S. Sedef Savas

Network adaptability to disaster disruptions by exploiting degraded-service tolerance

The rapid increase in network traffic with new bandwidth-hungry applications such as cloud computing and telemedicine makes disaster survivability a crucial concern as the data (and revenue) loss caused by large-scale correlated cascading failures can be very high. To alleviate their impact, new measures should be taken since the nature of the network changes dramatically as available resources decrease during disasters. We develop a metric, called degraded-service tolerance, which can reduce protection cost and network disruption, and support maximal carried traffic in case of disasters. Degraded-service-tolerant connections can be admitted and recovered with reduced bandwidth under resource crunch. Our scheme re-assigns resources among connections by leveraging their degraded-service tolerance. A case study shows how our proposal can be applied to boost network performance during the resource crunch following a disaster.

Disaster-aware service provisioning by exploiting multipath routing with manycasting in telecom networks

Disasters may cause large-area failures in high-capacity telecom networks, leading to huge data loss. Survivable service provisioning is crucial to minimize the effects of network / datacenter failures and maintain critical services in case of a disaster. We propose a novel disaster-aware service-provisioning scheme that multiplexes service over multiple paths destined to multiple servers/datacenters with manycasting. Our scheme maintains some bandwidth (i.e., reduced service) after a disaster failure vs. no service at all. Numerical examples show that our approach offers high level of survivability against link and node failures that may be caused by disasters and post-disaster failures at no extra cost compared to the other survivable schemes.

Traveling repairman problem for optical network recovery to restore virtual networks after a disaster [invited]

Virtual networks mapped over a physical network can suffer disconnection and/or outage due to disasters. After a disaster occurs, the network operator should determine a repair schedule and then send repairmen to repair failures following the schedule. The schedule can change the overall effect of a disaster by changing the restoration order of failed components. In this study, we introduce the traveling repairman problem to help the network operator make the schedule after a disaster. We measure the overall effect of a disaster from the damage it caused, and we define the damage as the numbers of disconnected virtual networks, failed virtual links, and failed physical links. Our objective is to find an optimal schedule for a repairman to restore the optical network with minimum damage. We first state the problem; then a mixed integer linear program (MILP) and three heuristic algorithms, namely dynamic programming (DP), the greedy algorithm (GR), and simulated annealing (SA), are proposed. Finally, simulation results show that the repair schedules using MILP and DP results get the least damage but the highest complexity; GR gets the highest damage with the lowest complexity, while SA has a good balance between damage and complexity.

Disaster-aware service provisioning with manycasting in cloud networks

Cloud services delivered by high-capacity optical datacenter networks are subject to disasters which may cause large-area failures, leading to huge data loss. Survivable service provisioning is crucial to minimize the effects of network/datacenter failures and maintain critical services in case of a disaster. We propose a novel disaster-aware service-provisioning scheme that multiplexes service over multiple paths destined to multiple servers/datacenters with manycasting. Our scheme maintains some bandwidth (i.e., degraded service) after a disaster failure vs. no service at all. We formulate this problem into a mathematical model which turns out to be an Integer Linear Program (ILP), and we provide heuristic optimization approaches as ILP is intractable for large problem instances. Numerical examples show that exploiting manycasting by intelligently selecting destinations in a risk-aware manner for service provisioning offers high level of survivability against link and node failures that may be caused by disasters and post-disaster failures at no extra cost compared to the other survivable schemes.

Disaster-resilient control plane design and mapping in software-defined networks

Communication networks, such as core optical networks, heavily depend on their physical infrastructure, and hence they are vulnerable to man-made disasters, such as Electromagnetic Pulse (EMP) or Weapons of Mass Destruction (WMD) attacks, as well as to natural disasters. Large-scale disasters may cause huge data loss and connectivity disruption in these networks. As our dependence on network services increases, the need for novel survivability methods to mitigate the effects of disasters on communication networks becomes a major concern. Software-Defined Networking (SDN), by centralizing control logic and separating it from physical equipment, facilitates network programmability and opens up new ways to design disaster-resilient networks. On the other hand, to fully exploit the potential of SDN, along with data-plane survivability, we also need to design the control plane to be resilient enough to survive network failures caused by disasters. Several distributed SDN controller architectures have been proposed to mitigate the risks of overload and failure, but they are optimized for limited faults without addressing the extent of large-scale disaster failures. For disaster resiliency of the control plane, we propose to design it as a virtual network, which can be solved using Virtual Network Mapping techniques. We select appropriate mapping of the controllers over the physical network such that the connectivity among the controllers (controller-to-controller) and between the switches to the controllers (switch-to-controllers) is not compromised by physical infrastructure failures caused by disasters. We formally model this disaster-aware control-plane design and mapping problem, and demonstrate a significant reduction in the disruption of controller-to-controller and switch-to-controller communication channels using our approach.

Exploiting Degraded-Service Tolerance to Improve Performance of Telecom Networks

Degraded-service-tolerant connections can operate with reduced bandwidth under failure conditions. We propose a provisioning scheme that accepts degraded services not only during failures but also during admission process to increase service acceptance and/or availability.

Network adaptability to disasters by exploiting degraded-service tolerance

This paper studies optimal traffic provisioning under service heterogeneity to alleviate the dramatic decrease in network resources, a.k.a. resource crunch, caused by large-area network failures or a traffic surge. Some services are sensitive to capacity provided, while others (e.g., video streaming) can operate with reduced bandwidth (degraded service). Degraded-service-tolerant connections can be admitted and recovered with reduced service under resource crunch. Our adaptive solutions distribute capacity among competing services when some may get degraded service under resource crunch but satisfying the Service-Level Objectives (SLOs).

Application-aware software-defined EPON access network

As bandwidth requirements of users in passive optical networks continue to increase rapidly, especially due to the growth of video-streaming traffic, the need to resolve bandwidth contention among competing users and applications becomes more compelling. To address this problem, we propose application-aware software-defined Ethernet passive optical network (EPON) architecture. It utilizes application-level feedback from the client side (for video users) to the network, through a software-defined network controller, to achieve better client- and service-level differentiation in the downstream direction of an EPON access network. We adopt adaptive video streaming that utilizes feedbacks regarding network congestion for improving Quality of Experience (QoE) of video-streaming clients. Numerical results for video-streaming applications demonstrate that the proposed architecture for downstream resource allocation considerably reduces video stalls, increases video buffer levels, and also reduces video-switching rate, leading to better QoE for users as a result of better client- and service-level differentiation.