Service level agreement (SLA) is a contract between the service provider (SP) and the customer. Among various parameters, SLA imposes penalties on the SP when the customer's quality-of-service (QoS) requirements are violated. Hence, it is desirable to avoid SLA violations and such penalties. SLA also states the maximum amount of downtime a connection may tolerate during its lifetime. In this study, we present a dynamic provisioning scheme called Preemption-Oriented SLA-Aware Provisioning (POSAP) for telecom mesh networks, which considers a connection's (1) availability requirements, (2) penalty, and (3) state. In addition to the connection's various parameters (availability, penalty, etc.), the amount of affordable downtime is used as a metric to determine the risk of violating the connection's SLA, and hence, its priority. Connection's priority is denoted by urgency level. Our scheme allows resource preemption among connections, i.e., high-priority connections may preempt backup resources from low-priority connections. Also, new connection requests may preempt resources from existing connections. Urgency level determines how resources can be preempted. Our scheme is applied to a US-wide network. Our results show the improved performance in terms of (1) reduced SLA violations, (2) increased network utilization, and (3) increased network revenue, compared to a dedicated primary-backup (P-B) scheme.

Survivability is a very important issue in telecom networks. Many previous works discuss availability-guaranteed and service-differentiated provisioning for survivable network design. Our work makes two new contributions. First, our work proposes a novel provisioning algorithm to explore the effect of link sharing among primary and backup paths. Second, the solution approach applies multi-backup protection scheme for those services which have extremely high availability requirements (relative to the network component availabilities). We develop a mathematical model to quantify a connection's availability with N (Nges1) backup paths with k-link (kges1) sharing between primary and backup paths. Based on the model, we propose a provisioning algorithm for dynamic connections with differentiated availability requirements. A connection can be either unprotected, or protected with N backup paths. The algorithm intelligently explores link sharing among primary and backup paths of a connection to target availability guarantee and resource efficiency. The illustrative numerical examples show that our scheme achieves better performance in both availability satisfaction and network resource utilization, compared to a traditional availability-constrained 1+1 provisioning strategy.

Survivability is a key concern in modern network design in order to achieve fast service restorability against network failures. This paper investigates the problem of survivable dynamic connection provisioning in general telecom backbone networks, which are mesh structured. These networks employ optical fibers, which may fail due to network outages such as fiber cuts, etc. Our study applies to survivability of optical wavelength-division multiplexing (WDM) as well as multi-protocol label switching (MPLS) networks. For survivability study, we assume differentiated services where connections may have different availability requirements, so they may be provisioned differently with protection (if needed) based on their availability requirements and current network state. Therefore, it may be possible that connections with the same source, destination, and availability requirement are provisioned differently (unprotected, shared-path protected, or dedicated-path protected) at different times based on current network state. Such differentiated provisioning can provide diverse levels of service performance and achieve network resource optimization flexibly. Our main contributions are as follows. First, we develop an analytical model to quantify the availabilities of connections with various protection modes, i.e., unprotected, dedicated-path protected, and shared-path protected. Particular emphasis is placed on computing a connection's availability with shared-path protection by employing the link-vector technique, because this technique can maximally explore the sharing potential among backup paths and achieve bandwidth-assignment flexibility. Based on the mathematical model, we then present a novel provisioning strategy for dynamic connection requests in which multiple levels of services are provided and different protection schemes may be applied to different connections. The strategy jointly considers both connection availability satisfaction and resource optimization. Numerical results show very good accuracy of our model and high effectiveness of our provisioning strategy

