No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Deep Learning-Based Channel Estimation Algorithm Over Time Selective Fading Channels
Abstract
The research about deep learning application for physical layer has been received much attention in recent years. In this paper, we propose a Deep Learning (DL) based channel estimator under time varying Rayleigh fading channel. We build up, train and test the channel estimator using Neural Network (NN). The proposed DL-based estimator can dynamically track the channel status without any prior knowledge about the channel model and statistic characteristics. The simulation results show the proposed NN estimator has better Mean Square Error (MSE) performance compared with the traditional algorithms and some other DL-based architectures. Furthermore, the proposed DL-based estimator also shows its robustness with the different pilot densities.
... The channel is represented by Rayleigh fading in the proposed work, where the total number of determined pathways M is taken into account to be 20. The signal is estimated and detected by the CSI using conventional SIC techniques like LS and MMSE [38]. Additionally, ML detectors are utilized to predict signals since U u signals are given greater power [39]. ...
... For the uplink CE of LS and MMSE, pilot data transmission is employed. So the conventional LS CE of (4) can be expressed as follows [38]: ...
... In addition for estimation of MMSE, the correction coefficient R hhLS is calculated. The estimation of MMSE can be formulated as follows [38]: ...
Deep learning (DL) techniques can significantly improve successive interference cancellation (SIC) performance for the non-orthogonal multiple access (NOMA) system. The NOMA-orthogonal frequency division multiplexing (OFDM) system is considered in this paper to develop a hybrid deep neural network (HyDNN) model for multiuser uplink channel estimation (CE) and signal detection (SD). The proposed HyDNN uses a combination of a bi-directional long short-term memory (BiLSTM) network and a one-dimensional convolutional neural network (1D-CNN) to optimize errors in the system. The extraction of input signal characteristics from OFDM is carried out using the 1D-CNN model and fed into the time series BiLSTM network to infer the signal at the receiver terminal. The HyDNN model learns through the simulated channel data during offline training. To optimize the loss during learning the model the Adam optimizer is utilized. After successful training, the transmitted symbols in the online deployment are instantly recovered with optimal prediction rates by using the proposed HyDNN model. In comparison to the traditional CE
and SD method for the NOMA scheme and other existing DL models, the proposed technique demonstrates satisfactory performance enhancements. In addition, the simulation outcomes show robustness with different training parameters such as minibatch sizes and learning rates.
... In [7], [8], time-varying and doubly selective channels are generated using Jacke's model and estimated with DLbased estimators. DL-based channel estimation is used to estimate time-varying Rayleigh fading channel in [9]. Besides, underwater acoustic (UWA) multipath channel estimation with DL is presented in [10]. ...
... Moreover, different kinds of neural networks (NN) are conducted for DL-based channel estimation in the literature such as forward neural network (FNN) and recurrent neural network (RNN). It is proved that RNN gives a better performance for long input data length than other neural networks [9]. Therefore, the bidirectional long-short term memory (BLSTM) which is an RNN based was adopted in [11] for channel estimation in orthogonal frequency division multiplexing (OFDM) system. ...
... Also, J 1 and J 3 denote the first-kind Bessel function of order 1 and 3, respectively. Based on (8), (9) and (10), the channel coefficient for each k frame can be expressed as [29] h k =h b ...
... There are many excellent research works [3][4][5][6][7] and reviews [8][9][10] in the literature dealing with the channel estimation. In recent years, the research on channel estimation mainly focuses on the following aspects: channel estimation based on deep learning, 11,12 channel estimation based on iteration, [13][14][15] and channel estimation based on compressive sensing (CS). 16 For example, a novel approach to learn a low-complexity channel estimator was presented in David Neumann,12 which is motivated by the structure of the MMSE estimator; to obtain accurate estimation of the channel parameters for a fluctuant multipath channel, a novel channel estimation algorithm based on iteration is proposed in Li et al. 14 ; and the channel estimation of SC-FDE system was modeled as CS-based reconstruction of sparse signal in Si et al, 16 which match pursuit algorithms based on greedy search were adopted to reconstruct channel information. ...
... The updates of MEE-KF involve (10), (11), and the equations in Table 2, and the corresponding floating-point operations are shown in Table 3. According to Table 3, the computational complexity of MEE-KF is ...
In this paper, the channel estimation and signal‐to‐noise ratio (SNR) estimation technique of single‐carrier frequency domain equalization (SC‐FDE) system under low SNR in aeronautical multipath channel are studied, a SNR estimation algorithm which is easy to implement in engineering and an improved LS channel estimation algorithm based on Kalman filter using minimum error entropy (MEE‐KF) are proposed. This paper first introduces the SC‐FDE system and introduces the principle of MEE‐KF, and then, the channel estimation flow based on MEE‐KF is obtained by combining it with the traditional LS channel estimation algorithm, which makes the estimation results perform better. Simulation results show that after getting more accurate noise variance, the channel estimation results can better follow the changes of the channel after MEE‐KF processing, so as to resist the doppler frequency offset effect and make the channel estimation results more accurate, that is the channel response results of the data part can be closer to the real situation, so that the communication performance of SC‐FDE system has also been greatly improved.
... Additionally several well researched papers are available for CSE Using DL under diverse communication scenarios. [13][14][15][16][17][18][19][20][21] In Reference 13, the authors have proposed LSTM based channel prediction for SISO in the presence of the Doppler shift. LSTM network is used to predict the future values of channel coefficients from their past values. ...
... From the basic communication system adopted, as shown in Figure 2, the output vector y depends on the input x, channel coefficient vector h, and the noise vector w. In (18). Therefore, Equation (19) can be rewritten as ...
