The rapid evolution of policy-based systems spans multiple application areas including QoS management within networks, security management and trust specifications for e-commerce. Policies provide a flexible means of adapting system behavior as they can be changed dynamically in order to change the management strategy and system response without interrupting its operation. Furthermore, access and admission control policies permit a clear, enforceable and auditable specification of the access privileges granted to users or other agents within the system.

This half-day tutorial will explain what is policy, why it is needed and will give an overview of the ongoing policy work in both the network management and the security communities. This will include the current activities within the IETF/DMTF communities, the Policy Definition Language (from Lucent), the Cfengine approach to system's management as well as the work on Role Based Access Control Models and IBM's Trust Policy Language. The Ponder policy framework including the language and associated management tools will also be presented. The tutorial will draw upon these various frameworks in order to analyze critical issues and characteristics of policy-based systems. This includes approaches to policy specification and policy languages, issues relating to policy analysis and conflict detection, architectural issues relating to implementation and deployment of policies in a distributed environment, scalability of policy frameworks as well as issues relating to refining high-level enterprise goals into implementable policies.

This tutorial is mainly intended for network managers, application vendors, architects and programmers working on policy-based systems. Researchers and research managers investigating policy-based solutions for systems management and security should also benefit from this tutorial.

The Java 2 Platform, Enterprise Edition (J2EE) defines the industry standard for developing multi-tier, vendor-independent enterprise applications. J2EE simplifies enterprise applications by basing them on standardized, modular components, by providing a complete set of services to those components, and by handling many details of application behavior automatically, without complex programming.

This tutorial takes an in-depth look at and demonstrates some of the hot J2EE topics such as: Enterprise JavaBeans, JavaServer Pages, Transactions, JNDI, Servlets, and Security plus new features such as Messaging Services and the Java Connector Architecture.

This tutorial is mainly intended for software architects, developers and project managers interested in knowing how J2EE meets telecom requirements such as scalability, reliability, security and manageability. Individuals interested in this tutorial are encouraged to consider attending tutorial T-7, which presents a case study of how the J2EE platform can be utilized to build large-scale telecom operations support systems.

The purpose of this tutorial is to elaborate on this subject by presenting an overview of the main technical solutions available for QoS in IP networks and their related strategic issues. In brief, the presentation will focus on:

• An overview of the principles and strategies adopted by the main standardized solutions (IntServ, DiffServ, MPLS and GMPLS);
• Their applicability and market-share; and
• Management and main research issues related to the technical aspects discussed.

This tutorial will be of interest to network engineers, network designers, and technical staff involved in managing and designing IP networks. A previous knowledge of IP and networking technologies basics is required.

This tutorial discusses the issues and solutions of management of dynamic networks and services based on distributed event correlation technologies. Distributed event correlation is becoming a widely accepted solution for managing the complexity and scale of modern telecommunication/enterprise networks and services. It has been used for various network management tasks, but the majority of the applications have so far been for network fault management. New important applications of event correlation are emerging in the area of network performance, traffic and service management.

The focus of this tutorial will be on the following challenging aspects of these applications: 1) extending the management models with dynamic properties of managed networks and services, 2) unified management using cross-domain correlation models, and 3) utilization of advanced management enabling technologies like data mining and temporal reasoning. The tutorial gives enough basic event correlation background information in order for the audience to follow the issues of management of dynamic networks and services. A significant part of the tutorial will be devoted to different practical solutions and commercially available tools supporting the development of distributed event correlation applications.

The construction of an end-to-end OSS from individual commercially available off-the-shelf components has, over the last ten years, proven to be a much more elusive goal than most realized. There are numerous reasons for the failure of component-based software to deliver on a plug-and-play-then-unplug-and-play-again solution. Two of the key reasons for this failure have been lack of a platform independent software component model (along with an associated component execution environment) and the lack of simple standards or specifications in terms of an OSS API. The development and continuing evolution of the Javaª 2 Extended Edition Platform and work of the OSS through Javaª Initiative have begun to address both of these issues to enable the construction of component-based OSS solutions. The J2EE platform provides the component model and the component execution environment, to which the OSS through Java^TM Initiative is defining precise standards-based APIs for the OSS marketplace.

This tutorial will provide in-depth coverage of the technical challenges in building very large-scale distributed OSS solutions using J2EE and the OSS through Javaª APIs. Individuals attending this tutorial will gain an understanding of the OSS through Java^TM APIs in terms of design patterns and XML; as well as information on how to deploy and make use of the freely available OSS through Java^TM reference implementations, test compatibility kits, and API specifications.

