Learning-based Signal Research Articles

Deep Learning-Based Signal Detection for Underwater Acoustic OTFS Communication

Orthogonal time frequency space (OTFS) is a novel two-dimensional (2D) modulation technique that provides reliable communications over time- and frequency-selective channels. In underwater acoustic (UWA) channel, the multi-path delay and Doppler shift are several magnitudes larger than wireless radio communication, which will cause severe time- and frequency-selective fading. The receiver has to recover the distorted OTFS signal with inter-symbol interference (ISI) and inter-carrier interference (ICI). The conventional UWA OTFS receivers perform channel estimation explicitly and equalization to detect transmitted symbols, which requires prior knowledge of the system. This paper proposes a deep learning-based signal detection method for UWA OTFS communication, in which the deep neural network can recover the received symbols after sufficient training. In particular, it cascades a convolutional neural network (CNN) with skip connections (SC) and a bidirectional long short-term memory (BiLSTM) network to perform signal recovery. The proposed method extracts feature information from received OTFS signal sequences and trains the neural network for signal detection. The numerical results demonstrate that the SC-CNN-BiLSTM-based OTFS detection method performs with a lower bit error rate (BER) than the 2D-CNN, FC-DNN, and conventional signal detection methods.

Multiagent Reinforcement Learning-Based Signal Planning for Resisting Congestion Attack in Green Transportation

Inefficient signal control will not only exaggerate traffic congestion, but also increase the fuel consumption and exhaust emissions. Thus, signal planning is highly important in green transportation. As the Connected vehicle (CV) technology has transformed today’s transportation systems by connecting vehicles and the transportation infrastructure through wireless communication, the CV-based signal control system has seen significant studies recently. Unfortunately, existing signal planning algorithms in use are developed for the signal-intersection, showing low traffic efficiency in the multi-intersection collaborative planning due to ignoring the traffic correlation among the neighboring intersections. In this work, we target the USDOT (U.S. Department of Transportation) sponsored CV-based traffic control system, and implement a multi-intersection traffic network. We model the multi-intersection collaborative signal planning problem as a multi-agent reinforcement learning problem, and present an actor-attention-critic algorithm to improve transportation efficiency and energy efficiency in green transportation, as well as resist congestion attack. Experiment results on the multi-intersection traffic network indicates that 1) compared to the baseline, our approach reduces the total delay by as high as 44.24%; 2) our method transports more vehicles passing the intersections meanwhile reduces the total CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> emissions by 2.40%; 3) under the congestion attack, our approach shows robustness and reduces the total delay by as high as 64.33%.

Deep Learning-Based Signal Quality Assessment for Wearable ECGs

Nowadays, use of the dynamic electrocardiogram (ECG) has developed rapidly because of the wide application of wearable devices [1]–[3]. Most ECG-based diagnostic algorithms require that the ECG signal have a clear waveform and accurate feature points. However, the collected wearable ECG signal usually contains a certain amount of noise and causes many false alarms in the ECG analysis system [4], [5]. Thus, signal quality assessment (SQA) plays a prominent role in ruling out the ECG segments with poor signal quality [6]. Compared with traditional static ECG signals, dynamic wearable ECGs contain more noise, which brings greater challenges to disease detection algorithms [7]–[9]. These artifacts and noises in dynamic ECG signals can seriously affect the R-peaks detection, ECG beat extraction, ECG morphological feature extraction and the detection of noise peaks, resulting in frequent false alarms [10]. In 2008, Li et al. [11] proposed the bSQI signal quality indexes: comparison of two beat detectors on a single ECG lead. Liu et al. [12] generalized the two QRS wave complex (QRS) detectors-based bSQI to multiple QRS detectors-based bSQI (GbSQI) to improve the SQA performance. Liu et al. [8] proposed an efficient real-time SQA method for healthy subjects.

Chain-structure time-delay reservoir computing for synchronizing chaotic signal and an application to secure communication

In this work, a chain-structure time-delay reservoir (CSTDR) computing, as a new kind of machine learning-based recurrent neural network, is proposed for synchronizing chaotic signals. Compared with the single time-delay reservoir, our proposed CSTDR computing shows excellent performance in synchronizing chaotic signal achieving an order of magnitude higher accuracy. Noise consideration and optimal parameter setting of the model are discussed. Taking the CSTDR computing as the core, a novel scheme of secure communication is further designed, in which the “smart” receiver is different from the traditional in that it can synchronize to the chaotic signal used for encryption in an adaptive manner. The scheme can solve the issues such as design constrains for identical dynamical systems and couplings between transmitter and receiver in conventional settings. To further manifest the practical significance of the scheme, the digital implementation using field-programmable gate array is conducted and tested experimentally with real-world examples including image and video transmission. The work sheds light on developing machine learning-based signal processing and communication applications.

