Fuzzy deception game using ant-inspired meta-heuristics

Abstract

Deception plays a significant role in many intelligent systems whether natural (insect colonies, animal colonies etc. to human beings) or artificial ones. In order to make artificial deception near the one existing in the real world, fuzzy theory seems a reasonable tool since it is the theory of mind through which the uncertainties can be modeled. In this work a hide and seek problem is considered in a maze imaginable in any form between two robots one of which is trying to deceive the other. Fulfilling the deception, the deceiver is supposed to use two signals (track and smell), the second of which is inspired from natural ants. The deceiver is free to choose whether or not to leave a track on the way and also it is the one who decides on the amount of smell deposited in order to confuse the robot under deception (the mark). After the deceiver decides where to go, the robot under deception (the mark) (on each cross-road) is to decide which path to choose based on a utility function calculated within a fuzzy inference system whose inputs are the value of deception signals. The results show that the robot under deception faces high utilities when coming across the roads the deceiver wants him to enter (less-beneficial roads to the deceiver) and consequently is convinced to do the thing desired by the deceiver.