Autoencoder with Channel Estimation for Marine Communications

Due to the special characteristics of the underwater environment such as pressure and temperature, many wireless communication technologies that can be implemented in a terrestrial environment cannot be implemented well in underwater environments. Therefore, it is very important to study the new generation of MWCNs. To solve the challenges mentioned in Sect. 1.2.2 , this chapter proposes a novel Orthogonal Frequency Division Multiplexing (OFDM) autoencoder featuring CNN-based channel estimation for marine communications with complex and fast-changing environments. We demonstrate that the proposed OFDM autoencoder system can be generalized to work under various channel environments, different throughputs, while outperform the traditional OFDM counterparts, especially when working at high throughputs. In addition, since OFDM systems require accurate channel estimations to function properly, this chapter also proposes a new channel estimation algorithm for OFDM systems that combine the power of deep learning with the philosophy of super-resolution reconstruction, which uses Dense convolutional neural Networks (Dense-Net) to reconstruct low-resolution pilot information images into high-resolution full Channel Impulse Responses (CIRs). The Dense-Net structure has the characteristics of dense connections and feature multiplexing. Simulation results show that under slow fading, the proposed channel estimator can estimate the CIRs perfectly. Under fast fading, the proposed channel estimator outperforms existing learning-based algorithms with fewer neural network parameters. Therefore, the proposed novel autoencoder scheme and the powerful channel estimator are potentially attractive approaches for MWCNs.