This tutorial will be of interest to software architects, developers and project managers interested to know how the APIs defined by the OSS through Java Initiative can be used for large-scale OSS integration. Individuals interested in this tutorial that do not have an understanding of the J2EE platform are strongly encouraged to attend tutorial T-2.

Amongst the many initiatives of the TeleManagement Forum (http://www.tmforum.org), one of the most interesting and valuable for Information and Communications Service Providers (ICSP) is the definition and development of the "eTOM Business Process Framework". eTOM is a business process model or framework that describes all the enterprise processes required for a Service Provider, and analyzes these at different levels of detail according to their significance and priority for the business. For these companies it serves as the blueprint for process direction, and provides a neutral reference point for internal process reengineering needs, partnerships, alliances, and general working agreements with other providers. For suppliers, eTOM framework outlines potential boundaries of software components to align with the customers' needs, and highlights the required functions, inputs, and outputs that must be supported by products.

This tutorial is aimed at specialists across the information and communications industry working on operations and business automation, as well as process designers. Moreover, it is also suggested for designers and integrators of business and operational management systems software.

The last decade has seen the advent and significant publicity of component-based software development for telecommunications management systems (OMG, TeleManagement Forum, DMTF, WC3 etc.). This emergent trend is based on the premise that a management solution's developer should be easily able to connect pre-existing software components in order to construct applications to solve specific business problems. However, the difficulties experienced in component based software construction are significant. Even though the software industries regularly claim 'easy plug and play', 'visual component assemble' and automatic software integration, in reality component assembly and integration is only seldom successfully achieved from a heterogeneous collection of components. Recently there has been an emergence of Model Based Development, which attempts to insulate design concerns from technology choices and allows for a later binding to technology implementation. This tutorial presents a fusion of Model Based Development and Component Based Development that aims to overcome some of the problems involved in developing and using separately sourced software components for constructing management systems.

This tutorial will present state of the art issues in management component construction and integration, and will present a model-based approach to component development and management system construction. This approach utilizes a specialization of Rationale's Unified Development Process and supports integrated information models and component selection. Two implementation cases studies that illustrate key aspects of component development and provide both theoretical and practical insight as to the realization of component driven management systems will be presented.

The aim of the tutorial is to address the needs of Management Component Developers, Management Integration Solution Providers, Network & Service Operators and members of the standards community e.g. TeleManagement Forum etc. The intention is to facilitate the exchange of ideas between those working in the research community in component based management system construction, as well as those involved with operational and deployment issues.

This tutorial provides a comprehensive description of existing and emerging broadband wireless networks. The tutorial classifies and surveys wireless networks, explains the differences in their technologies and applications. The presentation covers broadband fixed wireless access, MAN and WAN wireless standards and infrastructure, 802.11 LANS, 802.16, Bluetooth, 3G and 2.5G wireless systems. To be discussed are the technological challenges that face these standards, their data rates, their mobility management models and their possible futures. Other topics will include the current activities within 3GPP and 3GPP2 and their current approach to define and standardize service profiles, QoS requirements and methods for enforcing the QoS in IP-based wireless access networks. The tutorial concludes by discussing future trends including the emerging interest in integrating broadband wireless access and metropolitan optical networks.

This tutorial is aimed at technical managers, hardware and software engineers and other networking specialists. It is designed to give a comprehensive overview of wireless concepts and techniques to individuals not specializing in the wireless field.

T-11: MPLS-based Virtual Private Networks: Design, QoS, and Management
Raouf Boutaba, University of Waterloo, Canada

This tutorial will provide an overview of the general principles underlying Virtual Private Networks (VPN) design, will show the limitations of traditional VPN solutions using Frame Relay, ATM, or IP tunnels, and will focus on how MPLS can be used for supporting VPNs. Also discussed will be issues relating to MPLS-based VPNs design which includes addressing, security, scalability, quality-of-service support and service management.

Driven by the increasing demand for network bandwidth (customer pull) and the phenomenal advances in WDM technology (technology push), the next generation Internet is envisioned to be an IP-based optical WDM network (Optical Internet), where the integration of gigabit/terabit IP routers with WDM switching and transmission systems creates optimized transport networks supporting the future growth of the Internet. From the network control and management perspective, the shift from a static point-to-point architecture towards dynamic re-configurable and switched architectures supports two important trends: i) a shift away from static planned resource allocation and service provisioning towards dynamic on-demand resource allocation and service provisioning, and ii) a shift away from centralized management and off-line optimization strategies towards distributed control and on-line incremental heuristics in network and traffic management. In this tutorial, we will describe architecture alternatives of Optical Internet and discuss their associated control and management issues and technologies.

