Recent years have seen the maturing of ubiquitous computing middleware and software. Accompanied by research into sensor networks and other sensor-driven applications, widespread deployment and realisation of these technologies can now be expected in the coming years. As a cheap and quick method of prototyping applications and protocols, simulation will be a key part of the development cycle for these technologies. However, existing simulators only address a subset of ubiquitous computing environments and are unsuitable for modelling the desired complexity of the domain.This paper presents initial work on the design of a generic simulation tool suitable for the many scenarios encompassed by ubiquitous computing, such as simulation of sensors, actuators, and the environment. In addition, an emulation framework for middleware and software under development is provided which interfaces with the simulation tool. We provide a layered, flexible and modular approach to supporting the simulation of ubiquitous computing environments without constraining the simulator to one aspect of the many possible ubiquitous computing deployment scenarios. Finally, we present and discuss a proof-of-concept simulation.

Several wireless sensor network applications are currently appearing in various domains. Their goal is often to monitor a geographical area. When a sensor detects a monitored event, it informs a sink node using alarm messages. The area surveillance application needs to react to such an event with a finite, bounded and known delay: these are real-time constraints. This work proposes a real-time MAC protocol with realistic assumptions for a random linear network, where sensors are deployed randomly along a line. We present a formal validation of this protocol both for initialization and run-time, and present simulation results on a realistic scenario.

We describe an approach for the formal modeling and analysis of ad-hoc sensor networks, at various levels of abstraction. It is global because it takes into account all the following aspects: a precise modeling of the hardware that implements a single node; the protocol layers; the application code; an abstract model of the physical environment as viewed by the sensors. The global model is executable, to enable validation by simulations, but we also aim at analyzing the global model with various formal validation tools (automatic test, runtime verification techniques, model-checking and abstract interpretations). Each technique or tool may need particular abstractions of the model. In this paper, we illustrate the whole approach with a simple model, and show what formal analysis can be performed on the model.

We present a methodology based on physics laws and particles in order to represent, simulate, and architect advanced networking models. We introduce a mathematical formalism with its basic postulates seeing the messages signals and nodes as interacting/colliding particles and space-temporal zones, respectively. In particular we focus on using this "Particle" view for Ad-Hoc networks, where we can represent network nodes as the moving particles in physics. This idea can be used in several types of networks like sensor or mobile, and even to computational systems, this first presentation aims to continue with further work on robustness, tools, and examples on how this view can be used in specification of complex networks in conjunction with existing methodologies and tools.

Sensor networks become more and more important in various application areas and several application scenarios require connectivity between sensor network and the Internet. Especially scenarios of disaster management would benefit from IP-enabled sensor networks, providing a seamless integration in WAN infrastructures without a requirement for deploying proxies that convert between communication technologies.Preparing sensor networks for IP communication and integrating them into the Internet demand for certain features and specification work, e.g. the adaptation of the respective link technology for IP support, development of security mechanisms, and specification of ad hoc networking and autoconfiguration to support ad hoc deployment. Moreover, in case sensor networks are moving from the IP point of view, which depends on the actual application scenario, mobility management is required. Especially interesting for sensor networks is the deployment of IPv6, providing a huge address space suitable to address large sensor networks globally, providing built-in autoconfiguration via IPv6 neighbor discovery and IPv6 stateless address autoconfiguration, and supporting network mobility (NEMO).We will motivate the work presented in this paper by describing promising applications and by discussing the benefits of IP-enabled sensor networks compared to non-IP sensor networks. Afterwards we investigate the required functions and protocols to provide such a system and give examples for protocols and function realizations. We will investigate in more detail how to provide autoconfiguration and mobility management. Finally, we will design a system and its functional blocks that aid in disaster management. The system consists of an IPv6-enabled sensor network connected to a central management station via different WAN technologies.

There is growing interest in the networked sensing community in the technique of macroprogramming, where the end-user can design a system using a high level description without worrying about the node-level details. Since the burden of customizing the code to the target architecture is moved to the compiler that translates the high-level description to generate node-level codes, research on the issues involved in compilation of such a program assumes importance. In this paper, we list some issues that need to be resolved by the designers of a compiler for such a macroprogramming framework, including the decisions to be made in the choice of an abstraction, the design of the runtime system and the generating of the code for each node. We discuss some solution techniques that we are currently exploring to solve the above problems.

