An Innovative Approach to Improve Elasticity and Performance of Message Brokers for Green Smart Cities

Abstract

The smart cities vision aims at exploiting the most advanced Internet of Things (IoT) technologies to support added-value services for the mitigation of the problems generated by the urban population growth. Nowadays, pollution, is one of the most worrying of these problems due to its direct and documented impact in the health of living organisms. However, in order to make cities smart, there are some limitations to overcome such as the performance and scalability problems of the majority of the message brokers, which are the cornerstone of IoT and Smart City environments. In this paper we present an innovative design of a context broker which relies internally on a context-aware content-based publish-subscribe (CA-CBPS) middleware specifically designed to be elastic. This solution follows the FIWARE-NGSI v2 (Next Generation Service Interfaces) specification to manage the entire life cycle of context information across an IoT platform. Additionally, we present a use case using that specification for monitoring sky brightness with a network of photometers, as a first step to reduce and mitigate light pollution. Finally we demonstrate with an evaluation how the proposed broker overcomes the performance and elasticity limitations of the current reference implementation of the FIWARE IoT Platform's context broker (Orion), both implementing the same specification above mentioned.