arXiv subjects

Mohammad Reza Ghavidel Aghdam

Publications and source records attributed to Mohammad Reza Ghavidel Aghdam.

3 recordsLinked to original sources

arXiv2026

Deep Machine Learning in MIMO Communication Systems

This paper presents an innovative approach to enhancing machine learning based communication systems, specifically focusing on multiple-input multiple-output (MIMO) configurations using autoencoders. We optimize the transmitter, receiver, and channel simultaneously under conditions of noise and channel fading, aiming to minimize the bit error rate (BER). By incorporating the Rayleigh fading channel a widely recognized model for wireless channel impairments into the autoencoder framework, we directly train the communication system to handle real world conditions. We introduce a novel optimization process tailored for deep learning-based MIMO communication, and thoroughly analyze the resulting BER performance across various signal to noise ratio (SNR) levels. Our simulation results reveal that the proposed end-to-end wireless communication system achieves significantly lower BER compared to conventional block-based processing methods, highlighting its potential for more efficient and reliable wireless communication.

eess.SP↗

arXiv2023

Optimal Prototype Filter Design Based on New Mathematical Model of GFDM System

We introduce a new perspective and mathematical model for the generalized frequency division multiplexing (GFDM) to analyze the effect of self-interferences and remove their detrimental impacts. Using this model, we express the GFDM signal in a vector form where the self-interferences are represented using two independent terms. For the first time, we model these interferences in terms of prototype filter parameters in the frequency domain and derive their analytical expressions. Building on these expressions, we propose the idea of optimal filter design as one of the solutions to annihilate self-interference in GFDM systems. We introduce the relationship between the required bandwidth for each subcarrier and the number of subsymbols in the proposed optimal prototype filter and show how the prototype filter optimally exploits the total system bandwidth. We derive analytical expressions to evaluate the bit error rate (BER) of the system and show how it outperforms its recent counterparts in the literature. We also perform computer simulations to verify the analytical results and find that the analytical results match well with those of the computer simulations.

eess.SP↗

arXiv2023

Optimal Complex-Valued Prototype Filter Design for GFDM Systems

One of the main challenges with generalized frequency division multiplexing (GFDM) systems is prototype filter design. A poorly designed filter increases inherent and out-of-band (OOB) interferences. In this paper, we introduce a novel optimal prototype filter for GFDM systems that eliminates the negative effects of intrinsic interferences. We introduce a complex-valued pulse shape similar to a single-sideband (SSB) modulation scheme, which significantly improves bandwidth efficiency. Specifically, we introduce an optimization problem to design an optimal pulse shape filter to reduce all intrinsic interference to zero. We derive analytical expressions to evaluate the bit error rate (BER) of the system and show how the designed optimal prototype filter outperforms its current counterparts.

eess.SP↗