For Intelligent and Higher Spectrum Efficiency: A Variable Packing Ratio Transmission System Based on Faster-than-Nyquist and Deep Learning

TL;DR

This paper introduces a variable packing ratio transmission system based on Faster-than-Nyquist signaling and deep learning, enhancing spectrum efficiency and security in wireless communications through intelligent adaptation and secure dynamic switching.

Contribution

It proposes a novel VPR-based transmission scheme utilizing deep learning for packing ratio estimation, improving spectrum efficiency and security with minimal modifications to existing systems.

Findings

Achieves near the same packing ratio estimation accuracy with 20-fold complexity reduction.

Demonstrates significant spectrum efficiency gains over conventional Nyquist transmission.

Provides a closed-form expression for spectrum efficiency under various channels.

Abstract

With the rapid development of various services in wireless communications, spectrum resource has become increasingly valuable. Faster than Nyquist (FTN) signaling, proposed in the 1970s, is a promising paradigm for improving spectrum utilization. This paper proposes intelligent variable-packing-ratio (VPR)-based transmissions for high spectrum efficiency (SE) and security, respectively. Aided by deep learning (DL)-based estimation, the proposed scheme for high SE can achieve a higher capacity with negligible modification to existing communication paradigms (e.g., spectrum allocation or frame structure). Also, for VPR-based secure transmission, a dynamic generation scheme is proposed to produce randomly distributed positions to switch the packing ratio, which can effectively avoid detections and attacks. In addition, we propose a simplified DL-based packing ratio estimation for both of these two scenarios so that the receiver can estimate the packing ratio without any in-band or out-band control messages. Simulation results show that the proposed simplified estimation achieves nearly the same accuracy and convergence speed as the original multi-branch fully-connected structure with a complexity reduction of 20 folds. Finally, we derive the closed-form SE of the proposed VPR transmission under different channels. The numerical results validate the correctness of the derivation and demonstrate the SE gains of the VPR scheme beyond conventional Nyquist transmission.