Adaptive modulation

This research area focuses on leveraging index modulation (IM) to improve spectral and energy efficiency in next-generation wireless communication systems. IM transmits additional information by selectively activating subsets of resources such as subcarriers, antennas, or time slots, thereby boosting throughput without increasing power consumption. Understanding the classification, principles, transceiver designs, and practical applications of IM is critical for adaptive modulation implementations that seek to achieve green communication objectives and meet high data rate demands in 5G and future networks.

This theme explores state-of-the-art algorithms for automatic modulation recognition (AMR), which is essential for adaptive modulation systems to identify modulation types accurately under varying channel conditions. Emphasis is on statistical feature extraction (e.g., higher-order cumulants), machine learning classifiers (e.g., neural networks, KNN), and hybrid optimization methods (e.g., genetic algorithms) for improving classification accuracy, especially in low SNR and fading channels. These insights provide actionable approaches for real-time signal demodulation and modulation adaptation in both civilian and military communications.

This research direction addresses the adaptation of modulation schemes specifically for underwater acoustic communication networks, characterized by limited bandwidth, long delays, multipath effects, and dynamic environmental conditions. It centers on the design and evaluation of lightweight machine learning approaches, particularly Multi-Armed Bandit frameworks, that allow fully distributed, real-time modulation adaptation on resource-constrained underwater nodes. These methods aim to optimize network throughput, energy consumption, and packet error rates, supporting robust and efficient UWA communications under stringent operational constraints.