Clustering of multipath components utilizing geometrical parameters for indoor double-directional propagation channel

This paper proposes a new clustering method by using scattering points obtained from the measurement-based ray tracer (MBRT). Wideband MIMO channel sounder was used to measure propagation path in indoor environments. The parameters of multipath components (MPCs) were extracted by the space-alternating generalized expectation-maximization (SAGE) algorithm. Comparison between traditional and new clustering method by altering the parameter vector inputted into KPowerMeans algorithm (KPM) was carried out to assess the performance by using several cluster validity indices (CVIs). Simulation results suggest that the proposed method has a better performance than traditional algorithm.