Beam Squint and Channel Estimation for Wideband mmWave Massive MIMO-OFDM Systems

Abstract

With the increasing scale of antenna arrays in wideband millimeter-wave\n(mmWave) communications, the physical propagation delays of electromagnetic\nwaves traveling across the whole array will become large and comparable to the\ntime-domain sample period, which is known as the spatial-wideband effect. In\nthis case, different subcarriers in an orthogonal frequency division\nmultiplexing (OFDM) system will "see" distinct angles of arrival (AoAs) for the\nsame path. This effect is known as beam squint, resulting from the\nspatial-wideband effect, and makes the approaches based on the conventional\nmultiple-input multiple-output (MIMO) model, such as channel estimation and\nprecoding, inapplicable. After discussing the relationship between beam squint\nand the spatial-wideband effect, we propose a channel estimation scheme for\nfrequency-division duplex (FDD) mmWave massive MIMO-OFDM systems with hybrid\nanalog/digital precoding, which takes the beam squint effect into\nconsideration. A super-resolution compressed sensing approach is developed to\nextract the frequency-insensitive parameters of each uplink channel path, i.e.,\nthe AoA and the time delay, and the frequency-sensitive parameter, i.e., the\ncomplex channel gain. With the help of the reciprocity of these\nfrequency-insensitive parameters in FDD systems, the downlink channel\nestimation can be greatly simplified, where only limited pilots are needed to\nobtain downlink complex gains and reconstruct downlink channels. Furthermore,\nthe uplink and downlink channel covariance matrices can be constructed from\nthese frequency-insensitive channel parameters rather than through a long-term\naverage, which enables the minimum mean-squared error (MMSE) channel estimation\nto further enhance performance. Numerical results demonstrate the superiority\nof the proposed scheme over the conventional methods in mmWave communications.\n