Channel estimation is a significant prerequisite in wireless communication, especially where the multipath radio propagation incurs significant fading in a noisy environment. In such a scenario, Rayleigh fading model is traditionally adopted to represent the communication channel. In this article, a new method of getting the Rayleigh flat channel coefficients using deep learning is presented. Here, we presume that the channel state and the corresponding channel coefficients remain constant in a given communication context which depends on the locations of transmitter/receiver, time of the day and the communication environment. When the context changes, the channel state also changes and the corresponding coefficients switch to the respective matching values. Thus the channel coefficients can have several possible realizations or classes. In our scheme, the deep learning network, after training, acts as a classifier to detect the class or the context of the channel state and based on that determines the corresponding channel coefficients. In our proposed method, the percentage reduction in the percentage error is about 26% of that of its nearest competitor when the channel SNR is 10 dB. Percentage test error vs SNR performances of CNN for channel classification (CNN‐CC), channel prediction based on feed forward artificial neural network (ANN‐CC), and the minimum mean square error method are shown in the graph. From the plots, it can be seen that CNN‐CC is better than the other two methods.
... Furthermore, the inherent block structure of the spectrum inspires the joint model of structure-based sparsity [23]- [25] and label-based sparsity [26]- [28]. Therefore, by inspecting the limitation of [16] and further investigating the spectrum sensing with supervised learning in [29]- [32], it is promising that SDL can bring extra benefits to BTF in sparse recovery, so as to promote the improvement of BCSS. ...
... As a result, cooperated with (33) and (34), (32) becomes ...
With the development of communication systems towards the high-frequency band, the demand for spectrum resources is ever-increasing, where the research interest has changed from narrowband spectrum sensing to wideband spectrum sensing. The Nyquist-rate-based wideband spectrum sensing with high-rate sampling is being questioned whether it is suitable for real-time applications. On the contrary, the well-known compressive spectrum sensing (CSS) is more appealing due to the compressive sensing (CS) technology, resulting in lower signal acquisition costs. However, the CS algorithms for recovering sparse spectrum generally require multiple iterations, which presents a challenge to the low complexity implementation of spectrum sensing. To address this issue, this paper proposes a novel method for solving the block CSS (BCSS) problem, where the spectrum of primary users is modeled as a block structure signal. Specifically, the block threshold feature (BTF) is utilized to reconstruct the spectrum while bypassing any iterative operations. Furthermore, to improve the performance of the BTF-based BCSS, we develop a novel supervised dictionary learning (SDL) model, based on the theoretical analysis of mutual incoherence and restricted isometry properties. Simulation results not only verify the compatibility of BTF and the SDL model, but also demonstrate the effectiveness and robustness of SDL-BTF-based BCSS for practical implementation.
... Hence, many studies employed DL to develop solutions either for new complex systems or as an alternative to the old proven approaches. To mention a few but not the all, [1][2][3] examine signal detection, [4] uses DL for channel encoding and decoding, in [5][6][7][8][9][10][11] channel estimation, prediction, and compression are examined, in [12][13][14] resource allocation is studied. ...
... For example, the authors in [9] consider the time-frequency response of a fast fading communication channel as a two-dimensional image and try to find the unknown values of the channel response using some known values at the pilot locations and applying image processing. In [7], RNN is used for channel estimation with DL in time selective channel where pilot density was a consideration. In [8], the authors propose a new channel estimation method with the assistance of deep learning in order to support the low-cost least squares estimation to reduce to high channel estimation error. ...
In this study, we focus on realizing channel estimation using a fully connected deep neural network. The data aided estimation approach is employed. We assume the transmission channel is Rayleigh and it is constant over the duration of a symbol plus pilot transmission. We develop and tune the deep learning model for various size of pilot data that is known to the receiver and used for channel estimation. The deep learning models are trained on the Rayleigh channel. The performance of the model is discussed for various size of pilot by providing Bit Error Rate of the model. The Bit Error Rate performance of the model is compared to theoretical upper bound which shows that the model successfully estimates the channel.
... A similar network model was utilized in [24], where the authors presented deep channel estimator with untrained deep neural network to reduce pilot contamination. To trakle the time-varying Rayleigh fading channel, a sliding bidirectional gated recurrent unit channel estimator was designed to improve the channel estimation performance [25]. In [26], a residual learning based deep neural network (DNN) was designed for channel estimation. ...
... . The corresponding result can be expressed in (19). ) By calculating (25), the corresponding result S (n,k) can be used as the input of this layer, and the output of this layer is expressed as ...
The emerging intelligent reflecting surface (IRS) can significantly improve the system capacity, and it has been regarded as a promising technology for the beyond fifth-generation (B5G) communications. For IRS-assisted multiple input multiple output (MIMO) systems, accurate channel estimation is a critical challenge. This severely restricts practical applications, particularly for resource-limited indoor scenario as it contains numerous scatterers and parameters to be estimated, while the number of pilots is limited. Prior art tackles these issues and associated optimization using mathematical-based statistical approaches, but are difficult to solve as the number of scatterers increase. To estimate the indoor channels with an affordable piloting overhead, we propose an offset learning (OL)-based neural network for channel estimation. The proposed OL-based estimator can dynamically trace the channel state information (CSI) without any prior knowledge of the IRS-assisted channel structure as well as indoor statistics. In addition, inspired by the powerful learning capability of convolutional neural network (CNN), CNN-based inversion blocks are developed in the offset estimation module to build the offset estimation operator. Numerical results show that the proposed OL-based estimator can achieve more accurate indoor CSI with a lower complexity as compared to the benchmark schemes.
... As perfect CSI is not available in practical systems, channel estimation is usually carried out before HBF optimization. Many channel estimation algorithms with the HBF architecture have been proposed [19]- [30], including these [19]- [22] applying the DL method. The authors in [19] regarded the channel as a 2D image and used image processing technology and a convolutional neural network (CNN) to achieve super-resolution image restoration and channel reconstruction. ...
... The authors in [21] proposed to use NNs to learn the basic characteristics of the channel from low-rank measurements and map the characteristics to a channel matrix. The authors in [22] proposed a DL-based channel estimation algorithm in a time-varying Rayleigh fading channel to dynamically track the CSI. The authors in [23]- [25] converted channel estimation into a sparse signal recovery problem. ...