A versatile deep-neural-network-based music preprocessing and remixing scheme for cochlear implant listeners

While cochlear implants (CIs) have proven to restore speech perception to a remarkable extent, access to music remains difficult for most CI users. In this work, a methodology for the design of deep learning-based signal preprocessing strategies that simplify music signals and emphasize rhythmic information is proposed. It combines harmonic/percussive source separation and deep neural network (DNN) based source separation in a versatile source mixture model. Two different neural network architectures were assessed with regard to their applicability for this task. The method was evaluated with instrumental measures and in two listening experiments for both network architectures and six mixing presets. Normal-hearing subjects rated the signal quality of the processed signals compared to the original both with and without a vocoder which provides an approximation of the auditory perception in CI listeners. Four combinations of remix models and DNNs have been selected for an evaluation with vocoded signals and were all rated significantly better in comparison to the unprocessed signal. In particular, the two best-performing remix networks are promising candidates for further evaluation in CI listeners.

Distributed Deep Learning-Based Signal Classification for Time-Frequency Synchronization in Wireless Networks

Distributed Deep Learning-Based Signal Classification for Time-Frequency Synchronization in Wireless Networks

Traffic Signal Optimization for Multiple Intersections Based on Reinforcement Learning

In order to deal with dynamic traffic flow, adaptive traffic signal controls using reinforcement learning are being studied. However, most of the related studies are difficult to apply to the real field considering only mathematical optimization. In this study, we propose a reinforcement learning-based signal optimization model with constraints. The proposed model maintains the sequence of typical signal phases and considers the minimum green time. The model was trained using Simulation of Urban MObility (SUMO), a microscopic traffic simulator. The model was evaluated in the virtual environment similar to a real road with multiple intersections connected. The performance of the proposed model was analyzed by comparing the delay and number of stops with a reinforcement learning model that did not consider constraints and a fixed-time model. In a peak hour, the proposed model reduced the delay from 3 min 15 s to 2 min 15 s and the number of stops from 11 to 4.7 compared to the fixed-time model.

Two-Dimensional Convolutional Neural Network-Based Signal Detection for OTFS Systems

Orthogonal time frequency space (OTFS) modulation is a newly proposed modulation technique for providing a solution to high mobility doubly dispersive channel problems. In several recent research works, it is shown that OTFS has better performance over the existing conventional multicarrier modulations. OTFS modulate information symbols in a two-dimensional (2D) delay-Doppler domain rather than in time frequency domain, which can exploit the full channel diversity over time and frequency. This unique ability of OTFS can provide to design an advanced signal detection method. In this letter, we present a deep learning-based signal detection for OTFS systems. Since the input-output relation of OTFS is in 2D delay-Doppler domain, we propose a two-dimensional convolutional neural network (2D-CNN) based detector. We also employ data augmentation technique based on the widely used message-passing (MP) algorithm to improve learning ability of the proposed method. Simulation results show that the proposed method has an improved performance over the MP detector and achieves nearly the same performance as an optimal maximum <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">a posteriori</i> (MAP) detector with a very low time complexity.

Deep learning-augmented radioluminescence imaging for radiotherapy dose verification.

We developed a novel dose verification method using a camera-based radioluminescence imaging system (CRIS) combined with a deep learning-based signal processing technique. The CRIS consists of a cylindrical chamber coated with scintillator material on the inner surface of the cylinder, coupled with a hemispherical mirror and a digital camera at the two ends. After training, the deep learning model is used for image-to-dose conversion to provide absolute dose prediction at multiple depths of a specific water phantom from a single CRIS image under the assumption of a good consistency between the TPS setting and actual beam energy. The model was trained using a set of captured radioluminescence images and the corresponding dose maps from the clinical treatment planning system (TPS) for the sake of acceptable data collection. To overcome the latent error and inconsistency that exists between the TPS calculation and the corresponding measurement, the model was trained in an unsupervised manner. Validation experiments were performed on five square fields (ranging from 2 × 2 to 10 × 10cm2 ) and three clinical intensity-modulated radiation therapy (IMRT) cases. The results were compared to the TPS calculations in terms of gamma index at 1.5, 5, and 10cm depths. The mean 2%/2mm gamma pass rates were 100% for square fields and 97.2% (range from 95.5% to 99.5%) for the IMRT fields. Further validations were performed by comparing the CRIS results with measurements on various regular fields. The results show a mean gamma pass rate of 91% (1%/1mm) for cross-profiles and a mean percentage deviation of 1.15% for percentage depth doses (PDDs). The system is capable of converting the irradiated radioluminescence image to corresponding water-based dose maps at multiple depths with a spatial resolution comparable to the TPS calculations.

