An Analysis of Minimum Error Entropy Loss Functions in Wireless Communications

TL;DR

This work establishes MEE as a promising alternative for wireless communication tasks in deep learning models, enabling better resilience and adaptability and proposes a less complex, matrix based version of the MEE function to enhance practical usability in wireless communications.

Abstract

This paper introduces the minimum error entropy (MEE) criterion as an advanced information-theoretic loss function tailored for deep learning applications in wireless communications. The MEE criterion leverages higher-order statistical properties, offering robustness in noisy scenarios like Rayleigh fading and impulsive interference. In addition, we propose a less computationally complex version of the MEE function to enhance practical usability in wireless communications. The method is evaluated through simulations on two critical applications: over-the-air regression and indoor localization. Results indicate that the MEE criterion outperforms conventional loss functions, such as mean squared error (MSE) and mean absolute error (MAE), achieving significant performance improvements in terms of accuracy, over $20 \%$ gain over traditional methods, and convergence speed across various channel conditions. This work establishes MEE as a promising alternative for wireless communication tasks in deep learning models, enabling better resilience and adaptability.

Figures (6)

• Figure 1: Different loss functions used in DL
• Figure 2: The proposed system model involves a user transmitting a signal through a wireless channel to a CPU, which then predicts the corresponding target output.
• Figure 3: The comparison of MEE, MSE, and MAE setting SNR to $5$ dB using three different channel conditions: (a) ideal channel, (b) AWGN channel, and (c) Rayleigh fading channel.
• Figure 4: Convergence of the MEE loss function compared to MAE and MSE loss functions during training for the localization use case as a function of the training epochs.
• Figure 5: Comparison of the MED for the proposed MEE loss function compared to MSE and MAE loss functions. three different loss functions-MEE, MSE and MAE-in the context of the localization use case.