Wireless sensor network (WSN) architectures often feature a (single) base station in charge of coordinating the application functionality. Although this assumption simplified the path to adoption of WSN technology, researchers are now being attracted by more decentralized architectures with multiple sinks and heterogeneous nodes. These scenarios are brought to an extreme in Wireless Sensor and Actor Networks (WSANs), where sensing and acting nodes collaborate in a decentralized fashion to implement complex control loops. In these settings, new programming abstractions are required to manage complexity and heterogeneity without sacrificing efficiency.In this paper we propose and define a logical neighborhood programming abstraction. A logical neighborhood includes nearby nodes that satisfy predicates over their static (e.g., type) or dynamic (e.g., sensed values) characteristics. The span of the neighborhood and the definition of its predicates are specified declaratively, along with requirements about the cost of the communication involved. Logical neighborhoods enable the programmer to "illuminate" different areas of the network according to the application needs, effectively replacing the physical neighborhood provided by wireless broadcast with a higher-level, application-defined notion of proximity.This paper presents the definition of a declarative language for specifying logical neighborhoods, highlighting its expressiveness, flexibility and simplicity. Moreover, although the language constructs are readily implemented using existing communication mechanisms, we briefly report about a novel routing scheme we expressly designed to support efficiently our abstractions.

Dynamic wireless networks, and especially mobile ad hoc networks, have imposed new constraints regarding the design and the implementation of service-oriented middleware platforms dedicated to mobile computing. Indeed, from now, these platforms must be able to capture the context in which they operate in order to provide the services they host with an abstraction of their running context, and to notify them of the variations occurring in this context. These middleware platforms must also implement communication means enabling services to support the frequent and unpredictable connection disruptions. In this paper, we present the design and the implementation of a service-oriented middleware platform capable of environmental context introspection and of delay-tolerant asynchronous communications.

In this paper, we describe the middleware that has evolved from our attempt to develop location-aware and user-aware services in ubiquitous environments. User mobility can cause numerous problems and limitations: user-preferred applications are not installed in the new location or device, user-specific configurations are needed, there are different OS and platforms, limited resources in the new node and so on. These restrictions sometimes prevent the user from performing what is required, anywhere and anytime. We propose a system in which, by enabling code installation and execution on remote nodes, these limitations are controlled. A prototype implementation of this approach has been constructed and we present the rationale, design, implementation and applications of the prototype system.

Sensor networks have received wide attention in recent years for their revolutionary impact in numerous fields. To harness their full potential, researchers are beginning to build end-to-end solutions that integrate heterogeneous sensor deployments with traditional networks. While such efforts will bring true value to the use of sensor networks, there are several challenges that need to be kept in perspective. This paper highlights some of the key challenges, and presents a system software methodology to meet the difficult demands of development, deployment and long term management of sensor network systems.

The project MarathonNet develops wireless sensor networks for monitoring runners during marathon events. The application requires a high degree of connectivity in order to provide actual data for runners and spectators. Depending on the distribution of the runners on the track, the communication range and the number of base stations network partitions might occur that reduce connectivity. To investigate these dependencies in detail we conducted various simulations on connectivity. In this paper we first introduce the application scenario and discuss the impact of the different parameters and their interrelations. We then present the simulation results and discuss their consequences for the application design.

Optimizing energy consumption in wireless sensor networks is of paramount importance. There is a recent trend to deal with this problem by introducing mobile elements (sensors or sink nodes) in the network. The majority of these approaches assume time-driven scenarios and/or single-hop communication between participating nodes. However, there are several real-life applications for which an event-based and multi-hop operation is more appropriate. In this paper we propose to adaptively move the sink node inside the covered region, according to the evolution of current events, so as to minimize the energy consumption incurred by the multi-hop transmission of the event-related data. Both analytical and simulation results are given for two optimization strategies: minimizing the overall energy consumption, and minimizing the maximum load on a specific sensor respectively. We show that by adaptively moving the sink, significant power saving can be achieved, prolonging the lifetime of the network.

Mobility is one of the most invigorating features, having an enormous impact on how communication is evolving into the future. Mobility in 4G networks requires new level of mobility support as compared to traditional mobility. There is plenty of related research on mobility in next generation networks, which promises support for emerging ambient and ubiquitous communications. This paper aims to identify and explore the different issues and challenges related to mobility management in 4G heterogeneous networks. A review of the existing solutions and ongoing projects related to these mobility management issues is presented; which can help in discovering a unified approach to seamless mobility in future generation networks.

We propose a multichannel wireless network architecture that supports nodes with differing multichannel capabilities. Both contention and contention-free medium access control (MAC) protocols are proposed for this heterogeneous multichannel network. In the contention MAC protocol, a new LRN control packet is defined that allows a pair of nodes to re-negotiate channel selection. This relaxes the assumption that nodes can sense the carrier on all supported channels in a short period of time. We model the scheduling problem for our contention-free MAC protocol as a weighted bipartite matching problem that minimizes the sum of the completion times with polynomial time complexity. We close with an illustration showing how the weighted bipartite matching generates a transmission schedule.

