In our previous studies, the NBFC (Network-Based Fog Computing) model is proposed to energy-efficiently realize the IoT. Here, each fog node generates output data from sensor data and sends the output data to a succeeding fog node. We consider a linear application where a sequence of application processes are sequentially executed on sensor data. In our previous studies, on receipt of data from another fog node, only if each fog node supports a process p which can process the data, every process following the process p is executed. We more reduce the total energy consumption of fog nods, where for each process p to be executed on a node, we decide on which succeeding fog node processes following p are to be energy-efficiently executed. In this paper, we propose an EC algorithm to select a succeeding fog node for each fog node, where the average energy consumption of processes of the succeeding fog node is the smallest. In the evaluation, we show the total energy consumption of the nodes in the EC algorithm is 40 to 66% of the other algorithms.