Hybrid analog and digital beamforming (HBF) has been regarded as a key technology for future millimeter wave (mmWave) communication systems due to its ability to obtain a good trade-off between achievable beamforming gain and hardware cost. In this paper, we investigate the channel estimation and hybrid precoding for mmWave MIMO systems with deep learning. We adopt the hierarchical codebook based algorithm for channel estimation as it requires limited number of pilot transmissions, and enhance its performance by proposing a new codebook design algorithm based on manifold optimization (MO). With the estimated channel state information (CSI) as the input, we develop a robust HBF network (HBF-Net) by applying convolutional layers and attention mechanism, which can be trained to generate a robust HBF matrix targeting at spectral efficiency maximization with imperfect CSI. To further improve the performance, we propose a joint channel estimation and HBF optimization network (CE-HBF-Net). Considering that the adaptively selected HBF vectors in the hierarchical codebook based channel estimation are different for different channel realizations, we skillfully propose an index assign-and-input method to efficiently feed such information to the CE-HBF-Net to reduce the network input dimensions and make the network trainable. Furthermore, we propose a signal self-attention mechanism to enable the CE-HBF-Net to intelligently assign larger weight coefficients to those signals that contribute more to channel estimation. Simulation results show that the well-designed HBF-Net and CE-HBF-Net outperform the conventional HBF algorithms with imperfect channel and exhibit robustness to mismatches between offline training and online deployment stages.
... Citation information: DOI 10.1109/JIOT.2023.3274209 1. First, the generator and discriminator are trained iteratively to update the parameters by calculating the loss function according to equations (12) and (13). Then the parameters of transmitter and receiver are updated according to equation (2). ...
An end-to-end learning framework is proposed to optimize each module jointly in the communication system. Recently, Convolutional Neural Network (CNN) and conditional Generative Adversarial Network (cGAN) are used for end-to-end learning. However, deeper network layers will degrade the effect of CNN. cGAN suffers from unstable training and lacks generative diversity. In this paper, we propose the end-to-end learning based on Deep Residual Network (ResNet) and Wasserstein GAN (WGAN) for communication with unknown channels (ResNet-WGAN). First, ResNet is applied to solve the problem of network degradation to extract deeper data features. Second, for unknown channels, WGAN with conditional information is used to fit the channel effect to improve training stability and generative diversity. Finally, we present the simulation results of the ResNet-WGAN under additive white Gaussian noise (AWGN) channel, Rayleigh fading channel and frequency selective channel. The results demonstrate that the ResNet-WGAN reduces the communication Bit Error Rate (BER) and Block Error Rate (BLER). In particular, this paper applies ResNet-WGAN to the Internet of Vehicles (IoV) communication, and the results demonstrate that ResNet-WGAN is more effective.
... In this paper, we select the least square channel estimation (LS) for estimating the CSI due to its low complexity and extensive application [32]- [34]. Worthy of note is that the main thrust of this study remains on the statistical QoS provisioning analysis and performance optimization in xURLLCenabled massive MU-MIMO wireless networks, even though the channel estimation is taken into consideration. ...
In this paper, fundamentals and performance tradeoffs of the neXt-generation ultra-reliable and low-latency communication (xURLLC) are investigated from the perspective of stochastic network calculus (SNC). An xURLLC-enabled massive MU-MIMO system model has been developed to accommodate xURLLC features. By leveraging and promoting SNC, we provide a quantitative statistical quality of service (QoS) provisioning analysis and derive the closed-form expression of upper-bounded statistical delay violation probability (UB-SDVP). Based on the proposed theoretical framework, we formulate the UB-SDVP minimization problem that is first degenerated into a one-dimensional integer-search problem and then can be efficiently solved by the integer-form Golden-Section search algorithm. Moreover, two novel concepts, error probability-based effective capacity (EP-EC) and energy efficiency (EP-EE) have been defined to characterize the tail distribution and performance tradeoffs for xURLLC. Subsequently, we formulate the EP-EC and EP-EE maximization problems and prove that the EP-EC maximization problem is equivalent to the UB-SDVP minimization problem, while the EP-EE maximization problem is solved with a low-complexity outer-descent inner-search collaborative algorithm. Extensive simulations validate and demonstrate that the proposed framework in reducing computational complexity compared to reference schemes, and in providing various tradeoffs and optimization performance of xURLLC concerning UB-SDVP, EP, EP-EC, and EP-EE.
... The traditional SIC methods such as LS and MMSE are used, and are also applied for CSI estimation and detection of the signal [26]. In advance, the correction coefficient R hh for MMSE estimation is calculated. ...
Non-orthogonal multiple access (NOMA) has great potential to implement the fifth generation (5G) requirements of wireless communication. For a NOMA traditional detection method, successive interference cancellation (SIC) plays a vital role at the receiver side for both uplink and downlink transmission. Due to the complex multipath channel environment and prorogation of error problems, the traditional SIC method has a limited performance. To overcome the limitation of traditional detection methods, the deep-learning method has an advantage for the highly efficient tool. In this paper, a deep neural network which has bi-directional long short-term memory (Bi-LSTM) for multiuser uplink channel estimation (CE) and signal detection of the originally transmitted signal
is proposed. Unlike the traditional CE schemes, the proposed Bi-LSTM model can directly recover multiuser transmission signals suffering from channel distortion. In the offline training stage, the Bi-LTSM model is trained using simulation data based on channel statistics. Then, the trained model is used to recover the transmitted symbols in the online deployment stage. In the simulation results, the performance of the proposed model is compared with the convolutional neural network model and traditional CE schemes such as MMSE and LS. It is shown that the proposed method provides feasible improvements in performance in terms of symbol-error rate and signal-to-noise ratio, making it suitable for 5G wireless communication and beyond.
... Therefore, a growing number of researchers have started to study the application of deep learning in communication systems, hoping that deep learning techniques can be used to solve communication challenges that are difficult to be solved by traditional communication algorithms. There are many studies on the introduction of deep learning into communication systems, including channel estimation [10], channel decoding [11], channel equalization [12], channel modeling [13], modulated signal identification [14], or other local performance optimization. For signal modulation identification, the literature [15] addresses cognitive radio and proposes the use of deep learning methods for automatic signal modulation identification. ...
