Subtree-Based Fog Computing in the TWTBFC Model

Abstract

In our previous studies, the TBFC (Tree-Based Fog Computing) model of the IoT is proposed to reduce the electric energy consumed by fog nodes and servers. Here, fog nodes are hierarchically structured in a height-balanced tree. A root node shows a cloud of servers and leaf nodes indicate edge fog nodes which communicate with sensors and actuators. First, sensor data is sent to edge fog nodes. Each fog node receives data from child fog nodes, processes the data, and sends the processed data to a parent fog node. Finally, servers receive sensor data processes by fog nodes and send actions to actuators. TWTBFC (Two-Way TBFC) model, fog nodes not only send processed data to a parent fog node but also send both the processed data to each child fog node. Fog nodes nearer to edge fog nodes decide on actions by using the data and send the actions to child fog nodes. In these two previous studies, we assume that every edge node has acquired every processed data and then makes a decision on actions to actuator. In order to process only a small amount of data scattered in various parts of the tree model and then make decision on actions to actuator, we newly propose a subtree-based fog computing in the TWTBFC model, where all edge nodes do not send the sensor data to processed and make decision on actions to actuators. Just some sensor data, i.e. a small amount of data scattered in various parts of the edge is sent to a root node, which needs to make a decision on actions to actuator.