Search

Cognitive radio has been recommended for enhancing wireless spectrum usage, and an important part of cognitive radio is spectrum sensing. Conventional spectrum sensing techniques rely on energy detection and feature extraction from a received signal at a specific position. The advancement of artificial intelligence and deep learning has provided a chance to increase spectrum sensing efficiency. The proposed investigation presented a hybrid model comprising CNN, BILSTM, and Transformer Networks (TN) for spectrum sensing. With the aid of Transformer networks, the proposed model improves sensing accuracy even for lower SNR signals, particularly at −20 dB. The results of the proposed technique show that CNN-BILSTM-TN enhances spectrum sensing in comparison to earlier models studied in this domain. The proposed approach outperforms prior methods in terms of F1 Score and has an improved probability of detection at −20 dB and a lower likelihood of missed detection. Analysis of ten modulation techniques' performance criteria, such as the Jaccard index, Cohen Kappa coefficient, and Matthew's correlation coefficient, demonstrates that the suggested method's spectrum sensing is better even at lower SNRs.

The microbial community composition and biogeochemical dynamics of coastal permeable (sand) sediments differs from cohesive (mud) sediments. Tide- and wave-driven hydrodynamic disturbance causes spatiotemporal variations in oxygen levels, which select for microbial generalists and disrupt redox cascades. In this work, we profiled microbial communities and biogeochemical dynamics in sediment profiles from three sites varying in their exposure to hydrodynamic disturbance. Strong variations in sediment geochemistry, biogeochemical activities, and microbial abundance, composition, and capabilities were observed between the sites. Most of these variations, except for microbial abundance and diversity, significantly correlated with the relative disturbance level of each sample. In line with previous findings, metabolically flexible habitat generalists (e.g., Flavobacteriaceae, Woeseaiceae, Rhodobacteraceae) dominated in all samples. However, we present evidence that aerobic specialists such as ammonia-oxidizing archaea (Nitrosopumilaceae) were more abundant and active in more disturbed samples, whereas bacteria capable of sulfate reduction (e.g., uncultured Desulfobacterales), dissimilatory nitrate reduction to ammonium (DNRA; e.g., Ignavibacteriaceae), and sulfide-dependent chemolithoautotrophy (e.g., Sulfurovaceae) were enriched and active in less disturbed samples. These findings are supported by insights from nine deeply sequenced metagenomes and 169 derived metagenome-assembled genomes. Altogether, these findings suggest that hydrodynamic disturbance is a critical factor controlling microbial community assembly and biogeochemical processes in coastal sediments. Moreover, they strengthen our understanding of the relationships between microbial composition and biogeochemical processes in these unique environments., (© 2021. The Author(s).)