With the advancement of an intellectual and numerical society, the coal mining industry has also begun to change to intelligence. As an important aspect of intelligent coal mine construction, coal mine communication has put forward more stringent standards for communication quality. For the complex communication environment in mines, the transmission of communication signals is always damaged by various noises and interferences, resulting in serious distortion of the communication signals received at the receiving end. Therefore, the use of traditional receivers for information recovery has the problem of high bit error rate (BER), which cannot meet the standard of intelligent coal mine construction. Based on this, the aim of this research is to combine convolutional neural networks (CNN) and multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM) communication systems to design an intelligent receiver model for complex mine communication systems. At the receiver side, CNNs are used to take the place of all the information processing processes. First, features are extracted from the received IQ signal by the convolutional neural network, and then the original information bit is recovered using a multi-label classifier to finally realize end-to-end information restoration. The experimental results show that the intelligent receiver model designed in this research has more accurate information recovery capability in the complex mine channel environment compared with the traditional receiver. In addition, they also verify that the intelligent receiver can still recover information effectively when the traditional receiver cannot recover information properly in the case of partial loss of received data.
... Here, the channel exchange is constant within the single data block of the OFCDM method (Murthy et al. 2006). The existing techniques for the estimation of the channel methodologies are categorized into three groups such as blind methods, semi-blind methods, and training-related methods (Bai et al. 2020). ...
Channel estimation is a significant constraint for the communication system. The channel estimation in wireless communication systems is carried out by adopting distinct approaches. In addition to this, the more important techniques are Minimum Mean Square Error (MMSE), and Least Square (LS). Here, the LS process for estimating the channels is simpler, but it gets high error regarding Mean Square Error (MSE). On contrary to that, the efficiency of MMSE is more when evaluating with LS with Signal-To-Noise Ratio (SNR). It also has high computational complexity. Therefore, the integration of LS and MMSE approaches are used for receiving the exact signal and reducing the error rate through evolutionary programming. Moreover, the Invariable Step-Size Zero-Attracting Normalized Least Mean Square (ISS-ZA-NLMS) methodology has been adopted for exploiting the channel sparsity in Adaptive Sparse Channel Estimation (ASCE). On the other hand, ISS-ZA-NLMS faces inefficiency in performance in terms of a lack of good trade-off between the computational cost and convergence rate. Hence, the main scope of this research is to estimate a new channel estimation technique in broadband wireless communication systems with the help of a hybridized optimization algorithm. Here, the development of Hybrid Heuristic-based ISS-ZA-NLMS (HH-ISS-NLMS) is the main contribution that could enhance the ASCE. The integrated algorithm Tunicate Swarm-Deer Hunting Optimization (TS-DHO) is proposed for efficiently estimating the channels. The objective function of the advanced method is derived concerning MSE. Finally, results show that the channel estimation by the offered HH-ISS-NLMS algorithm estimated with the other traditional approaches shows enhanced performance in terms of MSE.
... However, this structure will lead to lengthy timestep, which means considerable training complexity. • Use sliding window for the whole sequences [19]. The waveform sequences within the window are considered as The details about BiLSTM layer in Figure 2 corresponding features for one label symbol. ...
Numerous researches on communication receiver design with deep learning algorithms have been implemented effectively recently. In this paper, an innovative scheme is proposed to construct an end‐to‐end neural network (NN) receiver to directly recover information bits from unsynchronized waveform sequences. Considering the correlation between samples, multiple bidirectional long‐short term memory (BiLSTM) layers to process oversampled signals. Then, shifted binary cross‐entropy (SBCE) function is devised to tackle the overfitting issue and eliminate instances with abnormal bit error rate (BER), which originate from standard binary cross‐entropy (BCE) loss function. Substantial simulation results demonstrate that the trained NN receiver can achieve theory BER values under excellent conditions for transceivers. Compared with traditional synchronization algorithms, the proposed method can obtain significant BER performance gain under harsh conditions, such as low oversample rate, small roll‐off factor, short frame length.
... Based on the deep learning (DL) algorithm, the authors in [10] introduced a sliding window Gated Recurrent Unit (GRU) channel estimator to acquire knowledge for the timevarying Rayleigh fading channel. Interleaver and channel coding schemes were merged with the proposed sliding window estimator to further enhance system performance. ...
In a non-orthogonal multiple access (NOMA) system, the successive interference cancellation (SIC) procedure is typically employed at the receiver side, where several user’s signals are decoded in a subsequent manner. Fading channels may disperse the transmitted signal and originate dependencies among its samples, which may affect the channel estimation procedure and consequently affect the SIC process and signal detection accuracy. In this work, the impact of Deep Neural Network (DNN) in explicitly estimating the channel coefficients for each user in NOMA cell is investigated in both Rayleigh and Rician fading channels. The proposed approach integrates the Long Short-Term Memory (LSTM) network into the NOMA system where this LSTM network is utilized to predict the channel coefficients. DNN is trained using different channel statistics and then utilized to predict the desired channel parameters that will be exploited by the receiver to retrieve the original data. Furthermore, this work examines how the channel estimation based on Deep Learning (DL) and power optimization scheme are jointly utilized for multiuser (MU) recognition in downlink Power Domain Non-Orthogonal Multiple Access (PD-NOMA) system. Power factors are optimized with a view to maximize the sum rate of the users on the basis of entire power transmitted and Quality of service (QoS) constraints. An investigation for the optimization problem is given where Lagrange function and Karush–Kuhn–Tucker (KKT) optimality conditions are applied to deduce the optimum power coefficients. Simulation results for different metrics, such as bit error rate (BER), sum rate, outage probability and individual user capacity, have proved the superiority of the proposed DL-based channel estimation over conventional NOMA approach. Additionally, the performance of optimized power scheme and fixed power scheme are evaluated when DL-based channel estimation is implemented.
... In [24], the authors regarded CSI as 2D images and used DL-based image processing techniques to estimate the channel. In [25], a channel estimator using the sliding bidirectional gated recurrent unit network was designed at the receiver, which can be combined with other channel estimation techniques. In [26], a DNN was constructed for channel estimation and direction of arrival estimation, improving performance without increasing complexity. ...
