摘要
由于海水的散射效应和水下湍流的影响,光在水下传输时会受到多种干扰,水环境越差干扰越复杂。鉴于此,在非对称限幅光正交时频空调制的水下光通信系统中引入了信道均衡技术,并根据水下光的传输特点提出一种基于卷积神经网络的均衡方法。在服从Gamma-Gamma分布函数的海水光弱湍流信道下进行仿真,为了验证仿真的有效性,采用蒙特-卡罗射线追踪模拟来确定实际信道特性,得出实际信道快速傅里叶变换的脉冲响应与Gamma-Gamma函数分布拟合良好。在接收端将接收到的光信号首先转换成电信号,仿真了在不同强度湍流下均衡前的误码率,并且在弱湍流条件下,将卷积神经网络均衡算法,反向传播神经网络均衡算法以及未均衡三者进行了性能对比。仿真结果表明:相比于BP算法,卷积神经网络误码性能有所提升。因此,在信噪比较差的情况下,采用所提方法能够有效提高水下光通信系统的通信性能,降低误码率,提高信道估计的可靠性。
Due to the scattering effect of seawater and the influence of underwater turbulence,light was subject to various interferences when transmitting underwater,and the worse the water environment,the more complex the interferences.The channel equalisation technology was introduced in the underwater optical communication system with asymmetric limited optical orthogonal time-frequency-space modulation,and an equalisation method based on convolutional neural network was proposed according to the transmission characteristics of underwater light.The simulation was carried out under the seawater light weak turbulence channel that followed the Gamma-Gamma distribution function.To verify the validity of the simulation,the Mont-Carlo ray-tracing simulation was adopted to determine the actual channel characteristics.It was concluded that the impulse response of the fast Fourier transform of the actual channel was excellently fitted with the Gamma-Gamma function distribution.At the receiving end,the received optical signal was first converted into an electrical signal,and the bit error rate before equalisation under different intensities of turbulence was simulated.Furthermore,under the condition of weak turbulence,the performance of the CNN equalisation algorithm,the BP equalisation algorithm and the unbalanced algorithm was compared.The simulation results showed that,compared with the BP algorithm,the bit error performance has been improved.Therefore,in the case of poor signal-to-noise ratio,the proposed method can effectively improve the communication performance of underwater optical communication systems,reduce the bit error rate,and enhance the reliability of channel estimation.
作者
丛艳平
张甜甜
封斌
曹明华
CONG Yan-ping;ZHANG Tian-tian;FENG Bin;CAO Ming-hua(School of Artificial Intelligence,Guangzhou Maritime University,Guangzhou Guangdong 510725,China;School of Computer Science and Artificial Intelligence,Lanzhou University of Technology,Lanzhou Gansu 730050,China;Center for Networking and Educational Technology,Guangzhou Maritime University,Guangzhou Guangdong 510725,China)
出处
《广州航海学院学报》
2025年第3期31-36,共6页
Journal of Guangzhou Maritime University
基金
国家自然科学基金(62265010)
甘肃省自然科学基金(24JRRA183)。
关键词
通信技术
水下无线光通信
非对称限幅的光正交时频率空
卷积神经网络
信道均衡
communication technology
Underwater Wireless Optical Communication
Asymmetrically Clipped Optical Orthogonal Time Frequency Space
Convolutional Neural Network
channel equalisation
