Subprocess Transmission Strategies for Recovering from Faults in the Tree-Based Fog Computing (TBFC) Model

Abstract

In order to increase the performance in the IoT (Internet of Things), the fog computing model is proposed. Here, subprocesses to handle sensor data are performed on fog nodes in addition to servers. Output data processed by a subprocess of a fog node is sent to a succeeding node. If a fog node is faulty, the preceding nodes are disconnected, i.e. no output data of preceding nodes can be delivered to servers. Another operational node which supports the same subprocess as the faulty node is an alternate node. In our previous studies, the FTBFC (Fault-tolerant TBFC) and MFTBFC (Modified FTBFC) models are proposed where disconnected nodes send the output data to alternate nodes. In this paper, we newly propose another strategy where a subprocess of a faulty node is transmitted to surrogate nodes which hold data to be processed. In this paper, we propose an SMSGD (Selecting Multiple Surrogates in Grandparent and Disconnected nodes) algorithm to select surrogate nodes. In the evaluation, we show the energy consumption and execution time of each surrogate node can be reduced compared with the data transmission strategy.