This paper studies the user selection problem for a cooperative nonorthogonal multiple access (NOMA) system consisting of a base station, a far user, and N near users. The selected near user receives its own message and assists the far user by relaying the far user’s message. Firstly, we propose a user selection strategy to maximize the selected near user’s data rate while satisfying the quality-of-service (QoS) requirement of the far user. Considering that the channel state information (CSI) of users in actual communication is usually imperfect, we then analyze the outage probability of the NOMA system based on the user selection strategy under imperfect CSI and obtain a closed-form expression. The theoretical analysis shows that the diversity order of the NOMA system under imperfect CSI is 0, which means the multiuser diversity order disappears. In order to improve the impact of imperfect CSI on system performance, we use the deep learning method to identify and classify channels of imperfect CSI and improve the accuracy of CSI. The simulation results show that the theoretical analysis of outage performance is consistent with the numerical results. Compared with the strategy without the deep learning method, the proposed deep learning-based user selection scheme significantly improves the system performance. Furthermore, we verify that our scheme recovers the diversity gain.
... In [27], the NN and DL methods have been used to predict the behavior of the Rayleigh channel, and it has been reported through simulations that the MSE performance compared with the traditional algorithms has improved. In [22,28], DL has been thoroughly investigated and provided a review of the various ML-based techniques for wireless communication. ...
Wireless communication systems have evolved and offered more smart and advanced systems like ad hoc and sensor-based infrastructure fewer networks. These networks are evaluated with two fundamental parameters including data rate and spectral efficiency. To achieve a high data rate and robust wireless communication, the most significant task is channel equalization at the receiver side. The transmitted data symbols when passing through the wireless channel suffer from various types of impairments, such as fading, Doppler shifts, and Intersymbol Interference (ISI), and degraded the overall network performance. To mitigate channel-related impairments, many channel equalization algorithms have been proposed for communication systems. The channel equalization problem can also be solved as a classification problem by using Machine Learning (ML) methods. In this paper, channel equalization is performed by using ML techniques in terms of Bit Error Rate (BER) analysis and comparison. Radial Basis Functions (RBFs), Multilayer Perceptron (MLP), Support Vector Machines (SVM), Functional Link Artificial Neural Network (FLANN), Long-Short Term Memory (LSTM), and Polynomial-based Neural Networks (NNs) are adopted for channel equalization.
... A channel estimator based on deep learning(DL) for time-varying Rayleigh fading channels is proposed in [9]. It use neural networks(NN) to build train a nd test channel estimators. ...
... Due to its strong ability to learn, recognize, and predict, DL has been regarded as a promising tool in solving the complex communication problems in unknown or complex channels. A comprehensive introduction and overview of the application of DL in wireless communications has been reported in [24][25][26][27][28][29][30][31][32][33], including the fully connected deep neural network (FC-DNN)-based channel estimation, nonlinearity compensation, signal detection, and decoding scheme. In addition, DL has also been applied in VLC to enhance the system performance. ...
The inherent impairments of visible light communication (VLC) in terms of nonlinearity of light-emitting diode (LED) and the optical multipath restrict bit error rate (BER) performance. In this paper, a model-driven deep learning (DL) equalization scheme is proposed to deal with the severe channel impairments. By imitating the block-by-block signal processing block in orthogonal frequency division multiplexing (OFDM) communication, the proposed scheme employs two subnets to replace the signal demodulation module in traditional system for learning the channel nonlinearity and the symbol de-mapping relationship from the training data. In addition, the conventional solution and algorithm are also incorporated into the system architecture to accelerate the convergence speed. After an efficient training, the distorted symbols can be implicitly equalized into the binary bits directly. The results demonstrate that the proposed scheme can address the overall channel impairments efficiently and can recover the original symbols with better BER performance. Moreover, it can still work robustly when the system is complicated by serious distortions and interference, which demonstrates the superiority and validity of the proposed scheme in channel equalization.
... However, in actual fact, compared with multi-tone interference occupying on different frequency bins, narrowband or wideband interference are more common [28]. Therefore, to obtain a compete channel response, the restored frequency domain pilots must be interpolated, with frequently used constant interpolation, Gaussian interpolation, or cubic spline interpolation [29]. ...
The Internet of Things (IoT) leads the era of interconnection, where numerous sensors and devices are being introduced and interconnected. To support such an amount of data traffic, wireless communication technologies have to overcome available spectrum shortage and complex fading channels. The transform domain communication system (TDCS) is a cognitive anti-interference communication system with a low probability of detection and dynamic spectrum sensing and accessing. However, the non-continuous and asymmetric spectrum brings new challenges to the traditional TDCS block-type pilot, which uses a series of discrete symbols in the time domain as pilots. Low efficiency and poor adaptability in fast-varying channels are the main drawbacks for the block-type pilot in TDCS. In this study, a frequency domain non-uniform pilot design method was proposed with intersecting, skewing, and edging of three typical non-uniform pilots. Some numerical examples are also presented with multipath model COST207RAx4 to verify the proposed methods in the bit error ratio and the mean square error. Compared with traditional block-type pilot, the proposed method can adapt to the fast-varying channels, as well as the non-continuous and asymmetric spectrum conditions with much higher efficiency.
... The applications include designing signals, estimating channels, detecting data, and developing modulation/demodulation schemes. For example, studies in [607]- [611] develop deep learning-based schemes for channel estimation in OFDM and MIMO systems. ...
Wireless transceivers for mass-market applications must be cost effective. We may achieve this goal by deploying non-ideal low-cost radio frequency (RF) analog components. However, their imperfections may result in RF impairments, including phase noise (PN), carrier frequency offset (CFO), and in-phase (I) and quadrature-phase (Q) imbalance. These impairments introduce in-band and out-of-band interference terms and degrade the performance of wireless systems. In this survey, we present RF-impairment signal models and discuss their impacts. Moreover, we review RF-impairment estimation and compensation in single-carrier (SC) and multicarrier systems, especially orthogonal frequency division multiplexing (OFDM). Furthermore, we discuss the effects of the RF impairments in already-established wireless technologies, e.g., multiple-input multiple-output (MIMO), massive MIMO, full-duplex, and millimeter-wave communications and review existing estimation and compensation algorithms. Finally, future research directions investigate the RF impairments in emerging technologies, including cell-free massive MIMO communications, non-orthogonal multicarrier systems, non-orthogonal multiple access (NOMA), ambient backscatter communications, and intelligent reflecting surface (IRS)-assisted communications. Furthermore, we discuss artificial intelligence (AI) approaches for developing estimation and compensation algorithms for RF impairments.