In this paper, a QoS routing protocol is proposed for a multihop wireless network with infrastructure support. Such networks will benefit existing applications in WLAN as the network coverage is greatly extended without incurring additional cost. The network resources can also be more efficiently utilized as load-balance can be achieved among neighboring access points (APs) through multihop connections. Neighboring APs usually operate at orthogonal channels to reduce interference and improve channel utilization. The channel diversity will impact the route choosing. In this paper, a load estimation metric is proposed to effectively estimate the load at AP. Unlike previous load-balance algorithms, mobile host (MH) will connect to AP with more residual bandwidth. MHs will also monitor the QoS to react quickly to the dynamics of the wireless networks. Upon detected QoS violation at the MH or AP, the MHs will perform channel switching and connect to other APs so that congestion is alleviated. Through simulation study, our proposal is proven to be able to efficiently utilize the network resources even when the network traffic changes very fast.

Mobile Ad-hoc Networks (MANETs) are seen as an essential technology to support future Pervasive Computing Scenarios and 4G networks. In a MANET, efficient support of multipoint communications is essential in order to provide services like group audio and video conferencing, dissemination of data to a set of receivers or collaboration of a group of users. However, most of these interactive services have very strong requirements regarding delay and bandwidth. Controlling the end-to-end delay and maintaining low packet loss rate is vital to support interactive multimedia applications in MANETs. In this paper, we present QAMNet, an approach to improve the Quality of Service (QoS) for multicast communication in MANETs. We extend existing mesh based multicast routing protocols by introducing traffic prioritization, distributed resource probing and admission control mechanisms and adaptive rate control of non-real-time traffic based on Medium Access Control (MAC) layer feedback to maintain low delay and required throughput for real-time multicast flows. Simulation results show that, by reusing control messages already used by those protocols, our approach does not significantly increase control overhead nor state stored at nodes compared while it manages to increase packet delivery ratio and reduce latency for real-time traffic.

Location-based routing (LBR) is one of the most widely used routing strategies in large-scale wireless sensor networks. With LBR, small, cheap and resource-constrained nodes can perform the routing function without the need of complex computations and large amounts of memory space. Further, nodes do not need to send energy consuming periodic advertisements because routing tables, in the traditional sense, are not needed. One important assumption made by most LBR protocols is the availability of a location service or mechanism to find other nodes' positions. Although several mechanisms exist, most of them rely on some sort of flooding procedure unsuitable for large-scale wireless sensor networks, especially with multiple and moving sinks and sources. In this paper, we introduce the Anchor Location Service (ALS) protocol, a grid-based protocol that provides sink location information in a scalable and efficient manner and therefore supports location-based routing in large-scale wireless sensor networks. The location service is evaluated mathematically and by simulations and also compared with a well-known grid-based routing protocol. Our results demonstrate that ALS not only provides an efficient and scalable location service but also reduces the message overhead and the state complexity in scenarios with multiple and moving sinks and sources, which are not usually included in the literature.

Wireless sensor networks are expected to be deployed in harsh environments characterized by extremely poor and fluctuating channel conditions. With the generally adopted single-sink architecture, be it static or mobile, such conditions arise due to contention near the sink as a result of multipath data delivery. The compactness of sensors with limited energy resources restricts the use of sophisticated FEC or ARQ mechanisms to improve the reliability of transmissions under such adverse conditions.We propose a novel virtual sink architecture for wireless sensor networks that mitigates the near-sink contention by defining a group of spatially diverse physical sinks. Reliability and energy efficiency is achieved through multipath data delivery to the sinks without the need for sophisticated FEC or ARQ mechanisms. This architecture is especially suitable for indoor environments, where channel conditions are harsh due to severe multipath fading, as well as emerging applications like underwater sensor networks where the predominant physical layer is acoustic communications, which is characterized by long propagation delays and severely fluctuating link conditions. We present our proposed architecture and demonstrate its efficacy using mathematical analysis.

Pervasive computing applications typically involve rich interactions and heterogeneous network types; e.g. involving the collation of data from a sensor network into a replicated repository in a fixed network. Although the middleware approach has been highly successful in supporting application development in networked environments, current middleware technologies cannot handle the accelerating complexity in interaction types, and diversity in networks types, seen in pervasive computing environments. Therefore, we propose a middleware solution (called Gridkit), which uniformly supports an extensible set of middleware interaction types (e.g. RPC, publish-subscribe, streaming, etc.), and handles network heterogeneity by layering itself over virtual overlay networks which it manages and transparently instantiates on demand. We focus in this paper on Gridkit's generalized architecture for the transparent deployment and management of overlay networks. We also consider the application of the Gridkit approach in two application scenarios.

