Simulation can be an important step in the development of software for wireless sensor networks and has been the subject of intense research in the past decade. While most previous efforts in simulating wireless sensor networks have focused on protocol-level issues utilizing models of the software implementation, a significant challenge remains in precisely measuring time-dependent properties such as radio channel utilization. One promising approach, first demonstrated by ATEMU, is to simulate the behavior of sensor network programs at the machine code level with cycle-accuracy, but poor performance has so far limited its scalability. In this paper we present Avrora, a cycle-accurate instructionlevel sensor network simulator which scales to networks of up to 10,000 nodes and performs as much as 20 times faster than previous simulators with equivalent accuracy, handling as many as 25 nodes in real-time. We show how an event queue can enable efficient instruction-level simulation of microcontroller programs and allow the hidden parallelism in finegrained sensor network simulations to be extracted, once two core synchronization problems are identified and solved. Avrora's ability to measure detailed time-critical phenomena can shed new light on design issues for large-scale sensor networks.

This paper presents a new carrier sensing mechanism called DVCS (Directional Virtual Carrier Sensing) for wireless communication using directional antennas. DVCS does not require specific antenna configurations or external devices.

Network computing attracts the attention of many computer researchers and scientists because it can better utilize existing computing resources. The key challenge of network computing is the search for the best method to distribute computing resources to submitted tasks. This thesis demonstrates a distributed dynamic scheduling of composite tasks on a grid computing system. It describes how a computer program was written to simulate a real world computer network. Submitted tasks consist of subtasks represented by DAGs. The adopted scheduling and mapping include two steps: one external and the other internal. External scheduling and mapping are performed on the task level, and internal scheduling and mapping are done on the subtask level. A task and its subtask must go through these two steps to be allocated computing resources. This research analyzes different factors on the distributed dynamic scheduling algorithm. The factors include Subtask Waiting Queue size, submitted task number, task submission interval, and network infrastructure. The percentage of tasks completed before deadline

In wireless networking, due to the high complexity of analytical and theoretical models, simulations are generally considered as the most convenient methodology for performance evaluation. Nonetheless, the physical complexity of the wireless medium induces a clear tradeoff between accuracy and scalability in a wireless network simulator design. In this paper, we focus on this tradeoff and study the impact of physical layer (PHY) modeling accuracy in the computational cost of simulations. We first discuss the main aspects of the wireless medium and briefly show how they have been handled in existing simulators. Then, we introduce a flexible and modular PHY simulation framework to analyze in more details their influence on the scalability of simulations. 1.

On behalf of:

The Future Combat Systems (FCS) program is developing the FCS System of Systems Simulation Environment (FSE) to provide the “real world wraparound ” to the FCS System of Systems Simulation Framework (S2F). A primary component of the FCS is the Communication Effects Server (CES) whose objective is to develop a flexible, scalable, and high-performance, packet-level, discreteevent simulator that will accurately portray the behavior of the FCS communications architecture to eventually support the live, constructive, and virtual simulations envisaged in the FSE. In particular, the CES is required to compute, in real-time, accurate end-end latency for every communication message sent over a wireless network in a FSE experiment. This paper provides an overview of the CES that has

Abstract. The pervasive computing paradigm promises great abilities whenever and wherever a user goes. However, as people are shifting from the desktop to more resource-constrained devices, issues due to scarce resources may appear preventing from the use of the available services and applications. In this paper, we consider the adaptive deployment as a mainstream solution to suit service-oriented applications to different context constraints such as the users requirements, the hosts resources, the services properties and the surrounding environments. We put forward a graph-based deployment approach for service-based applications so as to make these applications adaptable to the runtime contextual constraints. We introduce the AxSeL architecture, A conteXtual Service Loader in which services and their dependencies are represented as a bidimensional graph. The dependency graph is then coloured through a process taking into account the devices, services and users constraints. This process aims to choose to load or not a service according to its execution context. A prototype based on Java and OSGi technologies is implemented in order to demonstrate and evaluate our approach. 1