In recent years, the large increase in connected devices and the data that are collected by these devices have caused a heightened interest in distributed processing. Many practical distributed networks are of heterogeneous nature, because different devices in the network can have different specifications. Because of this, it is highly desirable that algorithms operating within these networks can operate asynchronously, since in that case there is no need for clock synchronisation between the nodes, and the algorithm is not slowed down by the slowest device in the network. In this paper, we focus on the primal-dual method of multipliers (PDMM), which is a promising distributed optimisation algorithm that is suitable for distributed optimisation in heterogeneous networks. Most theoretical work that can be found in existing literature focuses on synchronous versions of PDMM. In this work, we prove the convergence of stochastic PDMM, which is a general framework that can model variations such as asynchronous PDMM and PDMM with transmission losses.

This paper describes a data-driven approach to creating real-time neural network models of guitar amplifiers, recreating the amplifiers’ sonic response to arbitrary inputs at the full range of controls present on the physical device. While the focus on the paper is on the data collection pipeline, we demonstrate the effectiveness of this conditioned black-box approach by training an LSTM model to the task, and comparing its performance to an offline white-box SPICE circuit simulation. Our listening test results demonstrate that the neural amplifier modeling approach can match the subjective performance of a high-quality SPICE model, all while using an automated, non-intrusive data collection process, and an end-to-end trainable, real-time feasible neural network model.

Virtual analog (VA) modeling using neural networks (NNs) has great potential for rapidly producing high-fidelity models. Recurrent neural networks (RNNs) are especially appealing for VA due to their connection with discrete nodal analysis. Furthermore, VA models based on NNs can be trained efficiently by directly exposing them to the circuit states in a gray-box fashion. However, exposure to ground truth information during training can leave the models susceptible to error accumulation in a free-running mode, also known as “exposure bias” in machine learning literature. This paper presents a unified framework for treating the previously proposed state trajectory network (STN) and gated recurrent unit (GRU) networks as special cases of discrete nodal analysis. We propose a novel circuit state-matching mechanism for the GRU and experimentally compare the previously mentioned networks for their performance in state matching, during training, and in ex-posure bias, during inference. Experimental results from modeling a diode clipper show that all the tested models exhibit some exposure bias, which can be mitigated by truncated backpropagation through time. Furthermore, the proposed state matching mechanism improves the GRU modeling performance of an overdrive pedal and a phaser pedal, especially in the presence of external modulation, apparent in a phaser circuit.

In this paper the effects of quantisation on distributed convex optimisation algorithms are explored via the lens of monotone operator theory. Specifically, by representing transmission quantisation via an additive noise model, we demonstrate how quantisation can be viewed as an instance of an inexact Krasnosel' skiľ-Mann scheme. In the case of two distributed solvers, the Alternating Direction Method of Multipliers and the Primal Dual Method of Multipliers, we further demonstrate how an adaptive quantisation scheme can be constructed to reduce transmission costs between nodes. Finally for the Gaussian channel capacity maximisation problem, we demonstrate convergence even in the presence of one-bit uniform quantisation based on the aforementioned adaptive quantisation scheme.

We propose a new robust distributed linearly constrained beamformer that utilizes a set of linear equality constraints to reduce the cross power spectral density matrix to a block-diagonal form. The proposed beamformer has a convenient objective function for use in arbitrary distributed network topologies while having identical performance to a centralized implementation. Moreover, the new optimization problem is robust to relative acoustic transfer function (RATF) estimation errors and to target activity detection (TAD) errors. Two variants of the proposed beamformer are presented and evaluated in the context of multimicrophone speech enhancement in a wireless acoustic sensor network, and are compared with other state-of-the-art distributed beamformers in terms of communication costs and robustness to RATF estimation errors and TAD errors.

Large-scale networks of computing units, often characterised by the absence of central control, have become commonplace in many applications. To facilitate data processing in these large-scale networks, distributed signal processing is required. The iterative behaviour of distributed processing algorithms combined with energy, computational power, and bandwidth limitations imposed by such networks, place tight constraints on the transmission capacities of the individual nodes. In this paper we investigate the effects of subtractive dithered uniform quantisation in PDMM for the synchronous distributed averaging problem. This is done by deriving expressions for the mean squared error (MSE) that include quantisation noise. Also, the required data rate for quantised PDMM is considered. It was found that for practical applications quantisation in PDMM can be applied with a fixed-rate quantiser, such that significant data rate reduction can be achieved, without compromising the rate of convergence.

In this paper we propose a distributed reformulation of the linearly constrained minimum variance (LCMV) beamformer for use in acoustic wireless sensor networks. The proposed distributed minimum variance (DMV) algorithm, for which we demonstrate implementations for both cyclic and acyclic networks, allows the optimal beamformer output to be computed at each node without the need for sharing raw data within the network. By exploiting the low rank structure of estimated covariance matrices in time-varying noise fields, the algorithm can also provide a reduction in the total amount of data transmitted during computation when compared to centralised solutions. This is particularly true when multiple microphones are used per node. We also compare the performance of DMV with state of the art distributed beamformers and demonstrate that it achieves greater improvements in SNR in dynamic noise fields with similar transmission costs.