Identification of Non-Linear RF Systems Using Backpropagation
This work addresses the problem of efficient non-linear RF system identification for full-duplex communications, offering significant computational savings, though it is incremental as it applies an existing deep unfolding method to a specific domain.
The paper tackled the problem of identifying cascaded non-linear RF systems by viewing them as model-based neural networks using deep unfolding, enabling efficient identification with neural network tools. The result demonstrated a 74% reduction in model parameters and a 79% reduction in operations per sample for digital self-interference cancellation in full-duplex communications, achieving approximately 44.5 dB cancellation performance.
In this work, we use deep unfolding to view cascaded non-linear RF systems as model-based neural networks. This view enables the direct use of a wide range of neural network tools and optimizers to efficiently identify such cascaded models. We demonstrate the effectiveness of this approach through the example of digital self-interference cancellation in full-duplex communications where an IQ imbalance model and a non-linear PA model are cascaded in series. For a self-interference cancellation performance of approximately 44.5 dB, the number of model parameters can be reduced by 74% and the number of operations per sample can be reduced by 79% compared to an expanded linear-in-parameters polynomial model.