On the Benefits of Clustered Capillary Networks for Congestion Control in Machine Type Communications over LTE

Abstract

One of the significant complexities that Machine Type Communications (MTC) is expected to bring about is congestion. One promising conceptual approach in the literature to reduce congestion is clustering or grouping based solutions. Designing group based solutions highly depend on particular cases and it is challenging to come up with a universal solution. This paper is based on a project designed to address a grouping based congestion control technique in a specific MTC application scenario using the Capillary Networks approach from an experimental point of view. In this paper, an MTC congestion scene is generated inside a network-simulation environment in a defined "parking lot scenario" during a busy hour, where a huge number of cars and parking space sensors will try to connect to an LTE node in a congestive fashion. Based on the impact of this congestion on the node, the simulation results were analyzed to come up with a reasonable cluster size, a comparison of technologies to communicate within a group, and the benefits and tradeoffs due to clustering and Capillary Network based solutions.