Wireless Acoustic Sensor Research Articles

Environmental noise monitoring using distributed hierarchical wireless acoustic sensor network

Acoustic noise pollution is one of many problems people face as cities grow. Long-term noise exposure can result in a series of physical and mental health diseases that are highly harmful to foetuses and newborns. Hence, many IoT-based wireless sensor network systems have been proposed for automated monitoring for long-term operation. However, these systems suffer from weaknesses in functionality, power consumption, costs, and scalability, which hinder large-scale deployment. In this study, we propose a distributed hierarchical wireless acoustic sensor network for environmental noise monitoring to do sound classification and A-weighted sound-pressure-level measurement to address the shortcomings of existing systems. A series of tests and comparisons are performed in diagnosing the performance with respect to recording continuity, packet loss, recording quality, accuracy on A-weighted sound pressure level calculations, and costs. Results show that this proposed network structure is feasible as a part of hardware implementation in a large-scale, low-cost, and high-scalable environmental noise monitoring system to classify sound.

Research on a Method for Classifying Bolt Corrosion Based on an Acoustic Emission Sensor System.

High-strength bolts play a crucial role in ultra-high-pressure equipment such as bridges and railway tracks. Effective monitoring of bolt conditions is of paramount importance for common fault repair and accident prevention. This paper aims to detect and classify bolt corrosion levels accurately. We design and implement a bolt corrosion classification system based on a Wireless Acoustic Emission Sensor Network (WASN). Initially, WASN nodes collect high-speed acoustic emission (AE) signals from bolts. Then, the ReliefF feature selection algorithm is applied to identify the optimal feature combination. Subsequently, the Extreme Learning Machine (ELM) model is utilized for bolt corrosion classification. Additionally, to achieve high prediction accuracy, an improved goose algorithm (GOOSE) is employed to ensure the most suitable parameter combination for the ELM model. Experimental measurements were conducted on five classes of bolt corrosion levels: 0%, 25%, 50%, 75%, and 100%. The classification accuracy obtained using the proposed method was at least 98.04%. Compared to state-of-the-art classification diagnostic models, our approach exhibits superior AE signal recognition performance and stronger generalization ability to adapt to variations in working conditions.

Distributed energy-saving speech enhancement in wireless acoustic sensor networks

Wireless acoustic sensor network (WASN) has shown great potential for speech enhancement compared with traditional microphone arrays. As dealing with all sensor measurements demands a powerful centralized processor, a distributed framework for speech enhancement is developed in this work. We propose to suppress noises by jointly exploiting the temporal and spatial information of audio signals, based on the information weighted consensus (IWC) filtering. In order to reduce the computational cost as well as the transmission power, we activate only an informative subnetwork by maximizing the energy efficiency in the WASN, where the subnetwork connectivity is guaranteed by incorporating more constraints. To select sensors rapidly, we propose a near-optimal greedy method, which removes one sensor at each iteration until the subnetwork size equals a preset value. The proposed method can obtain the high-quality target signal in a distributed and energy-efficient way. Simulation and real-world experiment results confirm its validity.

M-estimate based diffusion active noise control algorithm over distributed networks and its performance analysis

Multi-channel active noise control (ANC) systems can achieve effective attenuation of low-frequency noise in spatial areas and have been extensively studied. Recent works have focused on the application of multitask diffusion strategies in ANC systems based on wireless acoustic sensor and actuator networks (WASAN), which show the advantages of saving computational burden and enhancing flexibility. However, these works all adopt the mean-square error minimization as the optimization criterion, which leads to severe performance degradation in the presence of impulsive noise interference. Therefore, in this paper, we exploit the M-estimate function and the augmented weight vector of the control filter to formulate a robust minimization problem and propose a distributed augmented diffusion filtered-x least mean M-estimate (DADFxLMM) algorithm for multi-channel ANC systems. The proposed algorithm leverages the M-estimate function to overcome the influence of outliers in the error signal and achieves robustness against impulsive noise, and also eliminates the dependence on the symmetry of the acoustic paths by integrating the augmented weight vectors. Moreover, we replace the score function with the update probability of the weight vector to avoid the need for integration and Price’s theorem, and analyze the mean and mean-square performance of the proposed DADFxLMM algorithm under the contaminated Gaussian (CG) noise model. Simulation results under different system configurations demonstrate that the proposed algorithm outperforms the existing multi-channel ANC algorithms in the presence of impulsive noise interference.

Acoustic Comfort Prediction: Integrating Sound Event Detection and Noise Levels from a Wireless Acoustic Sensor Network.

