MiNEMA Winter School 2009

March 23-26, 2009
Göteborg, Sweden


Session A
9h - 10h30
[09h00] Registration
[10h00] Welcome
Ad Hoc Routing Gossip-based Dissemination Tuple Spaces
Session B
11h - 12h30
Energy in WSN Opportunistic Networks Publish/Subscribe Vehicular Networks
Session C
14h - 15h30
Programming WSNs Population Protocols Application Layer Multicast (cancelled) Context-aware Applications
Session D
16h - 17h30
Mobility Models Wireless Mesh Networks SC meeting Dynamic Adaptation


Energy in Wireless Sensor Networks

Wireless sensor networks have the potential to be applied in different applications such as habitat, environmental and industrial monitoring. Each sensor node is responsible for performing sensing, processing and communication functions that depend on a power source, which is typically represented by a battery that has a limited energy capacity. Energy awareness is a very important design consideration for protocols and algorithms in sensor networks. Energy management in WSNs involves not only reducing the energy consumption of a single sensor node but also maximizing the lifetime of the entire network. Furthermore, energy awareness should be incorporated into every stage of the wireless sensor network design and operation with the goal of making dynamic tradeoffs between energy consumption, system performance, and operational fidelity. In this tutorial it is discussed the main energy techniques that can be applied to the design of applications, algorithms and protocols for wireless sensor networks. The discussion is based on a taxonomy of techniques used by algorithms and protocols to save energy. (slides)

Programming Wireless Sensor Networks

Robust programming of large-scale sensor networks is a difficult problem. The severe restrictions imposed by the underlying hardware platform and the choice of an appropriate set of abstractions to model the processes occurring in these networks makes the task quite formidable. Despite these difficulties, the potential pay-off is immense with applications in many areas from mundane (domotics) to state-of-the-art science (high energy physics). It is therefore not surprising that a considerable amount of research has been devoted to the subject and that some solutions have emerged with various degrees of success. This tutorial presents a survey of the current state of the art in programming languages and runtime systems for wireless sensor networks, while trying to identify problems not addressed adequately by the current body of research and, tendencies for future developments. (slides)

Mobility Models for Systems Evaluation

Mobility models play an essential role in the evaluation of protocols and algorithms for mobile systems, especially for the simulation of large-scale scenarios. In this tutorial we present a survey of the state-of-the-art of mobility modelling for networking and systems research. We discuss the most important synthetic models, considering both purely random mobility models and those based on real traces. We also present the most important analytical results about human connectivity derived from the analysis of real traces. Finally, we discuss recent developments related to the application of social network theory to mobility modelling and outline the existing research challenges. (slides)

Ad Hoc Routing

MANETs topology may change frequently as nodes move, thus turning multi-hop routing into a particularly difficult problem. In this tutorial, we focus on topology-based, and position-based routing schemes. In topology-based approaches, nodes exchange connectivity information with their neighbors to create routing tables. First topology-based routing protocols, were directly derived from distance vector IP protocols. Unfortunately, they are not suitable for wireless environments, because they waste precious resources collecting topological information for destinations that might never be contacted. To address this problem, newer protocols postpone collection of routes up to the moment when nodes need them. One shortcoming of topology-based routing is that nodes have to collect routing information that might be arbitrarily distant. As this might be impossible in a wireless network, position-based routing emerged as a less expensive alternative. By using location of the destination, of the node itself and of its neighbors, it is possible to reach destination with only a very limited view of the network, often restricted to the immediate one-hop neighborhood. We divide the presentation of position-based routing schemes in two major parts, as both play equally important roles: routing algorithms and pre-processing algorithms. The routing algorithm runs at each node and determines which neighbor should be the next hop for a packet. The pre-processing algorithm serves to create a graph for the routing algorithm. This usually means to create in each node a local view of a global planar graph. Unfortunately, simplicity of position-based routing comes with some costs, namely in the technology required to determine position and in routing failures, even in extremely simple scenarios. Hence, both approaches, topology and position-based, have their own advantages and disadvantages. (slides, 45MB!)

An Introduction to Population Protocols

Population protocols are used as a theoretical model for a collection (or population) of tiny mobile agents that interact with one another to carry out a computation. The agents are identically programmed finite state machines. Input values are initially distributed to the agents, and pairs of agents can exchange state information with other agents when they are close together. The movement pattern of the agents is unpredictable, but subject to some fairness constraints, and computations must eventually converge to the correct output value in any execution that results from that movement. This framework can be used to model mobile ad hoc networks of tiny devices or collections of molecules undergoing chemical reactions. This tutorial surveys results that describe what can be computed in various versions of the population protocol model. (slides)

