Parametric Sparse Channel Estimation Using Long Short-Term Memory for mmWave Massive MIMO Systems

Abstract

Terahertz (THz) communications is considered as one of key solutions to support extremely high data demand in 6G. One main difficulty of the THz communication is the severe signal attenuation caused by the foliage loss, oxygen/atmospheric absorption, body and hand losses. To compensate for the severe path loss, multiple-input-multiple-output (MIMO) antenna array-based beamforming has been widely used. Since the beams should be aligned with the signal propagation path to achieve the maximum beamforming gain, acquisition of accurate channel knowledge, i.e., channel estimation, is of great importance. An aim of this paper is to propose a new type of deep learning (DL)-based parametric channel estimation technique. In our work, DL figures out the mapping function between the received pilot signal and the sparse channel parameters characterizing the spherical domain channel. By exploiting the long short-term memory (LSTM), we can efficiently extract the temporally correlated features of sparse channel parameters and thus make an accurate estimation with relatively small pilot overhead.From the numerical experiments, we show that the proposed scheme is effective in estimating the near-field THz MIMO channel in THz downlink environments.

Massive Data Generation for Deep Learning-Aided Wireless Systems Using Meta Learning and Generative Adversarial Network

“…Specifically, to investigate the efficacy of D-WiDaC, we use two different types of benchmark datasets: model-based channel samples and real measured channel samples. As a model-based channel dataset, we exploit the samples generated from (7). As a measured channel dataset, we employ the softnull dataset obtained by massive MIMO systems at indoor environments [14].…”

“…As the wireless systems are becoming more complicated, it is very difficult to come up with a simple yet tractable mathematical model and algorithm. As an entirely-new paradigm to handle future wireless systems, deep learning (DL), an approach that the machine learns the desired function without human intervention, has received much attention recently [3]- [7].…”

Towards deep learning-aided wireless channel estimation and channel state information feedback for 6G

“…For example, a mmWave propagation channel (e.g., FR2 in 5G [13]) is characterized by the geometric parameters such as AoD/AoA, path delay, and path gain. Due to the the severe attenuation of signal power caused by the high diffraction and penetration loss, atmospheric absorption, and rain attenuation in the mmWave band, the number of effective paths is at most a few (LoS and at most one or two NLoS paths), meaning that the channel can be represented by a small number of geometric parameters [14]. Using this property as a side information, one can find out the sparse parameters used to reconstruct the mmWave channel.…”

“…While the CS-based channel estimation is effective in dealing with the sparsity of the mmWave channel, it might not work well in practical scenarios where the mismatch between the true angles {θ i , φ i } and the quantized angles in the angular bases [14]. By applying high-resolution angle quantization, one can reduce the error caused by the mismatch.…”

Genetic Algorithm Based BER Aware Channel Selection Using Break Point Technique For Next Generation Milli-Meter (mm) Wave Communication Systems

“…Other models as described in [28,29,30,31,32,33] aim at application specific mmWave channel modelling scenarios, which includes, railway networks, transmitter impaired conditions, blockage prediction for 6 GHz channels, hovering fluctuation consideration for unmanned aerial vehicles (UAVs), mobile to mobile (M2M) mmWave models, and urban microcell environments. On Similar models are discussed in [34,35,36,37,38,39] where BER optimizations are proposed for different scenarios. While the work in [40,41,42,43] proposes use of compact wideband antenna for 5G applications, along with improvement of QoS with routing protocols.…”