Spatiotemporal-wave-based computing on an active medium, which is known as wave computing paradigm, mimics the way biological systems process information. Recent anatomic and physiological studies and technological developments in very large-scale integration technology have encouraged the engineers to brace up with this new paradigm. In this paper, a simple network model that is essential in order to explore spatiotemporal behavior and to develop new algorithms based on wave computing technique has been presented. The presented network is a 2-D reaction-diffusion cellular neural network consisting of relaxation oscillators. The network can be programmed as an autowave generator and sources of the wave can be placed at any place on the network. The network has the simplest model that exhibits autowaves. Furthermore, the model is suitable for the implementation of a larger network. The propagation of autowaves between the cells whose dynamics is fixed shows that the network can be adapted to already developed wave computing algorithms in literature.