Deh-phone Hsing

A restoration methodology based on pre-planned source routing in ATM networks

This paper describes a restoration methodology for ATM networks that is built on top of a pre-planned source routing system in the network and operates on a connection-by-connection basis. Unlike most of the existing proposals that use dedicated spare facilities to protect network traffic against network fault conditions, this scheme simply takes advantage of unused network capacity to redirect traffic from the problem area. This restoration methodology requires the upstream neighboring switches to send messages back to the corresponding sources of the affected connections to find an alternative route. Given that this capability is being incorporated for new call setup in the signalling protocol for private ATM networks by the ATM Forum, the additional development cost and operations overhead associated with this restoration methodology is relatively insignificant. The performance of the proposed restoration methodology is discussed based on simulation studies for video on demand (VoD) applications.

A network management architecture for battlefield networks

Managing the future digital battlefield networks presents unique challenges: (1) the need to enable seamless, end-to-end network management interoperability across different network management systems that involves multiple network administrative domains, and heterogeneous network technologies, (2) the need to support user, terminal, and network equipment mobility management with large user population and wide geographical area coverage, and (3) the need to facilitate rapid network deployment and fault-tolerant network management. In this paper we describe a CORBA-based network management architecture to address these challenges. In particular, we propose to support network management applications utilizing a distributed object database framework where network information is managed, accessed, and manipulated as a collection of distributed CORBA objects. We will integrate the TMN tiered principle with the manager/agency distributed management paradigm to support management scalability for battlefield networks. An application of the proposed architecture to realize network self-configuration capabilities is also presented.

Comparison of restoration strategies based on pre-planned routing methodologies in ATM networks

The need for battlefield digitization, coupled with the growing demand for transporting diverse and bandwidth intensive applications both in the military and commercial sectors, have led to a rapid growth in broadband networks. The success of these broadband networks, however, is critically tied to their robustness. We propose and study two restoration strategies to restore the affected connections in the event of network facility failures in broadband ATM networks. The restoration strategies are characterized by the fact that they use the untapped resources of the underlying network rather than relying on dedicated backup. Such strategies will not only help in providing cost-effective network solutions for robustness in commercial communications networks, but will also be very useful in a scenario with diminishing network resources such as a battlefield. In a battlefield, the network elements such as switches and/or links can fail sporadically due to hostile conditions, causing a decrease in the available resources. Further, the proposed restoration strategies operate on a connection-by-connection basis thereby allowing network operators to provide a high degree of robustness for mission critical and premium service applications, and a comparatively lower degree of robustness to low priority applications depending on their specific QoS requirements.

Self-configuration management system design for tactical battlefield networks

As the impetus to extend computer and communications as far possible into the battlefield continues to grow, configuration management of these computer and communication networks becomes a challenging task. The challenges stem mainly from the volatile connectivity (due to network element failures) and sporadic mobility (due to the dynamics of the battlefield theatre). This paper proposes a self-configuration management system for such battlefield networks. The issues addressed and resolved by the proposed design are: (a) self-configuration-the ability to adapt and configure the network without human intervention when the network elements relocate, and (b) robustness-the ability to provide continued service in the midst of random network facility failures.

Trunking performance analysis of AAL2 for VBR voice transport

This paper analyzes the trunking performance of ATM Adaptation Layer Type 2 (AAL2) for variable bit rate (VBR) voice transport. In order to transmit voice traffic over ATM networks, unstructured and structured circuit emulation services using AAL1 were proposed as possible standard-based methods. However, although AAL1 is offered as a standard-based solution by some ATM equipment manufacturers, it is not considered an optimum solution, in terms of bandwidth utilization efficiency, for voice transport due to the lack of support of multiple individual voice users per ATM connection, compression, and silence suppression. In order to satisfy the critical need in the market for voice-over-ATM, AAL2 is defined in the ITU-T Recommendation I.363.2 which is designed with the capabilities to support multiple voice users per ATM connection and transport compressed/silence-suppressed VBR voice traffic. In this paper, we conduct detailed quantitative analysis of the trunking performance of AAL2.

