This paper proposes a mathematical framework to model and analyze the energy consumption of routing protocols in multihop wireless networks such Mobile Ad Hoc Networks (MANETs) or Sensor Networks (WSNs) by taking into account the protocol parameters, the traffic pattern and the network characteristics defined by the medium channel properties, the dynamic topology behavior, the network diameter and the nodes density. Moreover, our framework allows to classify routing protocols by exploring the relationship between application requirements and wireless network features. To the best of our knowledge, there is no such proposed framework that provides in-depth an analytical investigation of above relationship. We apply this model to a Beacon-less routing protocol Pizza-Forwarding (PF) compared to well known routing protocols such as Optimized Link State Routing, Ad Hoc On Demand Distance Vector Routing, Greedy Perimeter Stateless Routing and a Simple Random Walk Routing in order to underline their behavior according to the network features and application requirements. Finally, we show that Beacon-less routing protocol should be a best candidate to save energy in WSNs.

This paper presents a description of SensLAB(Very Large Scale Open Wireless Sensor Network Testbed) that has been developed and deployed in order to allow the evaluation through experimentations of scalable wireless sensor network protocols and applications. SensLAB’s main and most important goal is to offer an accurate open access multi-users scientific tool to support the design, the development tuning, and the experimentation of real large-scale sensor network applications. The SensLAB testbed is composed of 1024 nodes over 4 sites. Each site hosts 256 sensor nodes with specific characteristics in order to offer a wide spectrum of possibilities and heterogeneity. Within a given site, each one of the 256 nodes is able both to communicate via its radio interface to its neighbors and to be configured as a sink node to exchange data with any other ” sink node ”. The hardware and software architectures that allow to reserve, configure, deploy firmwares and gather experimental data and monitoring information are described. We also present demonstration examples to illustrate the use of the SensLAB testbed and encourage researchers to test and benchmark their applications/protocols on a large scale WSN testbed.

Energy-efficient communication protocol is a primary design goal for Wireless Sensor Networks (WSNs). Many efforts have been done to save energy: MAC with duty cycle, energy-aware routing protocols, data aggregation schemes, etc. Recently, beacon-less strategies have emerged as new direction to improve considerably the WSN lifetime. However, the main contributions are not suitable to real radio environments because of hole avoiding strategies based on either planarization or explicit neighbor solicitations. We propose PFMAC (Pizza-Forwarding Medium Access Control), which combines beacon-less geo-routing and energy efficient MAC protocol via a cross-layer design to save more energy with higher reliability. PFMAC supports radio interferences, asymmetric radio links, etc. PFMAC supports a greedy forwarding strategy and, a reactive and optimized neighborhood discovery at 2-hop to deal with holes. Intensive simulations are proposed to highlight the behavior and the performance of PFMAC compared to BOSS over BMAC.

In this paper, we are interested in evaluating the performance of indoor wireless networks in LOS and NLOS conditions. We first provide a thorough analysis of received signal strength indicator (RSSI) measurements obtained in indoor experiments. We aim to better understand the temporal and spatial correlations of RSSI values. Moreover, we analyze in detail the stability of radio links over time. We show that human activity has a significant impact on the performance of wireless networks and that RSSI is not always a robust indicator for the radio link quality. Based on our results, we then propose a time-varying simulation model to reproduce the RSSI variability as observed in our experiments. We argue that this model results in more realistic and accurate RSSI values to be used in simulation.

The proliferation of mobile computing devices and local wireless networks has promoted a growing interest in location-aware systems. The problem with existing positioning techniques is that they are designed to position dimensionless objects. Such an assumption may lead to practical inconsistencies, as objects might overlap in the resulting coordinate system. Moreover, it is usual to neglect the effects of an object volume and its physical characteristics on signal propagation. In the scenario considered throughout this paper (positioning containers in a harbor), such characteristics can be finely estimated. We propose VAPS, a volume-aware positioning system that takes advantage of the waveguide effect generated by containers. Although VAPS is specific to the harbor scenario, its principles can be extended and adapted to other situations. VAPS maps discrete RSSI levels into hop-counts and relies on realistic propagation models to obtain near-perfect positioning at a very low control overhead. Our results demonstrate that, in scenarios where the assumptions made by traditional approaches fail, the new considerations of VAPS do make a difference.

In wireless sensor networks, data dissemination is generally performed from the sensor nodes towards a static sink. In this paper, we address the particular case where the sink is mobile, according to an unpredictable mobility pattern. First, we study existing approaches. As an alternative, we present the line-based data dissemination (LBDD) protocol. Next, we analytically evaluate the communication cost of this protocol and we compare it to other approaches. Finally, realistic simulations are performed and results show that LBDD outperforms previous approaches and presents the best tradeoff among the evaluated protocols.

In wireless multi-hop networks, there is a growing need for the performance evaluation of protocols and distributed applications. Due to the high complexity of analytical models and the difficulty of setting up large scale testbeds, simulations are generally considered as the most convenient methodology for evaluating the performance of wireless systems. However, evaluating a protocol can be a tedious task as it depends on several factors including the physical layer modeling and the dimensioning of protocol or environment parameters. As a consequence inferring an overall performance for a protocol is a real issue. In this paper, we highlight some main factors that may affect the accuracy of the performance evaluation of protocols using simulations. First, we focus on the physical layer (PHY) modeling which represent a key point for the sake of realism and confidence. Second, we discuss the impact of some protocol parameters on the behavior of the evaluated protocols.

