Proposal and Evaluation of Collaborative Event-Triggered Algorithms in Ultra-Dense Wireless Sensor Network

Abstract

Wireless sensor networks and monitoring of environmental parameters such as humidity, temperature, rainfall, and soil salinity, among others, play a crucial role in precision agriculture. Power consumption can be reduced while maintaining accuracy by using collaborative and event-triggered algorithms. In this fact, reducing the energy used by the network is possible. This paper aims to propose a set of algorithms to improve the Quality of Information and the efficiency of the data collection process in a collaborative wireless sensor network while reducing the measuring frequency. In this way, the system will only be activated when necessary, creating an alarm system that will notify the farmer of a problem. Four algorithms are compared through simulations using a cluster of 16 nodes in 41 iterations that involve the detection of events in four scenarios. The option of including broadcast or multicast in the collaborative algorithm is evaluated. The study shows that the event-driven collaborative algorithm performs equal to or better than event-driven algorithms in all tested scenarios. In Scenario 3, a collaborative-based streaming algorithm has only a 1.45% error compared to 3.20% for multicast and 5.44% for the event-driven algorithm.