Sensor networks are large scale networks of low-power devices that collaborate in order to perform a given task. The sensors are limited in battery energy, capacity and computational power. In recent years, researchers have proposed several protocols for such sensor networks assuming that all sensors belong to the same authority. In this paper, we introduce the concept of multi-domain sensor networks that was, to the best of our knowledge, never considered before. We propose a game-theoretic model to investigate the impact of cooperation and show the conditions for which cooperation is the best strategy.