Digital signal processing for underwater acoustic in-band full-duplex communication
Date
2023
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Acoustic signals are widely used for communication in the underwater environment. However, due to using narrow acoustic frequency bands, underwater transmission suffers from low data rates. One of the promising techniques to address this problem is in-band full-duplex (IBFD) communication. IBFD implies the transmission and reception at the same time and on the same frequency band. On the upside, sharing the time and frequency can potentially double the data rate. However, a large portion of the self-transmitted signal interferes with the desired signal at the receiver of IBFD. This contamination is known as self-interference (SI), which is usually stronger than the desired signal. As a result, SI cancellation is one of the main concerns in IBFD communication. Because of numerous and strong reflections, along with the dynamic condition of the water, tracking and eliminating the SI in underwater channels can be more challenging, compared to that in radio channels. ☐ Analog methods, such as physical transmitter-receiver isolation and using directional antennas, can be employed to attenuate the SI in underwater IBFD. However, it has been shown that analog methods alone cannot provide a desirable SI cancellation gain. As a result, using digital-based techniques is also required to suppress the SI residuals after analog attenuation. In this regard, in this dissertation, our concentration is on digital-based SI cancellation methods in underwater acoustic IBFD communication. ☐ We start with the statistical characterization of the SI channel. To this purpose, we use orthogonal frequency division multiplexing (OFDM) signal to characterize the SI channel in the experimental lake water. To verify the results, the SI channel estimation is performed in both the frequency and the time domains. We show that in our experiment, regardless of the depth of the hydrophone, the direct path is strong, stable, and easy to be eliminated; however, the reflection paths are weak and rapidly time-varying which can challenge the SI cancellation process. Among the reflections, the first bounce from the water surface is the prevalent path with a short coherence time of around 70 ms. ☐ At the receiver of IBFD, channel state information for both the SI and the remote transmission (RT) channels is required to remove the SI and equalize the SI-suppressed signal, respectively. However, because of the rapid time-variations of the underwater environment, real-time tracking of both channels is necessary. Therefore, we propose a receiver for underwater IBFD in which the variations of the SI and the RT channels are jointly tracked by using a recursive least squares (RLS) algorithm fed by feedback from the previously detected data symbols. We show that because of the joint channel estimation, SI cancellation is more successful compared to IBFD receivers with separate channel estimators. ☐ Nevertheless, when the SI channel is significantly stronger than that of the RT one, even a slight inaccuracy in estimating the SI channel with RLS estimator leads to large residuals after cancellation. As a result, we develop the multi-layered recursive least squares (m-RLS) algorithm and use it for joint channel estimation with enhanced accuracy in estimating the SI channel. The m-RLS estimator is composed of multiple layers, each of which employs an RLS to estimate and eliminate the SI residuals remaining from the previous layer's process. We show that when the SI is significantly stronger than the RT signal, the optimum number of layers in m-RLS becomes more than one, and the SI cancellation is performed better than that when we use RLS. ☐ Next, in order to further improvement, we develop a method that is a combination of adaptive decision feedback equalizer and SI cancellation (ADFE-SIC) to jointly eliminate the ISI and SI. An RLS algorithm is employed to adaptively estimate the filters in ADFE-SIC. By conducting simulations and experimental tests, we show that the ADFE-SIC method outperforms our previous approach in which equalization and SI cancellation tasks are performed separately.
Description
Keywords
Digital signal processing, In-band communication, Underwater acoustic communication, Full-duplex communication