Dynamic channel allocation in electromagnetic nanonetworks for high resolution monitoring of plants

We investigate techniques to enable communication in the THz band between graphene-based nanoscale devices and microscale network components for agricultural crop-monitoring applications. The properties of THz communications, in particular sensitivity to moisture levels on the communications path and attenuation by obstacles (e.g., leaves) mean that achieving a desired level of throughput of monitoring data can be difficult. Using a simplified model of plant structure and typical plant moisture patterns, we analyze the performance of four frequency selection strategies in terms of throughput and energy utilization for varying numbers of nano and microscale devices, moisture concentration patterns and plant leaf densities. We find that a Two-Phase optimization strategy for frequency selection performs best in a wide range of operational conditions and that leaf density has a significant impact on achievable throughput. Our plant model could serve as a useful basis for planning the necessary concentration of nano and microscale devices to deploy on particular crop types in order to meet given network performance targets.