Massive MIMO as an Extreme Learning Machine

Abstract

This work shows that massive multiple-input multiple-output (MIMO) with low-resolution analog-to-digital converters (ADCs) forms a natural extreme learning machine (ELM), where the massive number of receive antennas act as hidden nodes of the ELM, and the low-resolution ADCs serve as the activation function of the ELM. It is demonstrated that by adding biases to received signals and optimizing the ELM output weights, the system can effectively tackle hardware impairments, e.g., the power amplifier nonlinearity at transmitter side. It is interesting that the low-resolution ADCs can bring benefit to the receiver in handling nonlinear impairments, and the most computation-intensive part of the ELM is naturally accomplished by signal transmission and reception.

Novel Multilayer Extreme Learning Machine as a Massive MIMO Receiver for Millimeter Wave Communications

“…Moreover, setting hyperparameters takes a lot of time. In addition, the ELM is widely used in many technological areas, especially in telecommunications signal reception, due to the following: (i) Easy implementation, (ii) extremely fast training speed, and (iii) good generalization performance [11], [16], [26]- [28].…”

“…The work in [27] introduces an ELM receiver for MIMO light-emitting diode (LED) communications with nonlinearities and cross-LED interference. The work in [28] shows the performance of a standard ELM used as a MIMO-OFDM receiver, considering low-resolution ADCs. The investigation shows a MIMO combining process learned online with the help of reference signals.…”

“…The work in [32] considers multiuser scenarios where the inter-user interference is the limiting factor. Finally, the studies in [28], [32] are limited to low-band frequencies.…”

MIMO Detection under Hardware Impairments via Learning from Noisy Labels

“…MIMO detection methods based on machine learning approaches, other than DL, have also been studied to address the challenges posed by nonlinear and unknown distortion caused by hardware impairments [23]- [25]. A representative example is a supervised-learning approach in [23], [24] which explicitly learns the empirical distribution of the distorted received signals from training data.…”

Deep Learning Assisted Automatic Methodology for Implanted MIMO Antenna Designs on Large Ground Plane

“…Recently, machine learning and deep learning proved their talents in optimizing radio frequency designs and they can be a good solution for modeling and optimizing MIMO antennas [14]. They can model the complex designs by considering the relationships between the input and output data that in turn can be design parameters and/or design specifications, respectively [15].…”