In the TBFC (Tree-Based Fog Computing) model to energy-efficiently realize the IoT, fog nodes are tree-structured but it is not easy to be tolerant of stop-faults of fog nodes and adaptive to the change of sensor data traffic. The NBFC (Network-Based Fog Computing) model is proposed to make the FC (Fog Computing) model of the IoT (Internet of Things) fault-tolerant and adaptive to the performance change. Here, a collection of fog nodes cooperate with one another by processing sensor data and exchanging processed data in networks. Each fog node receives input data from other fog nodes, generates output data by executing application processes for the input data, and finds a succeeding fog node which can process the output data. In this paper, we propose an SSFN (Selection of a Succeeding Fog Node) algorithm for each fog node to find an energy-efficient succeeding fog node in the NBFC model. In the evaluation, we show the total energy consumption of the fog nodes selected in the SSFN algorithm is smaller than the TBFC model.