Deep learning-enabled EPID-based 3D dosimetry for dose verification of step-and-shoot radiotherapy.

The study aims at a novel dosimetry methodology to reconstruct a 3D dose distribution as imparted to a virtual cylindrical phantom using an electronic portal imaging device (EPID). A deep learning-based signal processing strategy, referred to as 3DosiNet, is utilized to learn a mapping from an EPID image to planar dose distributions at given depths. The network was trained with the volumetric dose exported from the clinical treatment planning system (TPS). Given the latent inconsistency between measurements and corresponding TPS calculations, unsupervised learning is formulated in 3DosiNet to capture abstractive image features that are less sensitive to the potential variations. Validation experiments were performed using five regular fields and three clinical intensity-modulated radiation therapy (IMRT) cases. The measured dose profiles and percentage depth dose (PDD) curves were compared with those measured using standard tools in terms of the 1D gamma index. The mean gamma pass rates (2%/2mm) over the regular fields are 100% and 97.3% for the dose profile and PDD measurements, respectively. The measured volumetric dose was compared to the corresponding TPS calculation in terms of the 3D gamma index. The mean 2%/2mm gamma pass rates are 97.9% for square fields and 94.9% for the IMRT fields. The system promises to be a practical 3D dosimetric tool for pre-treatment patient-specific quality assurance and further developed for in-treatment patient dose monitoring.

현실 제약 조건을 반영한 강화학습 기반 교통 신호 제어

교통 신호 제어는 도로의 한정된 용량을 효율적으로 사용하기 위해 중요한 역할을 한다. 전통적인 교통 신호 제어 방법은 미리 설정된 신호를 기반으로 운영되기 때문에 수시로 변하는 교통 상황에 대응하기 어렵다. 최근에 복잡한 문제를 해결하기 위한 방법으로 강화학습이 주목을 받으면서 효율적인 교통 신호 제어를 위해 강화학습을 사용하는 연구가 진행되고 있다. 전통적인 방법과 비교하여 대기 시간이나 여행 시간이 개선되는 것을 시뮬레이션을 통해 입증하였으나 대부분의 연구가 실제 신호의 제약조건을 반영하지 않아 현실에 적용하기에는 부적합하게 설계되어 있다. 본 논문에서는 현실에 존재하는 신호 운영체계의 제약조건을 반영하여 비교적 현실 적용 가능하며 교통량 변화에 대응할 수 있는 강화학습 기반의 신호 제어 방법을 제안한다.

Deep learning-based signal processing for evaluating energy dispersal in bridge structures

Deep learning-based signal processing for evaluating energy dispersal in bridge structures

Machine Learning-based Signal Detection for PMH Signals in Load-modulated MIMO Systems

Phase Modulation on the Hypersphere (PMH) is a power efficient modulation scheme for the load-modulated multiple-input multiple-output (MIMO) transmitters with central power amplifiers (CPA). However, it is difficult to obtain the precise channel state information (CSI), and the traditional optimal maximum likelihood (ML) detection scheme incurs high complexity which increases exponentially with the number of transmitting antennas and the number of bits carried per antenna in the PMH modulation. To detect the PMH signals without knowing the prior CSI, we first propose a signal detection scheme, termed as the hypersphere clustering scheme based on the expectation maximization (EM) algorithm with maximum likelihood detection (HEM-ML). By leveraging machine learning, the proposed detection scheme can accurately obtain information of the channel from a few of the received symbols with little resource cost and achieve comparable detection results as that of the optimal ML detector. To further reduce the computational complexity in the ML detection in HEM-ML, we also propose the second signal detection scheme, termed as the hypersphere clustering scheme based on the EM algorithm with KD-tree detection (HEM-KD). The CSI obtained from the EM algorithm is used to build a spatial KD-tree receiver codebook and the signal detection problem can be transformed into a nearest neighbor search (NNS) problem. The detection complexity of HEM-KD is significantly reduced without any detection performance loss as compared to HEM-ML. Extensive simulation results verify the effectiveness of our proposed detection schemes.

Deep neural network-based target separation from mixed micro-Doppler signature of multiple moving targets

Under radar observation, every moving target contains a unique micro-Doppler (m-D) signature due to its structural component movement, which has been considered as one of the target class characteristics in their identification. It becomes challenging to identify the type of target and its motion parameter analysis for the simultaneous presence of multiple moving targets in the radar observation channel. This article aims to separate the desired target’s micro-Doppler (m-D) signature from the responses of multiple moving targets captured in a single channel radar with the help of a machine learning-based signal processing approach. A regressive deep neural network is designed to produce the probabilistic time-frequency mask, which separates the desired target time-frequency signature from the mixed m-D responses of more targets from different classes. The proposed method provides a high correlation of separated movement signature in comparison to the clean condition micro-Doppler signatures.