There is an increasing interest in accurately evaluating urban soundscapes to reflect citizens' subjective perceptions of acoustic comfort. Various indices have been proposed in the literature to achieve this purpose. However, many of these methods necessitate specialized equipment or extensive data collection. This study introduces an enhanced predictor for dwelling acoustic comfort, utilizing cost-effective data consisting of a 30-s audio clip and location information. The proposed predictor incorporates two rating systems: a binary evaluation and an acoustic comfort index called ACI. The training and evaluation data are obtained from the "Sons al Balcó" citizen science project. To characterize the sound events, gammatone cepstral coefficients are used for automatic sound event detection with a convolutional neural network. To enhance the predictor's performance, this study proposes incorporating objective noise levels from public IoT-based wireless acoustic sensor networks, particularly in densely populated areas like Barcelona. The results indicate that adding noise levels from a public network successfully enhances the accuracy of the acoustic comfort prediction for both rating systems, reaching up to 85% accuracy.

A geometric solution to microphone array position calibration and its confidence interval analysis

Wireless acoustic sensor networks (WASNs) comprised of spatially distributed acoustic nodes have been increasingly popular in many areas such as speech enhancement, blind source separation, and audio surveillance, where node position information is generally required to implement relevant processing algorithms. In this article, we focus on the node position calibration only using direction-of-arrival (DOA) information with each node consisting of a microphone array. State-of-the-art techniques translate the position calibration into optimization of DOA-based cost functions, and have shown notable success. However, a compromise between position calibration accuracy and computational load still exists. Furthermore, none of the existing studies can offer a quantitative evaluation of calibration performance before the optimization procedure is finished. To address these problems, a geometric solution to array node position calibration is proposed using DOA measurements. For a given array node, DOAs of three sources are first utilized to construct cylinder, plane, and cone equations. Next, the two-dimensional projection of intersections among these equations is calculated using Sylvesterresultantmatrix. Finally, a closed-form solution of node position in three-dimensional space is established. Moreover, explicit mathematical derivation of the confidence interval analysis on the calibration results under given node angle estimation error condition is presented, indicating that the proposed method can provide predictable performance with corresponding probabilities. The Cramér-Rao bound (CRB) is further investigated. Simulation results reveal that as compared with existing methods our approach achieves comparable calibration accuracy with much less computational burden. Real-world experiments also verify its effectiveness.

A long-term global and clustering analysis of the COVID-19 pandemic effects on the sound environment of Milan, Italy: Road traffic and other noise sources

The different measures taken by the competent authorities to tackle the effects of the COVID-19 pandemic affected the acoustic environment worldwide, resulting in significant reductions of noise levels. However, most of the investigations mainly focused on the lockdown during spring 2020, in which all non-essential industrial and commercial activities were prohibited, being people requested to stay at home. The analysis of the data collected and processed by the wireless acoustic sensor network deployed in Milan, Italy, by the DYNAMAP project showed a significant drop of sound levels, mainly due to road traffic as well as other noise sources, denoted as anomalous noise events (ANEs). This work extends that research through a long-term global analysis of the COVID-19 pandemic effects from January 2020 to August 2021, including data clustering to identify local patterns of the sound environment of Milan. As a result of the first hard and subsequent second soft lockdown imposed by the Italian authorities, a double V-shape reduction pattern of the mean sound levels Lden (with and without ANEs) is observed, with maximum drops of around 6 dB and 3 dB, respectively, surrounded with similar progressive reductions in both pre- and post-lockdown periods, showing statistically significant differences. A complementary pattern is obtained regarding the detection of ANEs. The clustering of Lden indicators yields three different sound patterns in the second lockdown, both in terms of noise levels and ANEs, depending on sensor locations and traffic flow variations, unlike the first lockdown with a homogeneous behaviour throughout the city.

A Systematic Analysis, Outstanding Challenges, and Future Prospects for Routing Protocols and Machine Learning Algorithms in Underwater Wireless Acoustic Sensor Networks

