Direct Clustering and Multi-Path Component Identification on THz Channel Measurements in a Factory Environment

Multi-path components pose both a challenge and an opportunity in high-frequency wireless communication, especially in environments with complex propagation conditions. In this paper, we derive a clustering algorithm to be applied directly to the measurements of indoor THz propagation. We show that such method does not require preprocessing to identify the peaks of multi-path components, but rather extract the time range of clustered multipath components in measurements. Ray-tracing experiments are performed together with classic clustering methods to validate our solution on the corresponding measurements. Our solution facilitates the identification of both clusters and multi-path components directly on measurements without the need to reconstruct scenario in ray-tracing to identify the sources of specular components.