Catálogo de publicaciones - libros

We consider wireless multihop data networks with random multi-access mechanisms at the MAC layer. In general, our aim is to study the performance as perceived by users in a dynamic setting where data flows are generated randomly by users and cease upon completion. This task comprises two major difficulties: first, the behavior of random multi-access algorithms at slot-level in a multi-hop network is even more complex than in the case of a single hop hotspot. Second, in order to study user-level performance accounting for a dynamic population of flows, one has to first characterize the so-called rate region when the population is fixed. The rate region is defined by the set of rates at which the various active users can generate packets without inducing any instabilities in the network. Since links interact with each other through interference, characterizing the rate region is as difficult as studying the behavior of a set of interacting queues. In addition, the behavior of the congestion control algorithm must be taken into account since it impacts the set of active links and thus the interference. We propose a model, based on the so-called mean field approach, that circumvents both difficulties and allows the derivation of explicit expressions for the rate region.

Fractal or scaling phenomena, such as self-similarity and long-range dependence, have been detected in network traffic, with important implications for network performance. This paper describes how the characteristics of traffic change when it traverses IEEE 802.11 networks. First, a simple, unidirectional traffic scenario with two nodes has been simulated. This scenario and two more general situations that include the hidden node situation and the effect of relaying (which appears in multihop communications) have been replicated in a real testbed. The most significant results are the smoothing effect of the MAC mechanisms on the traffic at the highest frequencies, while at the lowest frequencies a mitigation of fractality seems to be caused by packet loss induced by propagation impairments, with an overall effect of mitigation of the scaling characteristics on the output traffic. On the other hand, the scenarios with traffic aggregation show an increase of fractality.

Vehicular ad-hoc networks (VANETs) are promising means for distributed automotive applications. In addition to developing new communication protocols for VANETs, the widespread use of general purpose wireless network standards like WLAN IEEE 802.11a/b/g raises the need to investigate whether these networks are suitable for VANETs. The results from such experimental studies allow to introduce better simulation models and to evaluate the appropriateness of analytical models for wireless networks.

This work proposes a structured, scenario-based approach for measuring wireless networks in motion, which can easily be exploited by simulations. By applying the approach to IEEE 802.11g ad-hoc networks, experiments and results are described. The signal strength has been investigated for all mobility scenarios and is presented for the most significant cases as well as compared to the free space loss formula of electromagnetic waves. Furthermore, the throughput and packet loss are studied for traffic bursts while moving.

Information Technologies are called to play an important role in the world of transport services. One example is the combination between mobile Internet and automotive industry. This research is focused on the design of a new automotive network architecture called TrafficNet which, using IEEE 802.3 and 802.11 technologies, intends to support not only wireless communications to any type of vehicle driving along a circulatory route but also traffic management and road security services to the people in charge of the circulatory route. Results show that wired and wireless LAN technologies, together with new automotion-related protocols for optimization, can be used for high mobility applications in this type of network architecture.

Nowadays, some of the more exciting research areas in networking are based on cooperation between network nodes. Examples of this are ad-hoc or sensor networks, cooperative physical layer techniques known as , or other new cooperative mechanisms such asIn many scenarios, this cooperation could be further exploited using new mechanisms that fall into the L2-L3 protocols, leading to a . These mechanisms allow an easy use of the resources of adjacent nodes to increase communication capabilities. This combines particularly well with the characteristics of wireless networks thanks to the broadcast advantage inherent in wireless transmission. This paper proposes a novel cooperative relaying scheme that exploits transmitter diversity and performs a fast path repair procedure at L2. All these operations are made transparently to the content of the forwarding tables of the nodes, thanks to the use of a new addressing scheme. This protocol could work together with other cooperative protocols such as Cooperative ARQ, leading to an integrated mechanism for frame relaying. Analytical and simulation results show that cooperative frame relaying clearly improves network resilience.

Wireless ad hoc networks have the last 10 years gained a lot of attention within the research community. A wireless ad hoc network is a special type of wireless network where nodes are typically mobile and doesn’t rely on any fixed infrastructure to operate correctly. Much work has been done on developing robust and stable routing algorithms that consider the dynamic nature of an ad hoc network: nodes are very mobile and enter and leave the network in random ways. Much work has also been done on developing medium access control (MAC) algorithms that consider many wireless characteristics such as interference and the hidden terminal problems. Little work has so far been done on developing joint routing and MAC layer solutions that consider the variations of the wireless channel. We present a solution where the routing protocol can cooperate with the MAC layer to provide power control and channel dependent forwarding. By establishing non-disjoint multiple paths between each source and destination, nodes may be able to avoid links that are currently in a deep signal fade by choosing a more beneficial next hop path.

This paper presents a novel MAC protocol for Wireless Sensor Networks (WSN)s designated Tone-Propagated MAC (TP-MAC). This protocol is specially suited for early warning and tracking applications, where sensor nodes generate sporadic asynchronous traffic (mainly consisting of uplink alert messages and downlink control messages) with stringent latency requirements. This protocol aims to maximize energy-efficiency while minimizing latency in source-to-sink and sink-to-source communication. This difficult objective is achieved integrating scheduled channel polling (i.e. synchronized low power listening) with rapid fast path establishment based on the propagation of short wake-up tones. An analytical model was used to compare TP-MAC with SCP-MAC. The results show that TP-MAC is able to achieve better target latencies even when its duty-cycle is lower during periods of inactivity. The results also show that the advantage of using TP-MAC increases with the hop-distance between source and sink.

Achieving QoS (Quality of Service) in Mobile Ad-hoc NETworks (MANET) has been a research topic in the last years. In this paper we describe an adaptive QoS reservation mechanism for Multirate Ad-hoc Networks. The mechanism is targeted for sources requiring a bandwidth allocation. The scheme consist of the nodes initially computing the available bandwidth for new reservations, using the state information advertises among the neighbors. Then, the nodes adjust the available bandwidth by measuring the effective transmission rate achieved while transmitting packets. Simulation results show that the available bandwidth computed by the proposed scheme adapts fast and accurately to changing network conditions.

In this paper we study the expected difference between the QoS classes in an ad hoc network. The results have a direct bearing on the suitability of extending a fixed DiffServ architecture into an ad hoc network. Through simulation, we analyze the number of classes that can be used in the ad hoc network with separation between the observed QoS in the different classes. The results clearly depend on the type of traffic run in the network. With well behaved CBR traffic, the ad hoc network supports no more than four classes, but with more aggressive traffic like TCP no more than two classes are supported. In addition, there is a fairness problem; the performance for a particular flow is not well distributed among the nodes.

Most of the current topological based routing algorithms used in MANETs treat all the nodes in the network like independent peers, making them not scalable with respect to the number of nodes in the network. Since hierarchical routing techniques have been known to afford scalability in large networks (e.g. Internet), and taking advantage of the existence of scenarios in which the nodes of a MANET can be aggregated in a natural manner, we propose a two level hierarchical routing scheme for MANETs. In this paper we present the main components and an analytical performance evaluation of our proposal, with the number of control packets per second as the metric of our interest. The evaluation shows a significant overhead reduction from Θ(() to Θ((). However, a trade-off between the # of nodes in the network and the complexity of the system has to be achieved.