Joberto S. B. Martins

Routing in DS-TE networks with an opportunistic bandwidth allocation model

DS-TE (DiffServ-aware MPLS-TE) networks support Quality of Service (QoS) implementation for traffic classes (TCs). Bandwidth Allocation Models (BAMs) are used in DS-TE networks in order to define Bandwidth Constraints (BCs) in a per-class basis and effectively define how this resource is reserved and eventually shared by applications. Existing BAM models are aware of traffic classes and are enablers for QoS guarantees while trying to optimize the use of bandwidth resources on a per-link basis. This paper proposes a new bandwidth allocation model (AllocTC-Sharing) which has an opportunistic behavior on a per-link basis. Opportunistic behavior means that AllocTC-Sharing allows higher priority traffic classes to use non-allocated resources of lower priority traffic classes and the other way around. The paper presents initially the AllocTC-Sharing and its integration with CSPF (Constrained Shortest Path First) routing protocol for end-to-end MPLS LSP (Label Switched Path) path setup. Then, we compare AllocTC-Sharing features with RDM (Russian Dolls Model) and Adapt-RDM models. The results show that AllocTC-Sharing preserves the features of RDM and Adapt-RDM and improves overall network utilization.

A RDM-like bandwidth management algorithm for Traffic Engineering with DiffServ and MPLS support

MPLS-TE (Multiprotocol Label Switching - Traffic Engineering) and DiffServ are technical alternatives used for supporting Traffic Engineering and QoS-demanding applications, such those typically found in multimedia multiservice networks. This paper proposes the ADAPT-RDM algorithm, suitable to allocation or not of LSPs (Label Switch Paths) according DiffServ-aware MPLS Traffic Engineering (DS-TE) and with the utilization of the RDM (Russian Doll Model) bandwidth constraint model. The principal objectives of the ADAPT-RDM algorithm are fundamentally the class types (CTs) priority preservation and the management of the bandwidth sharing capabilities of the RDM model. The paper describes the proposed algorithm and evaluates its results considering different implementation scenarios.

Experiences and Practices in Modeling Distance Learning Curricula for Capillary Approaches and Limited ICT Resource Scenarios

Actual curricula design and development for distance learning courses have been challenged by constraints such as the absence of conventional face-to-face (F2F) (Keeton 2002) class communication, the required student’s autonomy and the need for ergonometric interfaces, just to mention some. In general, both the extensive use and the focused application of ICT components (Felder & Silverman 1988) may provide some significant contributions to overcome these drawbacks. The objective of this paper is to describe a distance learning curricula design and implementation strategy considering a capillary approach with limited ICT resources. It highlights the approaches undertaken and the obtained results. The implementation scenario is an undergraduate level course for active high-school teachers requiring certification for improving their pedagogical methods. The numbers, methods and implementation aspects presented are based on a real-life ongoing course.

Adapt-RDM - A bandwidth management algorithm suitable for DiffServ services aware traffic engineering

MPLS-TE and DiffServ are technical alternatives used for supporting traffic engineering and QoS-demanding applications, such those typically found in multimedia multiservice networks. This paper proposes an algorithm based on LSP rate reduction and LSP preemption, the ADAPT-RDM algorithm, suitable to manage bandwidth sharing according with the RDM (Russian doll model) model. The constraints used by the ADAPT-RDM algorithm are fundamentally the class priority preservation and the RDM models basic constraints (bandwidth sharing by classes). The proposed algorithm is suitable for use in DS-TE context. The paper describes the proposed algorithm and evaluates its results considering different implementation scenarios.

A quality of service framework for tele-medicine applications

Research reported in this paper presents the design principles, the services and the proposed architecture for a QoS provisioning and management framework. An experimental implementation and the measurements evaluating the QoS provided using different DiffServ mechanisms are also discussed. The research reported here relates to infraVlDA development scenario, a telemedicine system that intends to allow remotely located health professionals to make telediagnosis and get second medical opinion from consultants on central hospitals.

