492–504.

    Article  Google Scholar 

  18. Martín B, Sánchez Á, Beltran-Royo C, Duarte A. Solving the edge-disjoint paths problem using a two-stage method. Int Trans Oper Res. 2020;27(1):435–57.

    Article  MathSciNet  Google Scholar 

  19. Martinez A, Yannuzzi M, Lopez V, Lopez D, Ramirez W, Serral-Gracia R, Masip-Bruin X, Maciejewski M, Altmann J. Network management challenges and trends in multi-layer and multi-vendor settings for carrier-grade networks. Commun Surv Tutor IEEE. 2014;16(4):2207–30.

    Article  Google Scholar 

  20. Patel A, Jue J. Routing and scheduling for variable bandwidth advance reservation. IEEE/OSA J Opt Commun Netw. 2011;3(12):912–23.

    Article  Google Scholar 

  21. Prakash A, Kannan R, Bafna G Disjoint path computation systems and methods in optical networks. US Patent 10,003,867 2018

  22. Ramachandran M, Sivalingam KM Topology based path computation for provisioning in transport networks. In: \(9^{th}\) International Conference on Communication Systems and Networks (COMSNETS), 2017; pp 166–173

  23. Szczesniak I, Jajszczyk A, Wozna-Szczesniak B. Generic dijkstra for optical networks. IEEE/OSA J Opt Commun Netw. 2019;11(11):568–77.

    Article  Google Scholar 

  24. Tanaka J Path generating method, relay device, and computer product. US Patent 8,665,754 2014

  25. Xu D, Chiang M, Rexford J. Link-state routing with hop-by-hop forwarding can achieve optimal traffic engineering. IEEE/ACM Trans Netw. 2011;19(6):1717–30.

    Article  Google Scholar 

  26. Yen JY. Finding the k shortest loopless paths in a network. Manag Sci. 1971;17(11):712–6.

    Article  MathSciNet  Google Scholar 

  27. Yin Y, Zhang H, Zhang M, Xia M, Zhu Z, Dahlfort S, Yoo SB. Spectral and spatial 2d fragmentation-aware routing and spectrum assignment algorithms in elastic optical networks. IEEE/OSA J Opt Commun Netw. 2013;5(10):A100–6.

    Article  Google Scholar 

  28. Zhu Z, Lu W, Zhang L, Ansari N. Dynamic service provisioning in elastic optical networks with hybrid single-/multi-path routing. J Lightwave Technol. 2012;31(1):15–22.

    Article  Google Scholar 

  29. JGrapht - a Java library of graph theory data structures and algorithms. https://jgrapht.org/.

Download references

Acknowledgements

This work was supported by a Mid-Career Institute Research and Development Award (IRDA) from IIT Madras (2017–2020) and by a DST grant (EMR/2016/003016) from Government of India (2017–2020).

Author information

Authors and Affiliations

  1. Department of Computer Science and Engineering, Indian Institute of Technology Madras, Chennai, India

    Madanagopal Ramachandran & Krishna M. Sivalingam

Authors
  1. Madanagopal Ramachandran
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Krishna M. Sivalingam
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Madanagopal Ramachandran.

Ethics declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

An earlier version of this manuscript is available in arXiv and can be accessed using the link http://export.arxiv.org/pdf/1901.01531.

Appendix A

Appendix A

Preservation of Shortest Path Property

The proposed graph transformations preserve the shortest path property under the consideration that working and protection paths follow same set of topologies for protected path computation. The assumption is that the service provider transport networks are structured such that their core, aggregate and access segments of the network are formed as proper ring or dual homing ring predominantly. Wherever linear segments are present in the network, they would get protected by one or more parallel linear segments.

The Dijkstra’s shortest path algorithm is the base for protected path computation also and its proof of correctness is based on the induction on the number of visited nodes. As part of the invariant hypothesis, it is proved that the unvisited node’s cost is set as the sum of the cost of the sub-path up to the previous visited node and the cost of the current edge from the visited node to the unvisited node. Since the protected path computation removes the normal edges in the rings in the first step and considers only the special edges for the rings as described later in this paper, it has to be shown that the protected path computation using the special edges would result in the shortest path based on their weights. This is shown below for the three types of ring considered for which graph transformation by means of special edge creation is applicable.

For core ring, the ring is created such that it is transformed into full mesh of special edges. The weight for special edges are set as a fraction (a constant) of the sum of the weights of the involved edges in the core ring as described later in this paper. In the case of parallel or concentric core rings, it can be observed that the ring with the least weight would be chosen between a pair of nodes in those core rings by the proposed transformation. In the case of non-parallel rings available between a pair of nodes, it can be observed that the set of core rings whose sum of weights is lesser would be found by the shortest path algorithm.

For ring with aggregate node, the special edges are created from all the nodes in the ring to the aggregate node that transforms the ring into a star topology. The weight for special edges are set as a fraction (a constant) of the sum of the weights of the involved edges in the ring with aggregate node as described later in this paper. In this case also, it can be observed that the shortest path would involve the rings whose path cost is lesser for both parallel and non-parallel scenario.

For dual homing ring, a special hub node is created and special edges are created between each node in the dual homing ring to the hub node and from the hub node to the aggregate nodes of the dual homing ring. The weight for special edges to the hub node are set as a fraction (a constant) of the sum of the weights of the involved edges in the dual homing ring as described later. The weight for special edges from hub node to the aggregate nodes are set as zero as described later. Since a fraction of the sum of individual link weights are set as the special edge weight, the shortest path would use dual homing rings with lesser weights and so it can be observed that this transformation also preserves the shortest path property for both parallel and non-parallel dual homing ring scenario.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ramachandran, M., Sivalingam, K.M. Path Computation for Dynamic Provisioning in Multi-Technology Multi-Layer Transport Networks. SN COMPUT. SCI. 1, 304 (2020). https://doi.org/10.1007/s42979-020-00319-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1007/s42979-020-00319-4

Keywords

Search

Navigation