... A simple example of a single-layer RNN is given in Fig. 13, where the output of the previous time step t − 1 becomes a part of the input of the current time step t, thus capturing past information. Computation result performed by one RNN cell can be expressed as a following function [34]: ...
Wi-Fi systems based on the IEEE 802.11 standards are the most popular wireless interfaces that use Listen Before Talk (LBT) method for channel access. The distinctive feature of a majority of LBT-based systems is that the transmitters use preambles that precede the data to allow the receivers to perform packet detection and carrier frequency offset (CFO) estimation. Preambles usually contain repetitions of training symbols with good correlation properties, while conventional digital receivers apply correlation-based methods for both packet detection and CFO estimation. However, in recent years, data-based machine learning methods are disrupting physical layer research. Promising results have been presented, in particular, in the domain of deep learning (DL)-based channel estimation. In this paper, we present a performance and complexity analysis of packet detection and CFO estimation using both the conventional and the DL-based approaches. The goal of the study is to investigate under which conditions the performance of the DL-based methods approach or even surpass the conventional methods, but also, under which conditions their performance is inferior. Focusing on the emerging IEEE 802.11ah standard, our investigation uses both the standard-based simulated environment, and a real-world testbed based on Software Defined Radios.
communication performance in wireless networks are greatly affected by the fidelity of the channel estimator, revisiting old methods of the estimation of LS and MMSE channels by revealing their shortcomings, a new method based on deep-learning is proposed, demonstrating significant improvements in communications reliability as well as the reduction of the computational complexity of vehicular networks, under different high mobility environments ( high doppler frequency) with very short reflection times are tested by the new model Dl estimator
Sparse code multiple access (SCMA) is a promising non‐orthogonal multiple access scheme for cellular Internet of things (IoT) due to its ability to support massive connectivity, grant‐free transmission and scalability. Inspired by the recent developments of deep learning for physical layer communications, we present a design of an uplink SCMA receiver using deep learning. We propose the use of recurrent neural networks (RNNs) for joint channel estimation and multiuser data detection of uplink SCMA under time‐varying Rayleigh channel using a single deep learning structure. The use of RNNs enables the receiver to learn the time correlation between the received samples with a very low pilot density and with low complexity. Compared with the conventional SCMA receiver, the simulation results show that the proposed deep learning‐based receiver can achieve BER performance similar to that of the conventional SCMA receivers (such as sparse pilot channel estimator and message‐passing algorithm detector) with very low pilot density and with much lower complexity. Moreover, the proposed SCMA receiver shows good resilience to small changes in the receiver speed (second‐order channel statistics) which enables the proposed deep learning receiver to work over a reasonable range of receiver speeds without parameter tuning. Fine tuning the network parameters to capture the variations of the channel, during online transmission using small data sets and small training period, is also checked and provides additional BER benefits.
The trend of using larger scale antenna arrays will continue toward 6G systems, where the number of antennas will be further scaled up to improve spectral efficiency. However, the increase in the number of antennas will bring new challenges to the physical layer, such as frequent feedback in high-speed mobile communications, multiband coexistence overhead from sub-6 (gigahertz) GHz to terahertz (THz), and energy consumption due to increased antenna components and circuits. In this article, we introduce artificial intelligence (AI)-based channel extrapolation to address these problems. Specifically, we divide the channel extrapolation into time, frequency, and space domains according to different application scenarios. The channel propagation characteristics that affect the extrapolation of each domain, such as the spatial consistency property (SCP), partial reciprocity, and spatial nonstationarity, are analyzed. The motivations for selecting various AI models in each domain are explained, and the performance of AI models is compared. Furthermore, we find the gain of cross-domain channel extrapolation based on transfer learning (TL). The simulation results show that the experience of the AI model cross different domains can be mutually reinforcing. Finally, we introduce several challenges for AI-based channel extrapolation, which can be regarded as potential research directions for realizing future AI-powered 6G systems.
In wireless communication systems, channel estimation is one of the vital processes for determining channel characteristics. Deep learning based approaches construct a 2D image from the time-frequency grid of the channel. Further, they adopted an image pipeline with super-resolution convolutional neural network (SRCNN) and denoising convolutional neural network (DnCNN) algorithms. However, these two algorithms suffer from a large memory footprint and slower execution times due to fewer convolution operations and high resolution (HR) image processing. To cope with this, this paper aims to propose and implement two new advanced algorithms, namely fast SRCNN (FSRCNN) and fast and flexible DnCNN (FFDNet). FSRCNN uses additional convolutional layers to reduce the total number of operations without compromising performance. It also employs upscaling layers and smaller filter sizes, making FSRCNN faster and more efficient than SRCNN. FFDnet has significantly improved over DnCNN by injecting a noise channel as an input and performing downscaling and upscaling before and after non-linear mapping. These models are trained and evaluated over a range of SNRs and pilots values for comparing their performance with other existing models.
To address the problems of the high complexity and poor bit error rate (BER) performance of traditional communication systems in underwater acoustic environments, this paper incorporates the theory of deep learning into a conventional communication system and proposes data-driven underwater acoustic filter bank multicarrier (FBMC) communications based on convolutional autoencoder networks. The proposed system is globally optimized by two one-dimensional convolutional (Conv1D) modules at the transmitter and receiver, it realizes signal reconstruction through end-to-end training, it effectively avoids the inherent imaginary interference of the system, and it improves the reliability of the communication system. Furthermore, dense-block modules are constructed between Conv1D layers and are connected across layers to achieve feature reuse in the network. Simulation results show that the BER performance of the proposed method outperforms that of the conventional FBMC system with channel equalization algorithms such as least squares (LS) estimation and virtual time reversal mirrors (VTRM) under the measured channel conditions at a certain moment in the Qingjiang River.
