Hardware Evaluation of Cluster-Based Agricultural IoT Network

Abstract

In this paper, we present a real-world hardware evaluation of a robust, affordable, location-independent, simple, and infrastructure-less cluster-based agricultural Internet of Things (CA-IoT) network based on a commercial off-the-shelf (COTS) Bluetooth Low-Energy (BLE) communication technique and Raspberry Pi module 3 B + (RPI 3 B +) to address global food insecurity caused by climate change and increasing global population via precision farming and greenhouses. Using an engineering design approach, an initial centralized agricultural IoT hardware test-bed was implemented with the aid of BLE, RPi 3 B +, DHT22, STEMMA soil moisture sensors, UM25 meters, and LoPy /low-power Wi-Fi modules, among other devices. This test-bed was adapted and modified after the proposed cluster-based architecture to evaluate the performance of CA-IoT networks. This study provides holistic account of our location-independent CA-IoT solution covering the design and deployment experiences that can serve as a reference document to the agricultural Internet of Things (Agri-IoT) community. Additionally, the proposed solution performed satisfactorily when tested under indoor and outdoor (on-farm) environmental conditions in the USA and Senegal. Unlike existing Agri-IoT test-beds, a sample performance evaluation showed that our context-relevant CA-IoT technology is simple to deploy and manage by inexperienced users and is energy-efficient, location-independent, robust, and task- and size-scalable to provide a rich set of measurements for both educational and commercial purposes.