Water has covered a wide part of the earth’s surface. Oceans and other water bodies contain significant natural and environmental resources as well as aquatic life. Due to humans’ hazardous and unsuitable underwater (UW) settings, these are generally undiscovered and unknown. As a result of its widespread utility in fields as diverse as oceanography, ecology, seismology, and oceanography, underwater wireless sensor networks (UWSNs) have emerged as a cutting-edge area of study. Despite their usefulness, the performance of the network is hampered by factors including excessive propagation delay, a changing network architecture, a lack of bandwidth, and a battery life that is too short on sensor nodes. Developing effective routing protocols is the best way to overcome these challenges. An effective routing protocol can relay data from the network’s root node to its final destination. Therefore, the state of the art in underwater wireless acoustic sensor network (UWASN) routing protocols is assessed with an eye toward their potential for development. In real-world applications, sensor node positions are frequently used to locate relevant information. As a result, it is crucial to conduct research on routing protocols. Reinforcement learning (RL) algorithms have the ability to enhance routing under a variety of conditions because they are experience-based learning algorithms. Underwater routing methods for UWSN are reviewed in detail, including those that rely on machine learning (ML), energy, clustering and evolutionary approaches. Tables are incorporated for the suggested protocols by including the benefits, drawbacks, and performance assessments, which make the information easier to digest. Also, several applications of UWSN are discussed with security considerations. In addition to this, the analysis of node deployment and residual energy is discussed in this review. Furthermore, the domain review emphasizes UW routing protocol research difficulties and future directions, which can help researchers create more efficient routing protocols based on ML in the future.

Online distributed waveform-synchronization for acoustic sensor networks with dynamic topology

Acoustic sensing by multiple devices connected in a wireless acoustic sensor network (WASN) creates new opportunities for multichannel signal processing. However, the autonomy of agents in such a network still necessitates the alignment of sensor signals to a common sampling rate. It has been demonstrated that waveform-based estimation of sampling rate offset (SRO) between any node pair can be retrieved from asynchronous signals already exchanged in the network, but connected online operation for network-wide distributed sampling-time synchronization still presents an open research task. This is especially true if the WASN experiences topology changes due to failure or appearance of nodes or connections. In this work, we rely on an online waveform-based closed-loop SRO estimation and compensation unit for nodes pairs. For WASNs hierarchically organized as a directed minimum spanning tree (MST), it is then shown how local synchronization propagates network-wide from the root node to the leaves. Moreover, we propose a network protocol for sustaining an existing network-wide synchronization in case of local topology changes. In doing so, the dynamic WASN maintains the MST topology after reorganization to support continued operation with minimum node distances. Experimental evaluation in a simulated apartment with several rooms proves the ability of our methods to reach and sustain accurate SRO estimation and compensation in dynamic WASNs.

Pilot Agent-Driven Wireless Acoustic Sensor Network for Uninterrupted Data Transmission

Wireless acoustic sensor networks (WASN) usually demand uninterrupted and reliable data transmissions and an efficient path from the source nodes to the destination nodes, thus ensuring reliable delivery of sensitive or critical data with the use of multipath routing protocols. This paper presents a novel agent-driven WASN relying on a set of static and mobile agents. In the proposed technique, the location address is functioned into the operating node and the destination node. Calculations of the midpoint between the operating and the destination nodes are followed by identifying the shortest path based on a reference axis, using a mobile agent and a location address. Such an approach aims to ensure maximum utilization of the communicating paths. Based on the information concerning partial topology of the network, the destination node computes the weight factor and multiple paths using node distances, energy ratios and efficiencies of specific links. Then, the destination node selects the appropriate path from the multiple paths available, to transmit the data. This article also analyzes the performance of the proposed system with various parameters taken into consideration and shows its efficiency in comparison to other existing techniques.

Closed‐form solution for scaling a wireless acoustic sensor network

AbstractA closed‐form solution for localising and synchronising an acoustic sensor node with respect to a Wireless Acoustic Sensor Network (WASN) is presented. The aim is to allow efficient scaling of a WASN by individually calibrating newly joined sensor nodes instead of recalibrating the entire array. A key contribution is that the sensor to be calibrated does not need to include a built‐in emitter. The proposed method uses signals emitted from spatially distributed sources to compute time difference of arrival (TDOA) measurements between the existing WASN and a new sensor. The problem is then modelled as a set of multivariate non‐linear TDOA equations. Through a simple transformation, the non‐linear TDOA equations are converted into a system of linear equations. Then, weighted least squares is applied to find an accurate estimate of the calibration parameters. Signal sources can either be known emitters within the existing WASN or arbitrary sources in the environment, thus allowing for flexible applicability in both active and passive calibration scenarios. Simulation results under various conditions show high joint localisation and synchronisation performance, often compared to the Cramér‐Rao lower bound.

Automatic Heart Rate Detection during Sleep Using Tracheal Audio Recordings from Wireless Acoustic Sensor.

