In this paper, a novel dynamic awareness model is proposed and applied to coverage control over large-scale task domains using a decentralized multi-vehicle sensor network with intermittent communications and/or faulty sensors. For each vehicle, an individual state of awareness is defined. The individual vehicle's state of awareness continuously evolves based on the vehicle's motion and is updated at discrete instants whenever the vehicle establishes a communication link with other vehicles. This information sharing update step aids in reducing the amount of redundant coverage. Based on this awareness model, a decentralized control strategy is proposed that guarantees that every point within the task domain will be covered with a satisfactory state of awareness under intermittent communications and/or faulty sensors. We demonstrate the results using numerical simulations.