Digital Modulation Augmentations with Adaptive Modulation and Coding and AI in Cellular Networks.

Abstract

A user closer to a base station can operate using a higher order of quadrature  modulation, such as 64-QAM, for its communication link, however once the channel’s condition vary such as distance or noise, the throughput and spectral efficiency drops and results in unreliable and inefficient link. The study aimed to implement digital modulation techniques, for efficient data transmission in cellular networks. So the efficacy of Adaptive Modulation and Coding (AMC) and the prospect of Artificial Intelligence (AI) were considered as solution. Simulation was carried out in MATLAB R2025a and compared the performance of the various digital modulation techniques in their SNR and BER characteristics, in presence of Additive White Gaussian Noise. Also adaptive modulation and deep learning were applied to ascertain their effectiveness. The results showed that AMC augmented approaches gave the best BER/SNR performances, followed by M-QAM, QPSK, and BPSK in that order under fixed SNR situations. The use of Convolution Neural Network (CNN) gave 100% prediction for BPSK, 95% for QPSK, 90% for 16-QAM and 85% for the 64-QAM. So the confusion matrix favoured all the modulation schemes and mostly the lower order ones.