Catálogo de publicaciones - libros

The issue on optimality and robustness has become a major concern in large-scale network systems. While a star-like centralized network surtcture is optimal in terms of the average path length, it is vulnerable to the breakdown arising in the central node. Scale-free network (SFN) is known to be effective topology in terms of both the average path length and robustness against random breakdown. However, if the hub nodes are intentionally attacked, SFN is found vulnerable. In this paper, we propose an evolutionary network model which reconfigures a network suructure according to the various types of breakdown or intentional attacks while maintaining the system performance. The local evaluation indice and control parameters are introduced to regulate a balance between efficiency and robustness. Simulation results suggest that the proposed approach is promising.

Self-organizing systems are present in many areas of nature and science, and have more recently been increasingly applied to telecommunications. These systems often exhibit common structural properties, such as the small-world property, and can react to changes in their environment with no centralized control. With ever-increasing capacity requirements, Transparent Optical Networks (TONs) have been established as the enabling technology for future long-haul high-speed backbone networks. Designing fast security mechanisms is critical, particularly due to the high speeds and transparency inherent in TONs. In this paper, we propose a self-organizing small-world control plane for failure management in TONs, which can improve scalability and adapt to changes in the network.

A system is said to be self-organizing if its execution yields temporal global structures out of simple and local interactions amongst its constituents (e.g agents, processes). In nature, one can find many natural systems that achieve organization at the global level without a reference to the status of the global organization; real examples include ants, bees, and bacteria. The future of tuple-space systems such as lies on their ability to handle non-trivial coordination constructs common in complex applications, and their scalability to environments where hundreds and maybe thousands of nodes exist. The Achilles heel of scalability in current tuple-space systems is tuple organization. Legacy solutions based on antiquated approaches such as hashing are (unfortunately) commonplace. This paper gets inspiration from self-organization to improve the status quo of tuple organization in tuple-space systems. We present a solution that organizes tuples in large networks while requiring virtually no global knowledge about the system.

For an efficient usage of the transmission capacity of a QoS-supporting Next Generation Network, it is beneficial to influence the routing of traffic flows by the optimization of link metrics. Deploying a Network Admission Control at the network border helps to comply with assured service guarantees as it can effectively protect against overload situations, especially in times of varying traffic matrices or failures.

Since the manual adaptation of link metrics and NAC budgets is neither quick nor efficient, it makes sense to integrate these optimization algorithms into a self-configuration tool, which is able to autonomously keep the network in the best possible operational condition. This paper presents a management system that re-optimizes link metrics and NAC budgets when necessary. Different scenarios show the benefits of this approach for an increased network resilience and efficient operation.

Recent routing/multicast protocols in large-scale mobile ad-hoc networks (MANETs) adopt two-tier infrastructures by selecting backbone hosts (BHs) in order to avoid the inefficiency of the flooding. Further, previous MANET quality-of-service (QoS) routing/multicasting protocols determined bandwidth-satisfied routes for QoS applications. However, they suffer from two bandwidth-violation problems. In this paper, a novel algorithm that can avoid the two problems is proposed and integrated with the two-tier infrastructures to construct bandwidth-satisfied multicast trees for QoS applications in large-scale MANETs.

In this paper, we investigate how to incorporate an application metric into the construction of a multicast tree so as to facilitate the use of range constrained multicast. We first describe the construction and delivery protocols, show through an analysis drawing on stochastic geometry that the protocol is scalable, and provide simulations showing the performance of the protocol against trees derived from reverse path forwarding construction.

We consider load sharing in distributed systems where a stream of service requests arrives at a collection of identical servers. The goal is to provide the service with the lowest possible average waiting time. This problem has been extensively studied before, but most previous models have not incorporated the monitoring costs explicitly. This paper focuses on a rigorous study of maximizing the utility of monitoring.

We extend the Supermarket Model for dynamic load sharing by explicitly incorporating monitoring costs. These costs stem from the fact that the servers have to answer load queries, a task which consumes both CPU and communication resources. This Extended Supermarket Model (ESM) allows us to formally study the tradeoff between the usefulness of monitoring information and the cost of obtaining it. In particular, we prove that for each service request rate, there exists an optimal number of servers that should be monitored to obtain minimal average waiting time.

Based on this theoretical analysis, we develop an autonomous load sharing scheme that adapts the number of monitored servers to the current load. We evaluate the performance of this scheme using extensive simulations. It turns out that in realistic scenarios, where monitoring costs are not negligible, the self-adaptive load balancing scheme is clearly superior to any load-oblivious load sharing mechanisms.

The number ofcommunication devices one uses is increasing day by day. Configuring these devices for optimal functioning, especially Mobile Ad-hoc Networks (MANETs), has not been an easy task for users. Self-configuration of the devices for optimal networking performance, being imperative, has been studied by various researchers considering the tasks at hand in various layers of the network stack. We have made a comprehensive study on self-configuration at different layers, and discussed the required interactions amongst them from different perspectives, which we think important. We break down the design complexity and identify system-level merits with the help of earlier studies. We further present our vision to provide such merits with self-configuration architecture. Our work presented here aims at understanding the potential challenges posed by the newer applications and services, and identifying research directions while providing the foundation for design of self-configurable systems.

A knowledge network is a generic structure that organises distributed knowledge into a system that will allow it to be efficiently retrieved. The primary features of this network are its lightweight autonomous framework. The framework allows for smaller components such as pervasive sensors to operate. Stigmergy is thus the preferred method to allow the network to self-organise and maintain itself. To be able to return knowledge, the network must be able to reason over its stored information. As part of the query process, links can be stigmergically created between related sources to allow for query optimisation. This has been proven to be an effective and lightweight way to optimise. These links may also contain useful information for providing knowledge. This paper considers a number of possibilities for using these links to return knowledge through a distributed lightweight reasoning engine, thus upholding the main features of the network.

With the emergence of mobile and ubiquitous computing environments, there is a requirement to enable collaborative applications between these environments. As many of these applications have been designed to operate in isolation, making them work together is often complicated by the semantic and ontological differences in the meta-data describing the data to be shared. Typical approaches to overcoming ontological differences require the presence of a third party administrator, an approach incompatible with autonomous systems. This paper presents an approach to automatic ontology mapping suitable for deployment in autonomous, interacting systems for a class of collaborative application. The approach facilitates the collaboration of application-level data collections by identifying areas of ontological conflict and using meta-data values associated with those collections to establish commonality. A music sharing application has been developed to facilitate the sharing of music between peers.