Catálogo de publicaciones - libros

As universal high speed internet access becomes a reality, phone calls are increasingly being made over the Internet rather than the conventional PSTN. The danger to this trend is the un-availability of priority mechanisms for communication between emergency response personnel during times of disaster. We define a proposed architecture to enable ETS support for SIP-based VOIP systems.

In networks that support service differentiation, each in their own way, it is very important to manage how interactive multimedia applications use the networks’ enhanced services. Therefore, ways to ensure that these services are properly authorized and accounted for are needed. This paper addresses the role of Session Initiation Protocol (SIP) proxies to authorize QoS-enabled multimedia sessions, based on the session’s policy information and the network resources’ availability. Our inter-domain call authorization model provides call authorization status and adds more granularity to the authorization process. This model is implemented in a SIP testbed, and simulation results showed that the model is scalable at end domains.

The IP Multimedia Subsystem (IMS) is the latest framework for a seamless conversion of the ordinary Internet with mobile cellular systems. As such it has the backing of all major companies since it aims to offer a unified solution to integrated mobile services, including mechanisms for security, billing, quality of service and so forth. We provide a unique assessment of IMS based on our experimental test-bed, investigating functional and performance capabilities. Our study helps evaluating the level of maturity of state-of-the-art open source technologies in view of the deployment of IMS. We assess the suitability of SIP, IPv6, MIPv6 and IPsec as basic IMS enablers, highlighting crucial shortcomings which need immediate attention.

The need for providing applications with practical, manageable access to feature-rich capabilities of telecommunications networks has resulted in standardization of the OSA/Parlay APIs and more recently the Parlay X Web Services. Connectivity Manager is an existing, ‘stable’ API for establishing QoS parameters in a network. However, it falls short as an interface to the expected requirements based on recent drafts of ETSI TISPAN’s Resource Admission Control Subsystem. We analyze these requirements and suggest some improvements, which are incorporated into a new Parlay X Web Service specification. Furthermore, we describe our efforts to date in implementing a prototype of the specifications as well as our experience in utilizing the prototype to develop an example QoS-aware multimedia application.

Current OSA/PARLAY standard defines an architecture that enables service application located into an ASP to invoke network services capabilities through an open standardized interface. Nevertheless, architectures proposed within this standard work are isolated from other OSA/PARLAY infrastructures belonging to other network operators. In this research paper we deal with the invocation of network services located in visited networks from applications attached to a Home Network using OSA/PARLAY interfaces. We propose an extension to the OSA/PARLAY Framework signaling system so that it will work in a multi-OSA environment. This enhancement will require the definition of new interface classes among Frameworks in order to manage global service execution. The proposed method adds a new Framework-to-Framework interface that allows applications attached to a home network the access to service capability servers located in visited networks. In this enhanced OSA/PARLAY architecture, the applications may access remote services without integrating or dealing with any network negotiation.

In order to manage congestion problems and allocate network resources, many researchers have studied Internet pricing over the last decade. However, much of their research results have been limited by their reliance on the over-simplified demand model, and are not intended for adaptation to emerging multiple class environments such as the Diffserv network. For example, user utility is generally represented by a logarithmic form that is related to unit elasticity demand, but is not effective in representing user demand in the real Internet service market. We extend a dynamic pricing scheme by generalizing a demand model and applying it to the multi-class Diffserv network; and develop a simulation framework to compare the engineering and economic performance of our dynamic pricing model to those of static pricing.

Traditional telecommunications service providers are undergoing a transition to a shared infrastructure in which multiple services will be offered to customers. These services will be introduced, modified, and retired at a pace that tracks changing requirements and demands. In order to be cost-effective, these services will need to be delivered over a shared infrastructure that is managed to support delivery requirements at a given point in time. In this paper, we present an Autonomic Service Architecture (ASA) for the automated management of networking and computing resources. ASA ensures the delivery of services according to specific agreements between customers and service providers.

A mobile ad hoc network (MANET) is a temporary infrastructureless network, formed by a set of mobile hosts that dynamically establish their own network without relying on any central administration. Mobile hosts used in MANET have to ensure the services ensured by the powerful fixed infrastructure in traditional networks, the packet forwarding is one of these services.

Resource limitation of MANET’s nodes, particulary in energy supply, along with the multi-hop nature of these networks may cause a new problem that does not exist in traditional networks. To save its energy a node may behave , thus it misbehaves by not forwarding packets originated from other nodes, while using their resources to forward its own packets to remote recipients. Such a behavior hugely threatens the QoS (Quality of Service), and particulary the packet forwarding service availability. Some solutions for selfish nodes detection have been recently proposed, but almost all these solutions rely on the monitoring in the promiscuous mode technique of the watchdog [1], which suffers from many problems especially when using the power control technique. In this paper we propose a new approach to detect selfish nodes unwilling to participate in packet forwarding, that mitigates some watchdog’s problems. We also assess the performance of our solution by simulation.

TCP flows generated by applications such as the web or ftp require a minimum network throughput to satisfy users. To build this service, we propose a scheme with Admission Control (AC) using a small set of packet classes in a core-stateless network. At the ingress each flow packet is marked as one of the set of classes, and within the network, each class is assigned a different discarding priority. The AC method is based on edge-to-edge per-flow measurements, and it requires flows to be sent at a minimum rate. The scheme is able to provide different throughput to different flows and protection against non-responsive sources. We evaluate the scheme through simulation in several network topologies with different traffic loads consisting of TCP flows that carry files of varying sizes. In the simulation, TCP uses a new algorithm to keep the short-term sending rate above a minimum value. The results prove that the scheme guarantees the throughput to accepted flows and achieves high utilization of resources, similar to the ideal results of a classical hop-by-hop AC.

In this paper we present a policy-based architecture for aggregating (grooming) IP/MPLS flows (packet-based LSPs) within lightpaths taking into account the possibility of having to cope with further transport faults. The defined policies try to minimize the negative impact when a failure is detected in the optical transport network. Such policies deal with 1+1, 1:1 and 1:N schemes of protection. In our model, IP/MPLS flows are divided into High Priority (HP) and Low Priority (LP) traffics. The architecture is composed of an Admission Control responsible for receiving the requisitions from the IP/MPLS network and forward them to the Policy Manager which in turn is responsible for applying the policies. The architecture also has a Fault Manager responsible for accounting the failures and a Resource Manager responsible for managing the lightpaths. Our approach has been implemented to validate the policies and the results showed that the defined policies decrease the number of affected LSPs when a given lightpath fails.