Routing Issues in Opportunistic Networks

Opportunistic networking constitutes a medium-term application of general-purpose MANET for providing connectivity opportunities to pervasive devices when no direct access to the Internet is available. Pervasive devices, equipped with different wireless networking technologies, are frequently out of range from a network but are in the range of other networked devices, and sometime cross areas where some type of connectivity is available (e.g. Wi-Fi hotspots). Thus, they can opportunistically exploit their mobility and contacts for data delivery. Opportunistic networksthus aim at building networks out of mobile devices carried by people, possibly without relying on any pre-existing infrastructure. Moreover, opportunistic networks look at mobility, disconnections, partitions, etc. as "features" of the networks rather than exceptions. Actually, mobility is exploited as a way to bridge disconnected "clouds" of nodes and enable communication, rather than a drawback to be dealt with. More specifically, in opportunistic networking no assumption is made on the existence of a complete path between two nodes wishing to communicate. Source and destination nodes might never be connected to the same network, at the same time. Nevertheless, opportunistic networking techniques allow such nodes to exchange messages. By exploiting the "store-carry-and-forward" paradigm, intermediate nodes (between source and destination) store messages when no forwarding opportunity towards the final destination exists, and exploit any future contact opportunity with other mobile devices to bring the messages closer and closer to the destination. This approach to build self-organising infrastructure-less wireless networks turns out to be much more practical than the conventional MANET paradigm. (slides)

Wireless Mesh Networks

Wireless Mesh Networks have emerged as an important technology in building next-generation networks. They are seen to have a range of benefits over traditional wired and wireless networks including low deployment costs, high scalability and resiliency to faults. Moreover, Wireless Mesh Networks (WMNs) are often described as being autonomic with self-* (healing and configuration) properties and their popularity has grown both as a research platform and as a commercially exploitable technology. Initially this tutorial examines the challenges faced by traditional network technologies and discusses the role of WMNs in overcoming some of these problems. Following this an overview of Wireless Mesh Networks is presented, providing comparisons with similar network technologies including sensor networks and Mobile Ad Hoc Networks (MANETs). The core of the tutorial then details the state-of-the-art technologies involved in the construction of WMNs, including standards based activities, academic research, commercial products and deployment testbeds. (slides)

Gossip-Based Dissemination

This tutorial addresses the use of gossip-based protocols in mobile ad hoc networks. Gossip-based protocols have the advantage of requiring little or no structure to operate, making them particularly appealing to apply in dynamic systems such as wireless self-organizing networks. The use of gossip-based protocols is illustrated with three different, but related, middleware services, namely: broadcast, publish/subscribe, and data placement. (slides)

Application Layer Multicast

Application Layer Multicast (ALM) has increasingly gained interest, in particular because this approach makes it possible to overcome the shortcomings of IP multicast. Along that line, several approaches have been proposed, each focusing on a different set of criteria, such as reliability, scalability, control overhead, etc. The set criteria of an ALM solution then largely depends on its target application requirements. To provide a better understanding of key research results in this field, this presentation proposes a classification a large number of ALM solutions proposed in the literature. In doing so, we try to position each solution via three viewpoints: (1) quality of service, (2) underlying architecture and (3) applicative focus. Since many interesting ALM solutions were not explicitly designed for MANETs, we also try to assess to what extend these solutions may be suitable for MANETs deployment, by defining a set of criteria specific to deployments in MANETs.

Distributed Event Routing in Publish/Subscribe Systems

Distributed event routing has emerged as a key technology for achieving scalable information dissemination. In particular it has been used as preferential communication backbone within publish/subscribe communication system. Its aim is to reduce the network and computational overhead per event diffusion to a set of interested recipients. This tutorial introduces the reader to modern publish/subscribe systems through an overview on current techniques for event dissemination. The approach we follow proposes a decomposition of these architectures in functional layers. We survey current algorithms for event based routing, and possible overlay infrastructures in wired and mobile systems. (slides)

Tuple Space Middleware for Wireless Networks

Tuple spaces are a communication and coordination model based on data sharing. Applications based on this model are highly decoupled, a fundamental asset in the dynamic wireless setting. This tutorial looks back at almost a decade of efforts in the research community, by concisely describing some of the most representative tuple space systems and analyzing them along some fundamental dimensions of comparison. In doing so, it considers two main classes of applications that rely on wireless communication: mobile computing and wireless sensor networks. Some case studies are presented, borrowed from the instructor's own work on the Lime middleware and its adaptation to wireless sensor networks, TeenyLime. (slides)

Dynamic Adaptation

The fundamental characteristic of mobile entities connected via wireless networks (either fixed infrastructure or ad hoc) is change in environmental conditions. Dynamic adaptation is an essential technique to allow mobile systems to ensure they continually provide the required levels of service in the face of change. This tutorial investigates the software techniques for performing adaptation, and the adaptive middleware technologies that have been used to develop dynamic mobile applications. Finally, the future challenges in this emerging field of research are investigated. (slides)

Middleware Support for Context-Aware Applications

With computing devices becoming more mobile and pervasive, a stronger interaction between an application and its changing environment opens new horizons in terms of application functionalities. Location-based applications, such as GPSnavigation systems, are good examples of how information provided to an application on its surroundings offers new kinds of functionalities. Location is one of many environmental variables that might influence the behavior of an application. The notion of context encompasses these variables in the broad sense.

Vehicular Networks and Applications

Vehicular networks represent a particularly challenging class of mobile (ad hoc) networks that enable vehicles to communicate with each other and/or with roadside infrastructure. This tutorial describes potential applications, middleware approaches and communication protocols proposed for vehicular networks.

ESF Logo Chalmers Logo INESC-ID Logo DCS Logo


Design: José Mocito
Content: Georgios Georgiadis