Channel estimation is essential to wireless network system performance. Deep learning (DL) has also shown considerable advances in improving communication reliability and lowering computing complexity in 5G networks. In spite of this, least square estimation is frequently used to provide channel estimates due to its small cost. The LS approach has a high level of estimation error due to the complexity with which Minimum Mean Square Error adjusts for noise to obtain higher performance than LS. Deep learning has demonstrated its ability to lower computational complexity while simultaneously enhancing system performance in 5G and future networks. This review study's focus is deep learning-aided channel estimation. In this study, channel estimation methods employing traditional and deep learning techniques are examined, and comparisons of various estimation methods are given. From the comparison study, we know the optimal deep learning types for Channel Estimation and the past or current approaches utilized for it.
In this paper, reinforcement learning (RL) based end-to-end channel prediction (CP) and beamforming (BF) algorithms are proposed for multi-user downlink system. Different from the previous methods which either require perfect channel state information (CSI), or estimate outdated CSI and set constraints on pilot sequences, the proposed algorithms have no such premised assumptions or constraints. Firstly, RL is considered in channel prediction and the actor-critic aided CP algorithm is proposed at the base station (BS). With the received pilot signals and partial feedback information, the actor network at BS directly outputs the predicted downlink CSI without channel reciprocity. After obtaining the CSI, BS generates the beamforming matrix using zero-forcing (ZF). Secondly, we further develop a deep RL based two-layer architecture for joint CP and BF design. The first layer predicts the downlink CSI with the similar actor network as in the CP algorithm. Then, by importing the outputs of the first layer as inputs, the second layer is the actor-critic based beamforming layer, which can autonomously learn the beamforming policy with the objective of maximizing the transmission sum rate. Since the learning state and action spaces in the considered CP and BF problems are continuous, we employ the actor-critic method to deal with the continuous outputs. Empirical numerical simulations and the complexity analysis verify that the proposed end-to-end algorithms could always converge to stable states under different channel statistics and scenarios, and can beat the existing traditional and learning based benchmarks, in terms of transmission sum rate.
Optical transmission security has attracted much attention. In recent years, many secure optical transmission systems based on channel characteristics are proposed. However, there are many drawbacks with these systems, such as separated plaintext and key transmission, low key generation rate (KGR), insecurity when the eavesdropper has acquired the lengths of the local fibers utilized by legal parties. To solve the above problems, we propose a novel secure optical transmission system based on neural networks (NNs), which are employed to estimate channel characteristics. By training NNs locally and transmitting pseudo-keys, the proposed system can transmit the plaintext together with key, transforming the key dynamically. Moreover, since the channel characteristics for legal parties and eavesdropper are not completely identical, the NNs trained by legal parties and eavesdropper are inconsistent. Even though the eavesdropper has attained the lengths of local fibers wielded by legal parties, the NN model trained by the legal parties is still unavailable to illegal eavesdropper. The final key is generated by the trained NN and pseudo-key, so the keys generated by legal parties and eavesdropper are dissimilar. The simulation results prove the feasibility of the proposed system with the transmission distance of 100 km and the bit rate of 100 Gbps. Meanwhile, if plaintext and key have equivalent code length, the KGR of 50 Gbps for legal parties and the key disagreement rate (KDR) of 50% for illegal eavesdropper will be realized.
6G and beyond wireless networks will satisfy the requirements of a fully connected world and thereby provide ‘truly ubiquitous ‘connectivity everywhere. This chapter will discuss the following key technology for 6G and beyond wireless network: Massive Cell-Free MIMO, Quantum Communication, Intelligent Communication environment, Internet of Nano Things, Internet of Space Things with CubeSats, Artificial Intelligence for 6G and beyond Network, Terahertz band Communication, UAV-Based Communication, Visible Light Communication, Ambient Backscatter Communication, Context-aware cognitive radio, Reconfigurable Transceiver Front-ends for 6G and beyond, Advanced 6G Applications, Research and Project Activities. Each topic is discussed at length.Keywords6G and beyondCell-free massive MIMOQuantum communicationArtificial IntelligenceInternet of Things
In recent years, Cognitive Radio (CR) is an advanced and immense technology developed for enhancing the efficiency of spectrum utilization by permitting unauthorized Secondary Users (SUs) functions within service limit of authorized Primary Users (PUs), while creating bearable intervention to the PUs. Typically, secondary network is constructed as Half-Duplex (HD) and are carried out orthogonally in frequency or time domain. CR has been considered as an eminent technology to solve the issue of scarcity of spectrum in wireless communications. The main motive behind constructing CR technology is to increase the efficiency of spectrum utilization. In addition, there is a high need for sensing spectrum and this delay for sensing creates a huge constraint in construction of sensing approaches. This is because of minimum time consumed in spectrum sensing, and later much time is existed for transmission. Hence, this research proposes an effective approach for channel estimation in CR. Here, the optimization of pilot design in CR system is designed depending upon Least-Square channel estimation, where position to insert the pilot symbols are determined using proposed optimization algorithm called Adaptive Rider-Grey Wolf Optimizer (Adaptive RGWO). However, proposed Adaptive RGWO is devised by inheriting adaptive concept with merits of Rider Optimization Algorithm (ROA) and Grey Wolf Optimizer (GWO). Moreover, proposed Adaptive RGWO has achieved minimum BER of 0.0000028 for Rician channel with 768 carriers and minimum MSE of 0.000001 for Rician channel with 512 carriers.
We provide a new generation solution to the fundamental old problem of a doubly selective fading channel estimation for orthogonal frequency division multiplexing (OFDM) systems. For systems based on OFDM, we propose a deep learning (DL)-based blind doubly selective channel estimator. This estimator does require no pilot symbols, unlike the corresponding state-of-the-art estimators, even during the estimation of a deep fading doubly selective channel. We also provide the first of its kind theory on the testing mean squared error (MSE) performance of our investigated blind OFDM channel estimator based on over-parameterized ReLU FNNs.
