Demo: Design and Evaluation of MoleNet for Wireless Underground Sensor Networks Idrees Zaman † , Martin Gellhaar † , Jens Dede † , Hartmut Koehler ‡ , Anna Foerster † † Sustainable Communication Networks (ComNets), University of Bremen, Germany ‡ Centre of Environmental Research and Sustainable Technology (UFT), University of Bremen, Germany † {iz, mag, jd, anna.foerster}@comnets.uni-bremen.de ‡ hartmut.koehler@uni-bremen.de Abstract—In this demo, we introduce a new sensor node for underground monitoring called the MoleNet. The MoleNet is designed especially for wireless underground operation. The development of the platform is part of a revitalization project in Cameroon. Due to certain constraints, the project terrain requires a complete underground deployment of the soil monitoring system. The underground channel is quite different from the terrestrial channel and hence presents new challenges for the researchers. The MoleNet strives to overcome these challenges. It focuses on an increased operational life and enhanced com- munication range. The current revision of MoleNet achieves almost 5 years of battery life, an underground to underground (UG2UG) communication range of 7.5m and an underground to aboveground (UG2AG) communication range of 80m. Index Terms—Wireless Underground Sensor Networks, refor- estation monitoring, agricultural monitoring, precision agriculture, MoleNet I. I NTRODUCTION The MoleNet sensor node is designed for the reforestation of a barren land at the Revitec [1] site in Ngaoundere, Cameroon. The MoleNet monitors the volumetric water content (VWC) of the soil and assists in the revitalization process. The project site presents a challenge as its soil is extremely dry. The particles in the dry soil become hydrophobic and hence they do not absorb much water. Even when it rains the gravity drains a large portion of water into the ground and the soil does not hold the water for a longer period of time. Most of the plants cannot withstand the dehydration stress and hence no plant grows in such areas. In order to address this dilemma, old coffee bags filled with soil and seeds are used in order to preserve more water during rainy periods. The structure of coffee bags allows the roots of the plants to grow into the soil below. In order to measure the effectiveness of the coffee bags, constant monitoring of the water content below the placed coffee bags is required. [2] Figure 1 shows the MoleNet sensor node before underground placement. The main objective of this project is to monitor the VWC and temperature of the underground soil over a period of three to four years. In addition, the application demands the deployment of a complete underground wireless sensor network where all the sensor nodes must be buried underground except the base station. The MoleNet sensor node aims to meet these requirements. In this demo, we focus on the design and evaluation of Fig. 1. The MoleNet platform on the field our new MoleNet platform at the single node level. More precisely, we present: • The design and implementation of the MoleNet platform • The evaluation of the MoleNet platform II. DESIGN OF THE MOLENET NODE The design of the MoleNet is based on widely used ”Wattuino Pro Mini” [3] that is powered by Atmega328p microcontroller. The MoleNet uses power efficient components to increase its operational lifetime. It is equipped with 8Mhz external crystal, power efficient MCP1703 [4] regulator, 25LC1024 1Mbit EEPROM [5] for data logging and RV8523 [6] RTC clock for maintaining the time locally and waking the controller from deep sleep mode. Different frequency bands were investigated to maximize the range between two sensor nodes. The authors in [7] already investigated that 2.4GHz band is not suitable for UG2UG or UG2AG communication. Therefore different transceivers for 868MHz and 433MHz ISM bands are scrutinized. HopeRF RFM69CW [8] 433MHz transceiver is then selected based on the UG2UG and UG2AG communication range. Figure 2 shows the top view of the MoleNet sensor node. 2016 IEEE 41st Conference on Local Computer Networks Workshops © 2016, Idrees Zaman. Under license to IEEE. DOI 10.1109/LCNW.2016.10 145