Multi-Bit On–Off Digital Noise Modulation

Abstract

This letter proposes multi-bit on-off digital noise (M-OODN), a time-structured $M$ -ary extension of on-off digital noise (OODN) for noise-based signaling. Each symbol is partitioned into OODN-based sub-symbols, enabling multiple bits to be conveyed through temporal silence-versus-noise patterns without intermediate variance levels while preserving noncoherent energy detection. An exact analytical framework derives the decision-statistic distributions, optimal likelihood-ratio detector, and closed-form bit and symbol error probabilities. The analysis reveals a rate–reliability trade-off, since higher modulation orders reduce the samples available per bit decision. Monte Carlo simulations validate the analysis and show that M-OODN outperforms recent higher-order noise-based modulation schemes under the same average signal-to-noise ratio.