Presents the introductory welcome message from the conference proceedings.

Since 2005, IEEE 802.11-based networks are able to provide a certain level of Quality of Service by the means of service differentiation, thanks to the IEEE 802.11e amendment. However, no mechanism or method has been standardized to accurately evaluate the amount of resources remaining on a given channel. Such evaluation would however be a good asset for bandwidth-constrained applications. In multi-hop ad hoc networks, such evaluation becomes even more difficult. Consequently, despite the various contributions around this research topic, the estimation of the available bandwidth still represents one of the main issues in this field. In this article, we propose an improved mechanism to estimate the available bandwidth in IEEE 802.11-based ad hoc networks. Through simulations, we compare the accuracy of the estimation we propose to the estimation performed by others state of the art QoS protocols, BRuIT, AAC and QoS-AODV.

We present Worldsens,a prototyping and development work for wireless sensor protocols and applications.Our relies on two simulators,WSim and WSNet, and proposes a full simulation and performance estimation embedded platforms with instruction and radio byte curacy.During this demo,we propose to demonstrate the interest of the two simulators in stand-alone and in cooperative use.

In this paper we present Worldsens, an integrated environment for development and rapid prototyping of wireless sensor network applications. Our environment relies on software simulation to help the designer during the whole development process. The refinement is done starting from the high level design choices down to the target code implementation, debug and performance analysis. In the early stages of the design, high level parameters, like for example the node sleep and activity periods, can be tuned using WS-Net, an event driven wireless network simulator. WSNet uses models for applications, protocols and radio medium communication with a parameterized accuracy. The second step of the sensor network application design takes place after the hardware implementation choices. This second step relies on the WSim cycle accurate hardware platform simulator. WSim is used to debug the application using the real target binary code. Precise performance evaluation, including real-time analysis at the interrupt level, are made possible at this low simulation level. WSim can be connected to WSNet, in place of the application and protocol models used during the high level simulation to achieve a full distributed application simulation. WSNet and WSNet+WSim allow a continuous refinement from high level estimations down to low level real-time validation. We illustrate the complete application design process using a real life demonstrator that implements a hello protocol for dynamic neighborhood discovery in a wireless sensor network environment.

In wireless sensor networks, hello protocols for neighbor discovery are a basic service offered by the networking stack. However, their study usually rely on rather simplistic models. In this paper we study the impact of radio interferences on neighbor discovery and we introduce an analytical model to estimate the average number of neighbors that a node may expect to discover in his neighborhood.

In this article, we propose a method to enhance accuracy of the available bandwidth estimation in IEEE 802.11-based ad hoc networks. This method combines medium state monitoring, probability of collision estimation and backoff time evaluation, improving the method described in C. Sarr et al., (2005). We evaluate our solution by simulation on different scenarios and compare it with different QoS routing protocols based on different available bandwidth estimation techniques

This paper outlines the concept of auto-organization in IPv6 networks. If the auto-configuration of hosts is defined by IPv6 and mandatory, these mechanisms can not be assimilated to a fully plug and play feature as IPv6 routers still have to be manually configured. In order to succeed in new domains such as SOHO, a full auto-configuration feature must be offered. Indeed, the configuration of 64 bits IPv6 prefixes and the numbering of links are not really friendly and are error prone especially in home networking. This paper tackles the issues that arise, reviews actual solutions and proposes a totally distributed protocol named NAP for IPv6 routers auto-configuration.

Ever increasing network complexity and/or lack of network managers in a given organization justify the need for autoconfiguration mechanisms. IPv6 already provides mechanisms for autonomous address configuration of end-hosts. However, router address autoconfiguration is still lacking. This paper addresses this issue by studying the no administration protocol (NAP) distributed mechanism and comparing it to centralized mechanisms. Particularly, the evaluation is focused on the time it takes to distribute prefix information to all the nodes when powering up the network. A study on the dependency of this time on some characteristics of the network is also presented. Results show that, for the simulated scenarios, NAP performs better than centralized mechanisms, though, in some cases, the difference is not very important.

The dense multicast zone routing protocol (DMZ) presents a novel approach for robust multicast in mobile ad hoc networks by providing adaptive mesh oriented structures. This paper introduces the notion of dense zone, a high concentration of multicast members in a certain area of the network. Each dense zone owns leader(s) that connect to the multicast source. These leader(s) are the node members located on the upper frontier and thus nearest to the multicast source. This concept provides more robustness and scalability for multicast data transmission in ad hoc networks.

Two different strategies to provide a Virtual Interface Architecture (VIA) layer on top of the Myrinet interconnect network (Boden et al., 1995), are presented. The VIA protocol is based on network user level access to allow high performances and implements a point-to-point bidirectional connected communication model. The first alternative is the integration of a Myrinet device driver in the Modular-VIA architecture. This particular implementation allows an easy development of drivers to support VIA on top of new network devices. However, performances are not fully satisfactory. The second strategy is the development of a VIA interface on top of the Myricom message passing interface. It is a user-level library which achieves great portability, performance and scalability. It can also be integrated in a multi-protocol environment based on the Myrinet media.