Quantifying the impact of non-stationarity in reinforcement learning-based traffic signal control.

In reinforcement learning (RL), dealing with non-stationarity is a challenging issue. However, some domains such as traffic optimization are inherently non-stationary. Causes for and effects of this are manifold. In particular, when dealing with traffic signal controls, addressing non-stationarity is key since traffic conditions change over time and as a function of traffic control decisions taken in other parts of a network. In this paper we analyze the effects that different sources of non-stationarity have in a network of traffic signals, in which each signal is modeled as a learning agent. More precisely, we study both the effects of changing the context in which an agent learns (e.g., a change in flow rates experienced by it), as well as the effects of reducing agent observability of the true environment state. Partial observability may cause distinct states (in which distinct actions are optimal) to be seen as the same by the traffic signal agents. This, in turn, may lead to sub-optimal performance. We show that the lack of suitable sensors to provide a representative observation of the real state seems to affect the performance more drastically than the changes to the underlying traffic patterns.

Bearing Anomaly Recognition Using an Intelligent Digital Twin Integrated with Machine Learning

In this study, the application of an intelligent digital twin integrated with machine learning for bearing anomaly detection and crack size identification will be observed. The intelligent digital twin has two main sections: signal approximation and intelligent signal estimation. The mathematical vibration bearing signal approximation is integrated with machine learning-based signal approximation to approximate the bearing vibration signal in normal conditions. After that, the combination of the Kalman filter, high-order variable structure technique, and adaptive neural-fuzzy technique is integrated with the proposed signal approximation technique to design an intelligent digital twin. Next, the residual signals will be generated using the proposed intelligent digital twin and the original RAW signals. The machine learning approach will be integrated with the proposed intelligent digital twin for the classification of the bearing anomaly and crack sizes. The Case Western Reserve University bearing dataset is used to test the impact of the proposed scheme. Regarding the experimental results, the average accuracy for the bearing fault pattern recognition and crack size identification will be, respectively, 99.5% and 99.6%.

Signal detection with co-channel interference using deep learning

Signal detection using deep learning is a challenging and promising research topic. Several learning-based signal detectors have been proposed to produce significant results. However, most of them have ignored interference in their designs. In this paper, we evaluate the performance of learning-based signal detectors in the presence of co-channel interference under different channel conditions. Specifically, fully connected deep neural network (FCDNN) and convolutional neural network (CNN) are examined as the data-driven signal detector for blind signal detection without knowledge of the channel state information. Several important system parameters, including signal-to-interference ratio, number of interferences and type of interference, are considered. Numerical results show that FCDNN and CNN-based detectors have better performance and robustness to different SIRs conditions than traditional detectors in the presence of interference and FCDNN performs better than CNN when SIR is small and the order of interference modulation is high.

Corrections to “Learning-Based Signal Detection for Wireless OAM-MIMO Systems With Uniform Circular Array Antennas”

In the above article <xref ref-type="bibr" rid="ref1" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">[1]</xref> , the following corrections should be made:

Robust RF Mixture Signal Recognition Using Discriminative Dictionary Learning

RF signal recognition is an important element toward RF situational awareness and dynamic spectrum management. In this work, machine learning-based signal recognition algorithms are proposed. Our key contribution is to engineer feature learning such that the classifiers can perform robustly even when a mixture of heterogeneous signal classes is observed, although the training is still done using non-mixture single-label samples. To achieve this, discriminative dictionary learning algorithms are developed with various feature-shaping constraints. The signal detection can then be done in a way reminiscent of the multi-user detection in wireless communication, employing linear equalizers. The algorithms are tested using real wideband RF measurement data. It is verified that the proposed algorithms can robustly classify the component signals even when their powers are widely different and their number is not known a priori.

A Supplementary Explanation for Experimental Environment of “Implementation Methodologies of Deep Learning-Based Signal Detection for Conventional MIMO Transmitters”

To solve practical challenges for implementing deep learning-based algorithm in MIMO signal detector such as handling complex number or designing proper neural network for a specific communication system, (Baek et al., 2019) has proposed candidate implementation methodologies with simple verification experiments. According to (Baek et al., 2019), it was shown that the proposed algorithms can achieve the optimal ML performance. However, due to the lack of explanation on the experimental environment, it is difficult for readers to reproduce the presented experiments and obtain the same results. This document precisely explains on the experimental environments of (Baek et al., 2019) including the exact channel profiles.