A quality of service approach for managing tele-medicine multimedia applications requirements

Research reported in this paper presents the design principles, the services and the proposed architecture for a QoS provisioning and management framework. An experimental implementation and the measurements evaluating the QoS provided using different DiffServ mechanisms are also discussed. The research reported here relates to InfraVIDA development scenario, a tele-medicine system that intends to allow remotely located health professionals to make tele-diagnosis and get second medical opinion from consultants on central hospitals.

Network Autonomic Management: A Tutorial with Conceptual, Functional and Practical Issues

Computer networks are infrastructures largely used by our society supporting a broad range of applications. In effect, computer networks are diverse in terms of their scale, complexity, heterogeneity and application scenario (industrial, corporate, academic, telecommunications, sensors, domestic, mobile, smart grids, body networks, others). As such, network management became a challenging research issue mainly due to the fact that applications and services must be integrated with rather diverse requirements. The manager is another relevant and evolving aspect that has been considered by new OAM (Operation, Administration and Management) solutions. As a matter of fact, autonomic network management is a trend in the network management research area and the main focus is, in brief, to facilitate the network management task in a scenario of inherent complexity. As such, the autonomic management approach presents challenges that must be evaluated and understood. This tutorial explores the autonomic network management fundamentals which must be considered whenever any specific technical approach is conceived. In brief, the tutorial explores the state of the art, architecture, requirements and, in general, the strategic approaches that are commonly used in an autonomic management solution for networks in different utilization scenarios.

FIT@BR - a Future Internet Testbed in Brazil

A major objective of the Brazil-EU FIBRE project is the deployment in Brazil of FIT@BR, a wide-area network testbed to support user experimentation in the design and validation of new network architectures and applications. In such a testbed, a high degree of automated resource sharing between experimenters is required, and the testbed itself must be instrumented so that precise measurements and accounting of both user and facility resources may be carried out. In this article, we describe the design and implementation of the Control and Monitoring Framework (CMF) for the FIT@BR testbed, which is based on three CMFs developed in existing testbed projects. In order to take best advantage of different testbed functionalities at different sites, FIT@BR is being created as a federated testbed, which will facilitate future interoperation with international initiatives.

AllocTC-sharing: A new bandwidth allocation model for DS-TE networks

DiffServ-aware MPLS-TE (DS-TE) allows bandwidth reservation for Traffic Classes (TCs) in MPLS-based engineered networks and, as such, improves the basic MPLS-TE model. In DS-TE networks, per-Class quality of service guarantees are provided while being possible to achieve improved network utilization. DS-TE requires the use of a Bandwidth Allocation Model (BAM) that establishes the amount of bandwidth per-Class and any eventual sharing among them. This paper proposes a new bandwidth allocation model (AllocTC-Sharing) in which the higher priority traffic classes are allowed to use non allocated resources of lower priority traffic classes and vice versa. By adopting this “dual sense” allocation strategy for dynamic bandwidth allocation, it is shown that AllocTC-Sharing model preserves bandwidth constraints for traffic classes and improves overall link utilization.

A Policy-Based Autonomic Model Suitable for Quality of Service Management

The increased complexity resulting from computer networks technologies, devices and application diversity combined with the continuously growing service demands from users, requires a computer network that offers adequate treatment and support to multiservice (voice, video, web applications) and quality of service (QoS) requirements. This management scenario adds complexity to human administration even with the adoption of Policy- Based Management (PBM) systems. In effect, the effectiveness and efficiency of the management model is, typically, directly related with the knowledge and skills of the administrator in order to define a policy database that meets all possible states. In this paper is presented a model with autonomic capabilities that is able to manage quality of service in computer networks pushing towards an autonomic Full Policy-Based Management (FPBM) approach. The proposed model introduces a knowledge layer in a traditional Policy Based Management model so that policies are created in an autonomic way, replacing as such the human interaction in the process of defining the policy database.