Luiz Felipe Perrone

Simulation validation using direct execution of wireless Ad-Hoc routing protocols

Computer simulation is the most common approach to studying wireless ad-hoc routing algorithms. The results, however, are only as good as the models the simulation uses. One should not underestimate the importance of validation, as inaccurate models can lead to wrong conclusions. In this paper, we use direct-execution simulation to validate radio models used by ad-hoc routing protocols, against real-world experiments. This paper documents a common testbed that supports direct execution of a set of ad-hoc routing protocol implementations in a wireless network simulator. The testbed reads traces generated from real experiments, and uses them to drive direct-execution implementations of the routing protocols. Doing so we reproduce the same network conditions as in real experiments. By comparing routing behavior measured in real experiments with behavior computed by the simulation, we are able to validate the models of radio behavior upon which protocol behavior depends. We conclude that it is possible to have fairly accurate results using a simple radio model, but the routing behavior is quite sensitive to one of this models parameters. The implication is that one should: i) use a more complex radio model that explicitly models point-to-point path loss; or ii) use measurements from an environment typical of the one of interest; or iii) study behavior over a range of environments to identify sensitivities.

A scalable simulator for TinyOS applications

Large clouds of tiny devices capable of computation, communication and sensing, the goal of the Smart Dust project, will soon become a reality. Hardware miniaturization is shrinking devices and research in software is producing applications that allow devices to communicate and cooperate toward a common goal. Success on the software front hinges on the design of algorithms that can scale up with system size. Given that the number of individual cooperating devices will reach high orders of magnitude (hundreds of thousands or even millions), debugging and evaluating the software in such a large system can reap much benefit from simulation. This paper describes the design of a scalable and flexible simulator which allows for the direct execution, at source code level, of applications written for TinyOS, the operating system that executes on Smart Dust. This simulator also provides detailed models for radio signal propagation and node mobility.

Empirical Validation of Wireless Models in Simulations of Ad Hoc Routing Protocols

Computer simulation has been used extensively as an effective tool in the design and evaluation of systems. One should not, however, underestimate the importance of validation—the process of ensuring whether a simulation model is an appropriate representation of the real-world system. Validation of wireless network simulations is difficult due to strong interdependencies among protocols at different layers and uncertainty in the wireless environment. The authors present an approach of coupling direct-execution simulation and traces from real outdoor experiments to validating simple wireless models that are used commonly in simulations of wireless ad hoc networks. This article documents a common testbed that supports direct execution of a set of ad hoc routing protocol implementations in a wireless network simulator. By comparing routing behavior measured in the real experiment with behavior computed by the simulation, the authors validate the models of radio behavior upon which protocol behavior depends.

Enhancing the Credibility of Wireless Network Simulations with Experiment Automation

The last few years have witnessed a growing consensus around the notion that many papers discussing wireless network simulation are plagued by issues that weaken their scientific value. A number of articles have shown evidence of this crisis of credibility and identified many of its causes. In this paper, we show that the methodology flaws in wireless network simulation can be avoided with the use of a framework for experiment automation. We describe the rationale that drove us to develop tools for component-based simulators intending to guide the experimental process from first to last stages. We conclude that a framework that imposes the right constraints on the experimenter can lead to more credible simulation studies. The framework we present helps the construction of consistent models, the definition of model parameters, the design and the execution of experiments, the analysis of output data, and the preparation of data for the dissemination of results that allow experiments to be reproduced.

Using N-body algorithms for interference computation in wireless cellular simulations

A comprehensive simulation model of wireless cellular networks must include the computation of transmitter power levels. In such systems, as time evolves, powers are continuously updated to minimize interference and maintain signal quality. Transmitters operate at the minimum power required to meet a target signal to noise ratio (SNR), which, in the real system, can be promptly estimated since the values involved come front direct measurements. In a simulation model, however, the interference over each receiver is a quantity that must be computed and the associated costs are not low. A system with N pairs of transmitters and receivers requires that O(N/sup 2/) pairwise interactions be computed; its easy to see how very large the workload is when we consider that, in order to advance simulated time by one second, this large computation may have to be performed hundreds of times. We show that techniques devised for the simulation of systems of self-gravitating bodies (N-body problem) can be successfully applied to reduce the complexity of interference computations in simulations of wireless systems. However, our experiments suggest simple distance-based truncation may be the superior method.

Could a Caveman Do It? The Surprising Potential of Simple Attacks

The research projects our investigations have spawned span topics in attack modeling, quantifying the effects of attacks on network performance and robustness, and the construction of computational tools for managing our simulation experiments. We hope, however, that our works meta-contribution is to show, by proof of concept, that you must consider simple, sideways system attacks before you can rely on a system. We encourage engineers of mission-critical systems to look for then cavemen (or squirrels) and to study carefully what actions these adversaries might take against the products they roll out. In this paper we examine simple attacks on routing protocols for wireless networks. In our study, the system was to be a wireless ad hoc network, a collection of independent, possibly mobile computing devices that communicate using radio frequency technology.