Directory-enabled network management framework for battlefield networks

While the emerging CORBA technology holds many desirable characteristics of a distributed object system it is equally attractive as a platform for integrating different systems that will co-exist in the battlefield networks. In this paper, we investigate a framework that integrates emerging Java and directory service technologies with the CORBA based network management system to provide directory service enabled network management (DEMAN) capability. It is envisioned that this framework will enable DEMAN capability to be made available sooner (by using more matured technologies) and cover a wider portion of the battlefield network (by including Java-capable network elements).

Design of an adaptive configuration management architecture for tactical battlefield networks

Management of tactical battlefield networks presents a host of challenges. The challenges arise due to the volatile and sometimes, unpredictable nature of these networks. This is because, in a battlefield environment, the network elements (switches/routers) can appear and disappear sporadically, due to jamming, hostile attacks, etc. This results in sporadic changes in the network connectivity. Furthermore, mission re-assignments can occur and may also result in frequent traffic pattern shift. Thus an auto-configuration management system that can adapt to the dynamic network conditions is extremely important. In this work, we have designed an adaptive configuration management system (ACMS) that is sensitive to the battlefield dynamics. We use the concept of directory service systems to facilitate automated configuration management. We have also developed an initial design framework of a simulation model that can be used to demonstrate our ACMS design.

A CORBA-based interoperable auto-configuration system design for battlefield networks

The rapidly changing network topology of battlefield networks warrants an efficient configuration management system for the battlefield network to function efficiently. Furthermore, it is very important to provide uninterrupted services to the mission critical applications in the tactical battlefield network. In this work we propose an interoperable auto-configuration management system for the armys tactical networks. The interoperability stems from the fact that we support communications both through the tactical Internet and an ATM-based tactical backbone network. The proposed design also enhances network robustness by allowing for reconfiguration after a facility failure, thereby providing uninterrupted service to the mission critical applications. In addition, we propose the use of a directory service system to facilitate configuration management in the dynamic battlefield environment. In order to accommodate a distributed and heterogeneous computing environment, a CORBA-based architecture is proposed to realize a practical implementation of the proposed auto-configuration management system design.

Modeling and simulation study of the survivability performance of ATM-based restoration strategies for the next generation high-speed networks

The insatiable need for high bandwidth and low latency is spurring research into the next generation of high-speed networks (NGHSN). The design of these NGHSN will be typically motivated by the use of advanced networking technologies. The recent technological advancements in the field of WDM, render WDM-based networks a promising architectural candidate for the future NGHSN. Since ATM have the capacity for providing QoS guarantees, we propose a candidate architecture that combines the two technologies, i.e., ATM and WDM for the NGHSN. Furthermore, we observe that service survivability (i.e., continued service provision in the event of network facility failures) is a very important concern since service disruption at high-speeds can cause a considerable amount of loss. Hence, in this paper, we propose and analyze the performance of a restoration strategy that can tap the advantages of the two technologies to provide network robustness/service survivability.

A restoration strategy based on PNNI routing in ATM networks

As network speeds increase, the effects of network facility failures (link and/or switching node) can be devastating both to the end users and service providers, depending on the nature of the affected applications. Hence robustness in broadband communication networks, i.e., continued service provision in the event of network failures, is an important issue and has received significant attention. While most prior work in this regard has focused on incorporating redundancy to provide uninterrupted services, in this paper we propose and analyze a restoration scheme based on the PNNI (private network-network interface) signalling and routing protocols for ATM networks. This restoration scheme utilizes the existing unused resources of the underlying network in order to restore the affected connections during network failures. Detailed performance and complexity trade-offs analysis has been conducted. Our results indicate that for the network studied in this paper, a restoration scheme based on PNNI is an attractive option.