Agent-Based System for Reliable Machine-to-Machine Communication

Reliability is, along with energy efficiency and security, commonly referred to as one of the most important requirements which have to be met to successfully deploy Machine-to-Machine (M2M) communication systems. We propose an enhancement of agent-based system for M2M communication which ensures two reliability levels according to the intervals in which data obtained from sensors is sent to back-end system. Our assumption is that communication reliability is less important for reporting measurements sent with higher frequency than for reporting measurements sent with lower frequency. A mechanism is proposed which determines the appropriate reliability level. Introduced mechanism is implemented on two Libelium Waspmote devices which collect meteorological data in different intervals. Suitability for using one of the two proposed reliability levels is determined according to measured communication reliability within 24 h and energy consumption data for one operating cycle.