Measurement-Based Analysis of Millimeter-Wave Channel Sparsity

Abstract

Millimeter-wave (mmWave) channel sparsity is widely exploited to reduce the complexity of compressed sensing and beamtracking algorithms for 5G, yet there are still many misconceptions about it: that it arises from pencilbeam antennas admitting just a few paths, that the channel in fact has just a few paths, and that the channel is always sparse. What is missing to confirm these conjectures is a measurement-based analysis, to our knowledge conducted by only two papers thus far, however, both at <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">microwave</i> (5 GHz), in a single environment, and based on just a few ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\leq$</tex-math></inline-formula> 49) measurements. For conclusive analysis on <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">mmWave</i> sparsity, we have collected measurements with our high-resolution 3-D double-directional 60 GHz channel sounder in three indoor and two outdoor environments, comprising a total of 750 measurements while investigating both line-of-sight (LoS) and obstructed LoS (OLoS) conditions. We learned that in LoS, the channel is always sparse that sparsity drops notably from LoS to OLoS and that mmWave sparsity is inherent to the channel itself—not to pencilbeam antennas—since our sounder has omnidirectional field-of-view. We also propose a sparsity metric that accounts for channels with just a few paths, which we found to occur in some environments.