A Preliminary Fuzzy Markup Language based Approach for the Queue Buffer Size Optimization in Fog Nodes for Stream Processing

Abstract

The Internet of Things (IoT) is usually divided in three layers: Edge, Fog, and Cloud layers. The whole IoT infrastructure deals with large amount of data between layers. Focusing on the Fog layer, the sending/receiving data and further cascade processing of those data in the Fog layer enable the Stream Processing paradigm. Thus, aspects such as the number of connections, delays, buffer size, memory usage, among others, have to be considered to optimize the network traffic. Moreover, these characteristics are affected by uncertainty and imprecision since, for example, the number of connections or the buffer size may be considered low in some cases and high in others. Fuzzy Rule-Based Systems (FRBS) are suitable for addressing complex data and managing their imprecision. The objective of this paper is to propose an approach that optimizes network traffic with the main goal of dynamically and automatically adjusting the queue buffer size in a node to avoid network collapse. The IEEE std 1855-2016 for Fuzzy Markup Language and the open source library JFML are used for their flexibility and interoperability offered by these technologies. The proposal has been simulated in three basic different scenarios involving several network traffic states in a fog infrastructure.