Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Sparse Bayesian Multi-Task Learning of Time-Varying Massive MIMO Channels with Dynamic Filtering

Arash Shahmansoori

TL;DR

Using the dynamic information of the sparse channel is used to initialize the hyperparameters in the MT-SBL procedure for the next time step, and the power leakage is reduced due to considering angular refinement in the proposed algorithm.

Abstract

Sparsity of channel in the next generation of wireless communication for massive multiple-input-multiple-output (MIMO) systems can be exploited to reduce the overhead in the training. The multitask (MT)-sparse Bayesian learning (SBL) is applied for learning time-varying sparse channels in the uplink for multi-user massive MIMO orthogonal frequency division multiplexing systems. In particular, the dynamic information of the sparse channel is used to initialize the hyperparameters in the MT-SBL procedure for the next time step. Then, the expectation maximization based updates are applied to estimate the underlying parameters for different subcarriers. Through the simulation studies, it is observed that using the dynamic information from the previous time step considerably reduces the complexity and the required time for the convergence of MT-SBL algorithm with negligible sacrificing of the estimation accuracy. Finally, the power leakage is reduced due to considering angular refinement in the proposed algorithm.

Sparse Bayesian Multi-Task Learning of Time-Varying Massive MIMO Channels with Dynamic Filtering

TL;DR

Using the dynamic information of the sparse channel is used to initialize the hyperparameters in the MT-SBL procedure for the next time step, and the power leakage is reduced due to considering angular refinement in the proposed algorithm.

Abstract

Sparsity of channel in the next generation of wireless communication for massive multiple-input-multiple-output (MIMO) systems can be exploited to reduce the overhead in the training. The multitask (MT)-sparse Bayesian learning (SBL) is applied for learning time-varying sparse channels in the uplink for multi-user massive MIMO orthogonal frequency division multiplexing systems. In particular, the dynamic information of the sparse channel is used to initialize the hyperparameters in the MT-SBL procedure for the next time step. Then, the expectation maximization based updates are applied to estimate the underlying parameters for different subcarriers. Through the simulation studies, it is observed that using the dynamic information from the previous time step considerably reduces the complexity and the required time for the convergence of MT-SBL algorithm with negligible sacrificing of the estimation accuracy. Finally, the power leakage is reduced due to considering angular refinement in the proposed algorithm.

Paper Structure

This paper contains 8 sections, 16 equations, 2 figures, 1 algorithm.

Table of Contents

  1. Introduction
  2. The Proposed Method
  3. System Model
  4. MT-SBL learning method with DF
  5. The Algorithm
  6. Simulation Results
  7. Conclusion
  8. Optimal $\boldsymbol{\alpha}$ based on Dynamic Information

Figures (2)

  • Figure 1: Hierarchical model representation of Bayesian multi-task sparse channel learning with DF and the unknown underlying dynamic operation $\mathbf{f}_{t}(.)$.
  • Figure 2: (top-left) Required number of the iterations for the convergence of the algorithm \ref{['alg_sbl']} with respect to the time steps $t$ averaged over $100$ realizations. (top-right) The RMSE of estimated and tracked values of the norm of the virtual channel averaged over subcarriers with respect to the time steps $t$. (bottom) An example of virtual channel tracking over $0<t\leq T$.