NeTS-NBD: Designing Next-Generation Robust Telecom Networks

该项目将在设计、分析和优化新型系统架构、网络协议和软件算法方面创造新的知识，用于构建健壮的通信网络。此类网络的示例包括可存活的下一代光波分复用（WDM）骨干网，它可以(1)在多路径配置的情况下容纳不同带宽粒度和不同路径长度的连接，以及(2)防止多种网络故障，包括网络区域灾难。智力影响：该项目有三个主要的研究组成部分，如下所述。首先，本项目将研究网状网络中差异化保护服务的问题。注意到不同的连接在其可用性上可能有不同的服务水平协议（SLA），我们的目标是选择适当的保护机制（例如，专用或共享或根本没有保护），以经济有效地提供连接请求以满足其SLA。其次，该项目将研究可生存的多路径供应。具体而言，将设计新的保护和配置策略，以利用新一代SONET/SDH等网络中新兴的反向多路复用技术。通过反向多路复用在多个路径上分离和提供连接，即使在（多个）故障的情况下（相对于在目前的网络操作模式下根本没有服务），也可以实现降级服务。第三，本项目将研究重置保护带宽的问题。特别是，研究人员将研究备份重新配置以处理多个并发故障，这意味着在先前的故障被修复之前发生新的故障。这种备份重新配置策略可以更有效地利用防止网络故障所需的带宽资源。更广泛的影响：该项目的方法和结果可以具有更广泛的适用性，例如，多协议标签交换（MPLS）和其他网络中的带宽供应。除了通过学生培训加强国家教育基础设施外，该项目预计将对与pi有着密切工作关系的电信行业产生重大影响。因此，该项目的研究成果可以帮助提高国家（和全球）通信基础设施的稳健性，并带来所有随之而来的好处。

This project will create new knowledge on the design, analysis, and optimization of novel system architectures, network protocols, and software algorithms for building robust communication networks. Examples of such networks include survivable next-generation optical wavelength-division-multiplexing (WDM) backbone networks that can (1) accommodate connections of diverse bandwidth granularities and diverse path lengths in case of multi-path provisioning, and (2) protect against multiple network failures, including network-area disasters. Intellectual Impact: The project has three major research components, as outlined below.First, the project will study the problem of differentiated protection services in mesh networks. Noting that different connections may have different service-level agreements (SLAs) on their availabilities, the objective is to select a proper protection mechanism (e.g., dedicated or shared or no protection at all) to cost-effectively provision a connection request to meet its SLA.Second, the project will investigate survivable multi-path provisioning. Specifically, new protection and provisioning strategies will be designed to exploit emerging inverse-multiplexing technologies in networks such as Next-Generation SONET/SDH. By splitting up and provisioning a connection over multiple paths through inverse multiplexing, one can achieve degraded service even in case of (multiple) failures (vs. no service at all as in present mode of operation of networks).Third, the project will investigate the problem of reprovisioning protection bandwidth. In particular, the investigators will study backup reprovisioning to handle multiple concurrent failures, which means that a new failure occurs before a previous failure is repaired. Such backup reprovisioning strategies can lead to more efficient use of the bandwidth resources that are needed to protect against network failures.Broader Impact: The approaches and results of this project can have much broader applicability, e.g., bandwidth provisioning in Multi-Protocol Label Switching (MPLS) and other networks. Besides enhancing the national educational infrastructure through student training, this project is expected to have a strong influence on the telecom industry, with which the PIs have a close working relationship. Thus, research results from this project can help to improve the robustness of the national (and global) communication infrastructure, with all the attendant benefits.