A multi-layered model for survivability in intra-LATA transport networks

Abstract

The Intra-LATA transport network is modeled into 4 layers (from high to low): The switched layer, ~O S S -C O M K~ layer, multiplex layer, and physical layer. Capacity provided to meet demand at a given layer in turn becomes demand for the next lower layer. Thus, survivability provided for a subnetwork at a given layer protects the demand originating at that layer as well as higher layer subnetworks.In this paper we define the layered model, give a general characterization of survivability restoral methods, and then compare principal existing and proposed methods at each layer, especially the ~OSS-COMIZC~ layer. In particular, we discuss routing efficiency, speed, control architectures (centralized and distributed), feasibility, and, most important to Intra-LATA carriers, potential integration with other operations functions, such as demand provisioning, network operations, customer control, and network planning.

Rerouting in advance for preempted IR calls in QoS-enabled networks

“…From an implementation perspective, Doverspike [12] classified the rerouting approaches along three axes: link rerouting versus end-to-end rerouting; centralized vs. distributed schemes; and pre-computation versus dynamic computation rerouting schemes. The first of these focuses on the network point (e.g., source or unlink node) at which rerouting has to be initiated, while the second focuses on the controlling point of rerouting.…”

Rerouting Technique for Faster Restoration of Preempted Calls

“…The reactive schemes are free from the overheads of maintaining resources along a back-up path, but causes long connection rerouting times and/or no rerouting at all, if the network is highly loaded. Doverspike [13] investigated the rerouting approaches from an implementation perspective and grouped them along three axes: link rerouting vs end-to-end rerouting; centralized vs distributed schemes; pre-computation vs dynamic computation rerouting schemes. The link rerouting vs end-to-end rerouting focuses on the network point (e.g., source or unlink node) at which rerouting has to be initiated, while the centralized vs distributed schemes concentrates on the controlling point of rerouting.…”

Redundant trees for preplanned recovery in arbitrary vertex-redundant or edge-redundant graphs

“…The study of self-healing networks is often classified according to the following three criteria (e.g., [8], [16]): the use of link (line) rerouting versus path (or end- to-end) rerouting, the use of centralized computation versus distributed computation, and the use of precomputed versus dynamically computed routes. A succint comparison of the different options can be found in [75, pp.…”