Scholarlink

<p>Recent technological advances in wireless communication and the great demand for deep data mining in the sea have led to the research of an underwater communication system. As the channel passes through water, estimating the channel condition over different water effects is critical. Due to the steady transmission of large amounts of data, signal recovery is complicated. In this paper, an enhanced asymmetrically clipped DC biased optical orthogonal frequency division multiplexing system is proposed for accurate channel estimation and signal recovery. Here, a Multi-stage second order attention network with the multi-pooling method is employed to estimate the channel condition, and then a Deep tree training classifier is utilized to classify water type. Moreover, the received signal from the final classifier is detected using a Generative adversarial signal detection network (GA-SD Net). Subsequently, the transmitted signal is recovered through the linear combination of the recovered symbol with the classified water type. The proposed method is implemented using the Matlab tool, and the performance is analyzed using different evaluation metrics. The experimental findings show that the proposed method offers optimized system performance of the inconsistent underwater optical communication (UWOC) channel and proves the improvement under underwater communication circumstances. Also, the model tends to improve data transmission rate and reliability in contrast to conventional models. The throughput is measured at 18 Mbps and 22 Mbps for bandwidths of 1&#xa0;MHz and 20&#xa0;MHz, respectively. The relative entropy is minimized to approximately 0.1 at higher signal-to-noise ratio (SNR) levels (30&#xa0;dB), resembling better classification and channel condition estimation. Nevertheless, the research limitations involve the model complexity, which can pose challenges for real-time implementations in large-scale applications. The practical implication signifies that the proposed system can be effectively used in different underwater applications like subsea operations, environmental monitoring, and marine research, offering a robust solution for high-rate data communication. The originality and value of this paper lie in its innovative approach to incorporating advanced deep learning methods for real-time adaptation to change water conditions. This contributes to the advancement of the UWOC technology and its application.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

JCESR-Net: a joint channel estimation and signal recovery network for underwater optical communication with ADO-OFDM

  • 期刊论文

  • 发表日期：

    2025年8月6日

查看全文
  • B. Priyalakshmi

摘要

Recent technological advances in wireless communication and the great demand for deep data mining in the sea have led to the research of an underwater communication system. As the channel passes through water, estimating the channel condition over different water effects is critical. Due to the steady transmission of large amounts of data, signal recovery is complicated. In this paper, an enhanced asymmetrically clipped DC biased optical orthogonal frequency division multiplexing system is proposed for accurate channel estimation and signal recovery. Here, a Multi-stage second order attention network with the multi-pooling method is employed to estimate the channel condition, and then a Deep tree training classifier is utilized to classify water type. Moreover, the received signal from the final classifier is detected using a Generative adversarial signal detection network (GA-SD Net). Subsequently, the transmitted signal is recovered through the linear combination of the recovered symbol with the classified water type. The proposed method is implemented using the Matlab tool, and the performance is analyzed using different evaluation metrics. The experimental findings show that the proposed method offers optimized system performance of the inconsistent underwater optical communication (UWOC) channel and proves the improvement under underwater communication circumstances. Also, the model tends to improve data transmission rate and reliability in contrast to conventional models. The throughput is measured at 18 Mbps and 22 Mbps for bandwidths of 1 MHz and 20 MHz, respectively. The relative entropy is minimized to approximately 0.1 at higher signal-to-noise ratio (SNR) levels (30 dB), resembling better classification and channel condition estimation. Nevertheless, the research limitations involve the model complexity, which can pose challenges for real-time implementations in large-scale applications. The practical implication signifies that the proposed system can be effectively used in different underwater applications like subsea operations, environmental monitoring, and marine research, offering a robust solution for high-rate data communication. The originality and value of this paper lie in its innovative approach to incorporating advanced deep learning methods for real-time adaptation to change water conditions. This contributes to the advancement of the UWOC technology and its application.