Topics to be covered include: 1) IP and optical networking technologies: gigabit/terabit IP routers, optical cross-connects (OXC), 2) IP-over-optical networking architectures: network models, interconnection models-overlay/augmented/peering, service models, 3) Generalized MPLS (GMPLS or MP`Lambda'S): traffic engineering, routing and configuration management, link management, signaling, fault management and restoration, and 4) IP over switched WDM: optical burst switching and optical label switching.

This tutorial is targeted to those with knowledge of network management and broadband networking technology. The presentation will describe the basics of optical networks, architecture alternatives, GMPLS (MP`Lambda'S), traffic engineering, control and management issues of Optical Internet.

QoS technologies are entering the Internet domain. Among them are IntServ (RSVP), DiffServ and MPLS. This tutorial shortly introduces the basics of IntServ, DiffServ, and MPLS as the underlying QoS technologies. Scalability, manageability, and integration of these technologies are discussed. We believe that the key to getting QoS into the Internet lies in the control and management of QoS-enabled networks and transport services. Therefore, the tutorial concentrates on management issues, particularly for the combination of DiffServ and MPLS. Today we are facing several problems in this area, e.g., how do we build edge-to-edge or end-to-end guaranteed services?

The management technologies developed and being standardized in the IETF to provide solutions are mostly policy-based. An overview and discussion of the IETF management technologies such as SNMP, SNMPconf, COPS, and the Policy Framework to be used for QoS management are provided in this tutorial. Additionally, we present higher layer management architectures including Bandwidth Brokers/QoS Servers, Policy Servers, QoS routing, and Measurements. Finally, some issues and concepts for inter-domain QoS management will be addressed.

This tutorial requires basic knowledge of QoS and management technologies for IP. The first two thirds of the discussion is well suited for managers as well as for a technically interested audience. The last third which focuses on current research issues is more suited for technical experts in this area.

In the core and metro network, climbing Internet traffic growth rates are expected to find support from optical technology. The strength of optics is in its potential bandwidth of nearly 25 Terabits per fiber. The challenge, however, is to identify how this bandwidth can be made readily available to higher layer protocols, e.g., IP, providing the high speed network support required by bandwidth-greedy Internet applications.

This tutorial provides an overview of emerging optical technologies that are likely to become essential in the design of the Next Generation Internet, covering both the IP/MPLS and optical layer perspectives. Probable network architectures will be reviewed, including IP over Wavelength Division Multiplexed (WDM) optical signal. Issues and challenges of designing a cost effective high-speed network will be identified, presenting research proposals that address multiple access, protection and restoration, traffic grooming and TCP/IP internetworking. Short and long-term solutions, including testbeds from leading companies and consortia, will be reviewed, examining key components and anticipated advantages. CAD tools necessary to design the future Optical Internet will also be presented.

Researchers, students and practitioners in the field of Internet and optical networks that wish to acquire a broad knowledge of the enabling technologies and open challenges of the forthcoming Optical Internet will benefit from this tutorial.

Recent advances in the Java technology, Internet middleware, and IP telephony, along with a powerful traditional telecommunication services architecture gave rise to new potentials for the development of hybrid services that combine telecommunications and data services.

In this tutorial, we will review the intelligent network architecture of the traditional telecommunication network. We will focus on service creation using software components, service deployment, event triggering, service execution and management. An overview of IP telephony will be provided, focusing on signaling protocols and the JTAPI call model. The issue of integrating telecommunications and Internet technologies and building new hybrid services will be discussed. Then, two standard interfaces will be presented: JAIN and PARLAY. We will describe the JAIN architecture as well as the JCC and JSLEE APIs. Further, the Parlay architecture will be described along with various framework and service interfaces and its integration with JAIN.

This tutorial is intended for individuals who are interested in the use of the Java technology for the creation, management, and execution of hybrid, telecommunications and Internet, services. Service creators and application developers should benefit from new creation methodologies. Service providers should benefit from a technical understanding of the available standard technologies and trends. Equipment architects should benefit from modular, standard-based global service architecture.