Heart rate is an essential diagnostic parameter indicating a patient's condition. The assessment of heart rate is also a crucial parameter in the diagnostics of various sleep disorders, including sleep apnoea, as well as sleep/wake pattern analysis. It is usually measured using an electrocardiograph (ECG)-a device monitoring the electrical activity of the heart using several electrodes attached to a patient's upper body-or photoplethysmography (PPG). The following paper investigates an alternative method for heart rate detection and monitoring that operates on tracheal audio recordings. Datasets for this research were obtained from six participants along with ECG Holter (for validation), as well as from fifty participants undergoing a full night polysomnography testing, during which both heart rate measurements and audio recordings were acquired. The presented method implements a digital filtering and peak detection algorithm applied to audio recordings obtained with a wireless sensor using a contact microphone attached in the suprasternal notch. The system was validated using ECG Holter data, achieving over 92% accuracy. Furthermore, the proposed algorithm was evaluated against whole-night polysomnography-derived HR using Bland-Altman's plots and Pearson's Correlation Coefficient, reaching the average of 0.82 (0.93 maximum) with 0 BPM error tolerance and 0.89 (0.97 maximum) at ±3 BPM. The results prove that the proposed system serves the purpose of a precise heart rate monitoring tool that can conveniently assess HR during sleep as a part of a home-based sleep disorder diagnostics process.

Improved wireless acoustic sensor network for analysing audio properties

Improved wireless acoustic sensor network for analysing audio properties

A Non-Destructive Measurement of Trunk Moisture Content in Living Trees Based on Multi-Sensory Data Fusion

Water plays an important role in various physiological activities of living trees. Measuring trunk moisture content (MC) in real-time without damage has important guiding significance for transpiration research in forest ecosystems. However, existing standing tree MC detection methods are either too cumbersome to install or cause different degrees of damage. Here, we propose a novel Internet of Things (IoT) monitoring system that includes wireless acoustic emission sensor nodes (WASNs) and underground soil MC sensor nodes to efficiently detect and diagnose the MC level of living tree trunks. After the characteristic parameters were collected by the two sensors, a feature selection and multi-sensory global fusion method for MC diagnosis was designed and developed and several statistical parameters were selected as the input variables to predict the heartwood MC level with a support vector machine (SVM) model. Moreover, to achieve the highest prediction accuracy, an improved sparrow search algorithm (ISSA) is applied to ensure the most suitable parameter combinations in a two-objective optimization model. Extensive experiments result in a fusion of the environment, and AE signals show that the proposed mechanism has better diagnostic performance than state-of-the-art methods and is more adaptable to the fluctuation of working conditions.

Parallel processing of distributed beamforming and multichannel linear prediction for speech denoising and deverberation in wireless acoustic sensor networks

More and more smart home devices with microphones come into our life in these years; it is highly desirable to connect these microphones as wireless acoustic sensor networks (WASNs) so that these devices can be better controlled in an enclosure. For indoor applications, both environmental noise and room reverberation may severely degrade speech quality, and thus both of them need to be removed to improve users’ experience. For this goal, this paper proposes a parallel processing framework of distributed beamforming and multichannel linear prediction (DB-BFMCLP), which consists of generalized sidelobe canceler and multichannel linear prediction for simultaneous speech dereverberation and noise reduction in WASNs. By sharing a common desired response vector, the proposed DB-BFMCLP can provide a significant reduction in communication bandwidth without sacrificing performance. The convergence guarantee of the DB-BFMCLP to its centralized implementation is derived mathematically. Simulation results verify the superiority of the proposed method to the existing related methods in noisy and reverberant scenarios.

Study of rail damage diagnosis and localization method based on intelligent wireless acoustic sensor network

Many major derailments are closely related to rail damage, so it is especially important to conduct regular damage detection on in-service rails. In this paper, with the main objective of nondestructive detection and localization of rail crack damage, a wireless acoustic sensor network (WASN) based rail flaw detection and localization diagnosis system is designed and implemented: firstly, the WASN nodes rapidly sample the acoustic emission signal on the rail surface, then calculate and wirelessly transmit its characteristic parameters to the gateway, and then use the proposed information wasserstein generative adversarial network-gradient penalty (InfoWGAN-GP) algorithm for data augmentation of the collected AE features, followed by feature optimization of the augmented hybrid data set based on the proposed categorical boosting-borutashap (Catboost-BorutaShap) algorithm, furthermore, the proposed improved aquila optimizer-deep extreme learning machine (IAO-DELM) and binary dwarf mongoose simulated annealing optimization-newton iteration-triangulation algorithm (BDMSAO-NI-TA) strategies are then proposed to establish accurate identification and localization of rail crack damage derivation models. The experimental comparison results show that the IAO-DELM model constructed based on the preferred feature subset has the best effect on the performance enhancement of rail crack damage diagnosis, with the accuracy, precision, F1-score, weighted average and AUC of 99.86%, 99.58%, 98.75%, 0.9962 and 0.9923, respectively, which are higher than those of the optimization algorithm such as honey badger combined with DELM model. The evaluation indexes indicate that this method has excellent monitoring effectiveness.

