Simulating Robust Decentralized Spatial Algorithms

Abstract

Evaluating efficiency helps us understand how much of a system’s resources will be consumed by an algorithm. But efficient algorithms must also be robust: able to reliably generate useful information under a range of different circumstances. Most algorithms must strike a balance between efficiency and robustness; it is frequently possible to increase an algorithm’s robustness at the cost of decreased efficiency. Robustness becomes especially important in the practical contexts of algorithms for geosensor network deployments. Geosensor networks are also expected to operate in environments of uncertainty, for example, where sensors are inaccurate or network coverage is sparse. Further, it is in many cases desirable that decentralized algorithms continue to operate at some level even if the the assumptions upon which they are founded are violated, for example, when communication becomes unreliable.