Routing in Ad hoc Networks has received a significant amount of attention. In recent years, the focus of research has been in on-demand (or reactive) routing protocols due to the recognition that these protocols have the potential to achieve higher levels of scalability than proactive routing strategies. However, most on-demand routing protocols proposed to date attempt to increase routing efficiency by using existing knowledge about the destination or by increasing the stability of the routes. Little research has been done to reduce route discovery overhead when no previous destination information is available. We present a number of different strategies, which encourage a more distributed and localised approach to route discovery by allowing each intermediate node during route discovery to make forwarding decisions using localised knowledge and self-selection. The use of self-selection for route discovery enables nodes to independently make route request (RREQ) forwarding decisions based upon a selection criterion or by satisfying certain conditions. The nodes which do not satisfy the selection criterion do not rebroadcast the RREQs. This provides a more effective and efficient search strategy than the use of traditional brute force blind flooding. We implemented our self-selecting route discovery strategies over AODV using the GloMoSim network simulation package, and compared the performance with existing routing protocols. Our simulation results show that a significant drop in the number of control packets can be achieved by giving each intermediate node more authority for self-selection during route discovery. Furthermore, a significant increase in routing performance is achieved as the number of nodes in the network is increased.

Wireless Sensor Networks (WSNs) enabled with authenticated querying [1] allows only the authorized users to inject queries into the network and get responses thereof. Queries from unauthorized users are silently dropped. Providing authenticated querying for WSN is especially very challenging since the nodes are very much resource constrained and vulnerable to physical captures by the adversary. We present a fully symmetric key based solution for the above problem and analyze the same.

In this paper, we consider the sensor replacement problem in hybrid wireless sensor networks composed of mobile and static sensors. Mobility equipped mobile sensors are utilized to recover or to improve the overall coverage and connectivity. Unlike existing solution making assumptions of precise location information, our fault repair solution does not assume the localization information is available. Faulty sensors try to find available redundant sensors and ask for replacements. Mobile sensor nodes make use of simple geometric operation to locate and replace dying nodes to recover or increase the existing coverage and connectivity. Simulation results show that our proposed sensor replacement protocol successfully manages fault repair in absence of location information and is energy efficient.

Cooperation of nodes is a main issue of civilian applications of mobile ad hoc networks. More precisely, mobile nodes with few resources may try to maximize the benefits they get from the network without participating to its services (e.g. without forwarding packets of other nodes). A solution to this problem is to use virtual currency mechanisms: nodes have to pay to send packets and they are rewarded when they forward packets of other nodes. Several papers of the literature have shown that such models enforce individual node cooperation. We address in this paper two main issues of these models that have been left out: the computation by the sender of the price it will be charged and the possibility to reward nodes depending on their charge or resources. We show that it leads to a better load balancing of the network.

Motivated by the current trends in wireless technologies, we present challenging scenarios consisting of moving and uncoordinated wireless objects as well as algorithms that are likely to perform well in such environments. Our primary interest is to design network protocols to efficiently build and maintain distributed communication systems based on opportunistic design principles. Such systems should continue to operate well even if communication is sporadic, no end-to-end routes are available, and sources and consumers of information are not known in advance. Also the environment poses many challenges, that are typical for wireless mobile devices, such as wireless channel impairments, the impossibility of having up-to-date and accurate views of the network topology at every device, and limited energy resources. In the second part of the paper, we discuss and propose solutions for data dissemination in such an environment by exploiting network coding techniques. Furthermore, we report first results which reveal that network coding largely outperforms standard solutions which are based on the classical store and forward paradigm.

Wireless Mesh Networks (WMNs) have emerged as one of the new hot topics in wireless communications. They consist of a number of static wireless routers which form an access network for end users to IP-based services. Unlike traditional WLAN deployments, wireless mesh networks provide multihop routing, facilitating an easy and cost-effective deployment. In this paper, we focus in the provision of efficient multicast routing on such wireless mesh networks, and its seamless interconnection to wired IP multicast services. The proposed architecture and protocols allow for a ready deployment without changes on the existing networking equipments in the fixed network or in end user's devices. Our simulations and emphirical evaluation on a real wireless mesh network demonstrate that the proposed approach offers a good performance and is highly scalable.