Wireless sensor networks have resources of limited capacity (e.g. bandwidth, processing power, memory and energy). That is why these resources should be efficiently used. Node activity scheduling is a technique that allows nodes to alternate sleep and awake states. This technique spares energy insofar as the sleep state is the state using the smallest power. Moreover, by allowing several nodes to transmit simultaneously without interfering, spatial reuse of the bandwidth is obtained. Furthermore with a smart schedule, data gathering can be done in a single cycle. All these reasons render node activity scheduling very attractive in wireless sensor networks. We propose in this paper a three-hop coloring algorithm for data gathering applications. Simulation results allow us to evaluate the number of colors needed to color all network nodes and hence to determine the reduced size of the activity period in each polling cycle. The complexity of our algorithm is given in terms of number of messages sent per node. We can then determine the network configurations for which coloring brings interesting benefits, namely a more efficient use of the bandwidth and the node energy, as well as a shorter delay to collect data ensuring their time consistency.