On the Challenges of Acoustic Energy Mapping Using a WASN: Synchronization and Audio Capture

Acoustic energy mapping provides the functionality to obtain characteristics of acoustic sources, as: presence, localization, type and trajectory of sound sources. Several beamforming-based techniques can be used for this purpose. However, they rely on the difference of arrival times of the signal at each capture node (or microphone), so it is of major importance to have synchronized multi-channel recordings. A Wireless Acoustic Sensor Network (WASN) can be very practical to install when used for mapping the acoustic energy of a given acoustic environment. However, they are known for having low synchronization between the recordings from each node. The objective of this paper is to characterize the impact of current popular synchronization methodologies as part of the WASN to capture reliable data to be used for acoustic energy mapping. The two evaluated synchronization protocols are: Network Time Protocol (NTP) y Precision Time Protocol (PTP). Additionally, three different audio capture methodologies were proposed for the WASN to capture the acoustic signal: two of them, recording the data locally and one sending the data through a local wireless network. As a real-life evaluation scenario, a WASN was built using nodes conformed by a Raspberry Pi 4B+ with a single MEMS microphone. Experimental results demonstrate that the most reliable methodology is using the PTP synchronization protocol and audio recording locally.

Reinforcement learning for autonomous vehicle movements in wireless multimedia applications

We develop a Deep Reinforcement Learning (DeepRL)-based, multi-agent algorithm to efficiently control autonomous vehicles that are typically used within the context of Wireless Sensor Networks (WSNs), in order to boost application performance. As an application example, we consider wireless acoustic sensor networks where a group of speakers move inside a room. In a traditional setup, microphones cannot move autonomously and are, e.g., located at fixed positions. We claim that autonomously moving microphones improve the application performance. To control these movements, we compare simple greedy heuristics against a DeepRL solution and show that the latter achieves best application performance.As the range of audio applications is broad and each has its own (subjective) performance metric, we replace those application metrics by two immediately observable ones: First, quality of information (QoI), which is used to measure the quality of sensed data (e.g., audio signal strength). Second, quality of service (QoS), which is used to measure the network’s performance when forwarding data (e.g., delay). In this context, we propose two multi-agent solutions (where one agent controls one microphone) and show that they perform similarly to a single-agent solution (where one agent controls all microphones and has a global knowledge). Moreover, we show via simulations and theoretical analysis how other parameters such as the number of microphones and their speed impacts performance.

Intelligent Microsystem for Sound Event Recognition in Edge Computing Using End-to-End Mesh Networking

Wireless acoustic sensor networks (WASNs) and intelligent microsystems are crucial components of the Internet of Things (IoT) ecosystem. In various IoT applications, small, lightweight, and low-power microsystems are essential to enable autonomous edge computing and networked cooperative work. This study presents an innovative intelligent microsystem with wireless networking capabilities, sound sensing, and sound event recognition. The microsystem is designed with optimized sensing, energy supply, processing, and transceiver modules to achieve small size and low power consumption. Additionally, a low-computational sound event recognition algorithm based on a Convolutional Neural Network has been designed and integrated into the microsystem. Multiple microsystems are connected using low-power Bluetooth Mesh wireless networking technology to form a meshed WASN, which is easily accessible, flexible to expand, and straightforward to manage with smartphones. The microsystem is 7.36 cm3 in size and weighs 8 g without housing. The microsystem can accurately recognize sound events in both trained and untrained data tests, achieving an average accuracy of over 92.50% for alarm sounds above 70 dB and water flow sounds above 55 dB. The microsystems can communicate wirelessly with a direct range of 5 m. It can be applied in the field of home IoT and border security.

An integrated system for the acoustic monitoring of goat farms

Effective precision livestock farming requires the availability of reliable and up-to-date data characterizing the animals and environment of interest. This work presents the design, architecture, and implementation of a wireless acoustic sensor network for monitoring goat farms on a 24/7 basis. In addition, we define a hierarchical organization of the involved sound classes exhaustively covering every aspect of the encountered goat vocalizations. Moreover, we developed an annotation tool tailored to the specifics of the problem at hand, i.e. a big, real-world data environment, able to meaningfully assist the annotation of goat vocalizations via a suitable sound classification module. On top of that, a mobile phone application was developed facilitating authorized users to remotely access information describing the situation on the farm site. Importantly, such a non-invasive monitoring framework was installed in 4 different sites located in Northern Italy while taking into account their diverse characteristics.