Mapping Channels in Space and Frequency
Mapping (extrapolating) channels from an antenna set at a certain frequency/location to the channels at another antenna set at a (generally) different frequency/location
Key ideas
- Proving the existence of channel mapping functions in space and frequency
- Leveraging deep learning to map (extrapolate) channels in space and frequency
- Uplink to downlink channel prediction in FDD massive MIMO
- Sub-6GHz channels to mmWave beam prediction
- Using a few antenna channels to predict the other antenna channels
- Proactive channel prediction for high mobility mmWave applications
Applications
- Reduce the channel estimation overhead in FDD massive MIMO systems
- Reduce the beam training overhead in mmWave MIMO systems
More information about this research direction
Paper: Muhammad Alrabeiah and Ahmed Alkhateeb, “Deep Learning for TDD and FDD Massive MIMO: Mapping Channels in Space and Frequency,” 2019 53rd Asilomar Conference on Signals, Systems, and Computers, 2019, pp. 1465-1470.
Abstract: Can we map the channels at one set of antennas and one frequency band to the channels at another set of antennas- possibly at a different location and a different frequency band? If this channel-to-channel mapping is possible, we can expect dramatic gains for massive MIMO systems. For example, in FDD massive MIMO, the uplink channels can be mapped to the downlink channels or the downlink channels at one subset of antennas can be mapped to the downlink channels at all the other antennas. This can significantly reduce (or even eliminate) the downlink training/feedback overhead. In the context of cell-free/distributed massive MIMO systems, this channel mapping can be leveraged to reduce the fronthaul signaling overhead. In this paper, we introduce the new concept of channel mapping in space and frequency, where the channels at one set of antennas and one frequency band are mapped to the channels at another set of antennas and frequency band. First, we prove that this channel-to-channel mapping function exists under certain conditions. Then, we leverage the powerful learning capabilities of deep neural networks to efficiently learn this non-trivial channel mapping function, which is also confirmed by the simulation results.
@INPROCEEDINGS{Alrabeiah2019,
author={Alrabeiah, Muhammad and Alkhateeb, Ahmed},
booktitle={2019 53rd Asilomar Conference on Signals, Systems, and Computers},
title={Deep Learning for TDD and FDD Massive MIMO: Mapping Channels in Space and Frequency},
year={2019},
volume={},
number={},
pages={1465-1470},
doi={10.1109/IEEECONF44664.2019.9048929}}
To reproduce the results in this paper:
Example: Steps to generate the results in this figure
- Generate a dataset for scenario I1_2p4 using the settings in this table. Number of paths should be 1.
- Organize the data into a MATLAB structure named “rawData” with the following fields: channel and userLoc. “channel” is a 3D array with dimensions: # of antennas X # of sub-carriers X # of users while “userLoc” is a 2D array with dimensions: 3 X # of users.
- Save the data structure into a .mat file.
- In the file main, set the option: options.rawDataFile1 to point to the .mat file.
- Run main.m
