Ciril Rozic

Optical protection cost of IP fast reroute on a fully connected IP network over a WDM ring

Optical protection bandwidth costs for IP fast reroute for a fully connected IP network over a WDM ring is reduced with failure notifications from the WDM network. The costs are compared with MPLS fast reroute.

Benefits of multi-layer application-aware resource allocation and optimization

Internet traffic is generated by a multitude of applications, each one with diverse service requirements in terms of bandwidth, latency, reliability, etc. Today traffic engineering techniques can provide service differentiation at the IP/MPLS layer, but not at the optical layer. In this paper we propose a framework where application service requirements drive a dynamic multi-layer (IP/MPLS and optical) resource allocation and optimization. We compare by means of simulations such application-aware algorithmic framework with a multi-layer but application-unaware strategy. Results show that the application-aware approach, unlike the application-unaware one, is always able to guarantee the specified service requirements to those applications whose generated traffic is accepted by the network. In addition, the application-aware strategy does not consume more network resources than the application-unaware one, but only requires a network that is more dynamic and responsive.

Cost of loop-free alternates in IP-over-WDM networks

In Internet protocol (IP) networks, the failure recovery time of IP restoration, of several hundreds of milliseconds or up to seconds, is sufficient for most applications. But protection switching is needed for services that require very short recovery times of at most 50 ms. Loop-free alternates (LFAs) is a protection switching mechanism that is in the IP layer. We consider IP-over-WDM (wavelength division multiplexed) networks where IP routing eventually restores all traffic, while LFAs protect the fraction of the traffic that has high priority. Simulations show that if 50% of the traffic has high priority, then the additional cost to protect the high-priority traffic using LFAs is just a percent on average, under a particular cost model. Thus, with very little additional network cost, LFAs can guarantee that a large fraction of traffic will be protected between all source-destination pairs. In addition, a network design algorithm is presented that, given a traffic matrix and a physical network topology, will find a low-cost network design that includes determining the IP network topology, IP-link capacities, lightpath routing, and LFAs.

The cost of loop free alternates in IP over WDM

In IP networks, the failure recovery time of IP restoration is sufficient for most applications. But some services require very short recovery times (50 ms) that can be achieved only by protection switching. Loop-free alternates (LFAs) is a protection switching mechanism that is in the IP layer. We consider IP over WDM networks where IP routing eventually restores all traffic, while LFAs protect the fraction of the traffic that has high priority. Simulations show that if 50% of the traffic has high priority then the additional cost to protect the high priority traffic using LFAs is just a few percent, e.g., 1% on average. Thus, with very little additional network cost, LFAs can guarantee that a large fraction of traffic will be protected between all source-destination pairs. In addition, a network design algorithm is presented that, given a traffic matrix and physical network topology, will find a low cost network design that includes determining the IP network topology, IP link capacities, lightpath routing, and LFAs.

Protection costs of a fully connected IP core network over a WDM ring

Protection costs, optical and electronic, for a fully connected IP core network over a WDM ring network are compared under two implementations of IP Fast Reroute. The particular IP core network considered has its routers paired, and client routers are dual homed to the pairs. The IPFRR schemes are also compared with MPLS Fast Reroute.

Cost Efficient Survivable IP Over WDM With Dual Homing

Dual homing lightpath service is considered to support survivable IP over WDM networks, where a prescribed fraction of network bandwidth must survive fiber-link faults. Network costs can decrease by as much as 32%.

A Framework for Dynamic Multi-layer Resource Allocation in Application-Centric Networking

In an SDN-based network, connection requests can be accommodated according to application requirements. We devise a framework where such requirements drive IP and optical network resource allocation, dynamic optimization, and instantiation through an SDN orchestrator.

Network primitives based on latency and recovery time in orchestrated multi-layer networks

Other than merely a bandwidth pipe, current and future network applications will have other requirements such as maximum end-to-end latency and the type of recovery from failure. The added complexity will require a central intelligence such as a software-defined networking (SDN) orchestrator to compute and keep track of the resources allocated to such applications. In such an SDN-orchestrated network, we examine a planning scenario where not all the requirements are known in advance. Instead, they are computed by the network orchestrator and referred to as network primitives. The primitives are then presented to the applications so that the applications know what type of service the network can offer. With our planning algorithm, we show an example of computing the primitives by analysing the guaranteed maximum end-to-end latency and post-failure recovery time.