Self-Adaptive Learning and Cellular Automata based Mobile Crowdsensing

Abstract

Mobile Crowdsensing (MCS) is frequently utilized for computation assignments, but it is particularly useful for sensing complicated environments. Previously, the MCS platform spent a lot of time and effort establishing incentive mechanisms and task assignment algorithms to encourage mobile users to participate. In actuality, because of their sensing environment and other participants' methodologies, MCS participants face numerous uncertainties, and it is unknown how they interact with one another and make sensing decisions. This study uses the perspectives of MCS participants to develop a web detection arrangement that will maximize their payoffs through MCS participation. Self-adaptive cellular automata-based Markov decision process exhibits interactions among mobile clients and detecting contexts. With the help of Self-Adaptive Support Learning (SASL) and Cellular Automata (CA), we developed a novel method that uses the ideal detecting technique for each client to improve the predicted payoff against random detecting scenarios in a stochastic multi-agent environment. With distinct dynamic sensing, the SASL and CA based smart Crowdsensing enhances user’s payoff, as shown in the simulation.