Optical burst switching (OBS) is a promising technique for supporting high-capacity, bursty data traffic over optical wavelength-division-multiplexed (WDM) networks. An optical link failure may result in a huge amount of data (and revenue) loss, and it has been an important survivability concern in optical networks. In this paper, we study the fault- management issues related with a link failure in an OBS network. We propose to use pre-planned global rerouting to balance network load and to reroute bursts after a link fails. We apply optimization techniques to pre-plan explicit backup routes for failure scenarios. Our objective is to achieve optimal load balancing both before and after a failure such that the network state can still remain stable with minimum burst-loss probability when a failure occurs. We apply the pre-planned normal and backup routing tables to an OBS network, and study the network performance after a failure occur using illustrative numerical examples. The results show that the average burst-loss probability can be significantly reduced by 60% - from an average of 0.10 to 0.04 (when the normalized link load is less than 0.5) using globally-rerouted backup routes, when compared with the scheme without global rerouting. We also observe that the burst-loss probability is reduced by 43% - from an average of 0.07 to 0.04 (when the link load is less than 0.5) if the rerouting is done using optimization techniques, when compared with shortest-path routing.for Fault Management

Cost is a major concern in telecom industry. Cost-efficient network design has been attracting the attention of network operators (NO). Wavelength-division multiplexing (WDM) optical technology is being widely deployed in telecom networks for their high capacity. Backbone telecom networks, which are mesh-connected, employ optical crossconnects (OXC) at switching nodes; and OXC deployments today are of the opaque (optical-electronic-optical, OEO) variety. In such a network, the number of OXC ports is a major portion of the cost of the switching equipment (not counting the cost of the transmission equipment, which is needed anyway independent of the switching equipment). In this work, we focus on a practical situation in which a bundle of ports are fabricated on a single blade, called a line card (LC). If we are restricted to using only bundles of ports per LC, this will impose an extra constraint on cost-efficient network design, especially for asymmetric traffic between node pairs and arbitrary network topologies. Our design relies on a two-step approach using two algorithms whose objective is to reduce the number of LCs (ports). First, we route traffic connections on low-cost paths where cost depends on how much resources are needed to support the connection. The second step is a reassignment step, which tries to re-optimize the outcome of the first step. Simulation was conducted on a sample backbone network. Results show a reduction in cost (in terms of the number of deployed LCs) by nearly 6%.

Survivability is a very important problem in telecom networks. Many previous works discuss availability-guaranteed and service-differentiated provisioning. Our work makes two new contributions. First, we consider link sharing between a primary path and backup path(s). In order to achieve better capacity utilization, a connection can share some highly reliable links in common among its primary and backup paths. Second, we consider 1+N (Nges2) protection schemes for those connections which have extremely high availability requirements or have always-on service quality demands. Multiple backup paths are provided to an availability-stringent connection, compared to the traditional 1+1 protection scheme which may have to block some high-availability connections. We develop a mathematical model to quantify a connection's availability with 1+N (Nges2) protection with k-link (kges1) sharing. Based on the model, we propose a provisioning algorithm for dynamic connections with differentiated availability requirements. A connection can be either unprotected, 1+1 protected, or 1+2 protected using our current model with the 1+N extension being an open problem. One-link sharing is allowed among primary and backup paths. The illustrative numerical results show that our approach achieves better performance in both blocking probability and resource utilization, compared to a traditional 1+1 provisioning strategy.

Survivability is a key concern in modern telecom mesh networks because of the enormous capacity of a telecom link which is usually an optical fiber employing wavelength-division multiplexing (WDM). Backup bandwidth reprovisioning has been shown to be an effective approach for improving network survivability as well as preventing existing services from unnecessary interruption. We investigate the advantages of reprovisioning new backup paths for connections when a previous failure is repaired (as well as when a network failure occurs). We consider reprovisioning of backup paths either (1) for unprotected or vulnerable connections that lose their primary or backup paths due to a previous failure and fail to be reprovisioned when the failure happens due to resource limits, or (2) for all existing connections in the network. The pros and cons of the two policies are investigated. A wavelength-convertible network and shared-path protection are assumed in this study. We compare the performance of our dynamic reprovisioning approach with a conventional scheme which reprovisions backup paths for connections only when a network failure occurs. The simulation results demonstrate that our approach achieves more network robustness and better backup capacity optimization.