Survivability Analysis of Telephone Access Network ∗ Yun Liu † , Veena B. Mendiratta ‡ , and Kishor S. Trivedi † † CACC, Department of ECE, Duke University, Durham, NC 27708, USA liu, kst@ee.duke.edu ‡ Bell Labs, Lucent Technologies, 1960 Lucent Lane, Naperville, IL 60566, USA veena@lucent.com Abstract The telecommunications industry has achieved high re- liability and availability for telephone service over decades of development. However, the current design does not aim at providing service survivability when a local switching of- fice fails due to catastrophic damage. In this paper, several survivable architectures for telephone subscriber network are proposed based on common survivability principles. In order to quantitatively assess the effectiveness of design al- ternatives, a set of analytical models are developed to de- rive various survivability measures. Numerical results are provided to show how a comprehensive understanding of the system behavior after failure can be achieved through different survivability aspects. 1. Introduction Today’s telecommunications industry has achieved high quality, reliability and availability over decades of develop- ment. A number of requirements, methods, and techniques have been established and standardized by Telcordia [1] to quantitatively evaluate the capability of a system for pro- viding the desired service. On the other hand, due to the increasing competition and the growing dependence of the entire society on telecommunications services, survivability has become a crucial issue for vendors, service providers, and government agencies. Although survivability is emerg- ing as a new requirement for telephone networks, many sur- vivability issues such as definition, architecture, and quan- tification have not been sufficiently studied. The survivability research and development of telephone networks has mostly concentrated attention on two aspects: redundancy in switches and survivable transport network architectures. The former is accomplished by having all * This research was supported by an AFOSR MURI grant no. F49620- 1-0327. the essential components in switching offices duplicated to avoid a single point of failure [5]. When a primary com- ponent fails, the standby unit can promptly detect the fail- ure and take over the ongoing traffic without incurring cus- tomer perceivable call loss or service degradation. The lat- ter is realized by the evolution of robust transport archi- tectures from Automatic Protection Switching (APS) point- to-point protection to bi-directional self-healing fiber ring technologies, and multiplex section protection in the fu- ture [9]. However, these survivability enhancement features do not ensure of the continuity of service when a central office (CO) is physically destroyed by natural disasters or man-made accidents/attacks. The reason is two fold. First, due to the nature of the damage, the primary and redun- dant components will be destroyed together since in prac- tice they are co-located in the same CO, usually in the same box. Second, unlike Toll/Tandem offices and Toll centers, local central offices are at the lowest level in the telephone network hierarchy and have no office level redundancy in subscriber access networks. When a CO is down, customers will lose all the telephone services in the coverage area and have to wait for repair, which takes from several to tens of days. In order to improve survivability in telephone access net- works this paper proposes a comprehensive solution from architecture design to quantitative analysis. From an ar- chitecture perspective, a general reference hierarchical ar- chitecture is defined and described for subscriber networks. This layered abstraction facilitates the further presentation of alternative architectures for different functional elements, the choice of various survivability definitions, the quantifi- cation of survivability measures, and the interpretation of numerical results. From the perspective of analytical mod- eling, this paper discusses three typical survivability defi- nitions and shows how they can be quantified for assessing different facets of survivability. Note that these definitions are proposed and studied in other areas such as critical in- formation system and optical transport network. Both the definitions and the corresponding analysis methods cannot