A Mixture Density Channel Model for Deep Learning-Based Wireless Physical Layer Design

Martí, Dolores García; Beltrán, Joan Palacios; Lacruz, Jesus O.; Widmer, Joerg

doi:10.1145/3416010.3423229

Cited by 6 publications

(7 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is a standard technique for modulation and demodulation (decoding), which does not adapt based on the channel conditions. 8 IQ imbalance is a common issue in radio frequency communications that introduces distortions to the final constellation. mixture distributions.…”

Section: Discussionmentioning

confidence: 99%

“…We evaluate the performance of our adaptation for 20, 35 and 50 samples per symbol. We introduced an IQ imbalance-based distortion to the constellation 8 , and gradually increase the level of imbalance to the system. The BLER of the proposed adaptation methods and the baseline methods (16-QAM and no adaptation) is shown as a function of the IQ imbalance in Fig.…”

Section: Autoencoder Adaptation On Real Fpga Tracesmentioning

confidence: 99%

“…The training algorithm for jointly learning the autoencoder and channel model (based on [8]) is given in Algorithm 1. It is an alternating (cyclic) optimization of the channel parameters and the autoencoder (encoder and decoder) parameters.…”

Section: Limitations and Future Workmentioning

confidence: 99%

“…Aoudia et al [3] present a different model-free approach for autoencoder learning where the channel does not require a model and is allowed to be non-differentiable. Recently, [8] proposed to use a mixture density network to learn the channel since Gaussian mixtures are known to be universal approximators that can accurately capture the channel distribution, given sufficient parametric complexity. Garcia et al [8] also propose to adapt the MDN model to changing channel conditions by training the MDN with a small set of samples from a new distribution.…”

Section: A3 Related Workmentioning

confidence: 99%

“…For realistic wireless channels, the transfer function and conditional probability density are usually unknown and hard to approximate well with standard mathematical models. Recently, a number of works have applied generative models such as conditional generative adversarial networks (GANs) [4,5], MDNs [8], and conditional variational autoencoders (VAEs) [9] for modeling the wireless channel. To model a wireless channel, generative methods learn a parametric model P θc (y | x) (possibly a neural network) that closely approximates the true conditional density of the channel from a dataset of channel input, output observations.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Few-Shot Domain Adaptation For End-to-End Communication

Raghuram¹,

Zeng²,

Martí³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

The problem of domain adaptation conventionally considers the setting where a source domain has plenty of labeled data, and a target domain (with a different data distribution) has plenty of unlabeled data but none or very limited labeled data. In this paper, we address the setting where the target domain has only limited labeled data from a distribution that is expected to change frequently. We first propose a fast and light-weight method for adapting a Gaussian mixture density network (MDN) using only a small set of target domain samples. This method is well-suited for the setting where the distribution of target data changes rapidly (e.g., a wireless channel), making it challenging to collect a large number of samples and retrain. We then apply the proposed MDN adaptation method to the problem of end-of-end learning of a wireless communication autoencoder. A communication autoencoder models the encoder, decoder, and the channel using neural networks, and learns them jointly to minimize the overall decoding error rate. However, the error rate of an autoencoder trained on a particular (source) channel distribution can degrade as the channel distribution changes frequently, not allowing enough time for data collection and retraining of the autoencoder to the target channel distribution. We propose a method for adapting the autoencoder without modifying the encoder and decoder neural networks, and adapting only the MDN model of the channel. The method utilizes feature transformations at the decoder to compensate for changes in the channel distribution, and effectively present to the decoder samples close to the source distribution. Experimental evaluation on simulated datasets and real mmWave wireless channels demonstrate that the proposed methods can quickly adapt the MDN model, and improve or maintain the error rate of the autoencoder under changing channel conditions.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Autoencoder Adaptation On Real Fpga Tracesmentioning

confidence: 99%

Section: Limitations and Future Workmentioning

confidence: 99%

Section: A3 Related Workmentioning

confidence: 99%