Chaos-based communications can be applied to high-speed vehicular information transmissions thanks to the anti-jamming and anti-interference capabilities of chaotic transmissions. In traditional practical chaotic systems, reference chaotic signals are required to be delivered to remove complex chaotic synchronization circuits. However, the direct transmission of reference signals will degrade the security performances, while interferences and noises imposed on the reference signals due to imperfect channel conditions will deteriorate the reliability performances. In order to enhance the reliability and security performances over vehicular channels such as the railway channel and the channels undergoing fast fadings, in this paper, we propose a deep learning (DL) aided intelligent OFDM-DCSK transceiver. In this design, no reference chaotic signals are delivered, and we propose to utilize the time-delay neural network (TDNN) to learn the chaotic maps, followed by the long short-term memory (LSTM) units to extract and exploit the correlations between chaotic modulated signals, and multiple fully connected layers (FCLs) to estimate the user bit data. With the aid of the constructed deep neural network (DNN), after the offline neural network training, the receiver can recover the transmitted information with lower bit error rate (BER) and enhance security performances. Theoretical performance is then analyzed for the proposed intelligent transceiver. Simulation results validate the proposed design, and demonstrate that the intelligent DL-based OFDM-DCSK system can achieve better BER and security performances over fast fading and railway channels compared with the benchmark systems.
In this article, we propose a unified simultaneous wireless information and power transfer (SWIPT) signaling and architecture to take advantage of both single tone and multitone signaling by adjusting only the power allocation ratio of a unified signal. Toward this, we design a unified and integrated receiver architecture for the proposed unified SWIPT signaling via an envelope detection with low power consumption. To lower the computational complexity of the receiver, we propose an adaptive control algorithm where the transmitter adjusts the communication mode through temporal convolutional network (TCN)-based
asymmetric processing
. To this end, the transmitter optimizes the modulation index and power allocation ratio in short-term scale while updating the mode switching threshold in long-term scale. We demonstrate that the proposed unified SWIPT system improves the achievable rate under the self-powering condition at low-power Internet of Things (IoT) devices. Consequently we will facilitate the effective deployment of low-power IoT networks that concurrently supply both information and energy
wirelessly
to the devices by using the proposed unified SWIPT and adaptive control algorithm at the transmitter side.
The power of big data and machine learning has been drastically demonstrated in many fields during the past twenty years which somehow leads to the vague even false understanding that the huge amount of precious human knowledge accumulated to date seems to no longer matter. In this paper, we are pioneering to propose the knowledge-driven machine learning (KDML) model to exhibit that knowledge can play an important role in machine learning tasks. Compared with conventional machine learning, KDML contains a unique knowledge module based on specific domain knowledge, which is able to simplify the machine learning network structures, reduce the training overhead and improve interpretability. Channel estimation problem of wireless communication is taken as a case verification because such machine learning-based solutions face huge challenges in terms of accuracy, complexity, and reliability. We integrate the classical wireless channel estimation algorithms into different machine learning neural networks and propose KDML-based channel estimators in Orthogonal Frequency Division Multiplexing (OFDM) and Massive Multiple Input Multiple Output (MIMO) system. The experimental results in both communication systems validate the effectiveness of the proposed KDML-based channel estimators.
Multiple lines of evidence suggest a potential role for the kappa dynorphin system in schizophrenia and its therapeutics. Kappa stimulation acutely and chronically modulates dopamine in opposite ways, where acutely it decreases dopamine transmission and chronically it increases it and it can induce D2 sensitization. In addition, pharmacological evidence from studies using agonist and antagonists at KOR have indicated a therapeutic potential of KOR antagonists for psychosis. We present here a brief overview of the evidence supporting this viewpoint.
In this paper, online deep learning (DL)-based channel estimation algorithm for doubly selective fading channels is proposed by employing the deep neural network (DNN). With properly selected inputs, the DNN can not only exploit the features of channel variation from previous channel estimates but also extract additional features from pilots and received signals. Moreover, the DNN can take the advantages of the least squares estimation to further improve the performance of channel estimation. The DNN is first trained with simulated data in an off-line manner and then it could track the dynamic channel in an online manner. To reduce the performance degradation from random initialization, a pre-training approach is designed to refine the initial parameters of the DNN with several epochs of training. The proposed algorithm benefits from the excellent learning and generalization capability of DL and requires no prior knowledge about the channel statistics. Hence, it is more suitable for communication systems with modeling errors or non-stationary channels, such as high-mobility vehicular systems, underwater acoustic systems, and molecular communication systems. The numerical results show that the proposed DL-based algorithm outperforms the existing estimator in terms of both efficiency and robustness, especially when the channel statistics are time-varying.
In this paper, we implement an optical fiber communication system as an end-to-end deep neural network, including the complete chain of transmitter, channel model, and receiver. This approach enables the optimization of the transceiver in a single end-to-end process. We illustrate the benefits of this method by applying it to intensity modulation/direct detection (IM/DD) systems and show that we can achieve bit error rates below the 6.7\% hard-decision forward error correction (HD-FEC) threshold. We model all componentry of the transmitter and receiver, as well as the fiber channel, and apply deep learning to find transmitter and receiver configurations minimizing the symbol error rate. We propose and verify in simulations a training method that yields robust and flexible transceivers that allow---without reconfiguration---reliable transmission over a large range of link dispersions. The results from end-to-end deep learning are successfully verified for the first time in an experiment. In particular, we achieve information rates of 42\,Gb/s below the HD-FEC threshold at distances beyond 40\,km. We find that our results outperform conventional IM/DD solutions based on 2 and 4 level pulse amplitude modulation (PAM2/PAM4) with feedforward equalization (FFE) at the receiver. Our study is the first step towards end-to-end deep learning-based optimization of optical fiber communication systems.
This article presents our initial results in deep learning for channel estimation and signal detection in orthogonal frequency-division multiplexing (OFDM). OFDM has been widely adopted in wireless broadband communications to combat frequency-selective fading in wireless channels. In this article, we take advantage of deep learning in handling wireless OFDM channels in an end-to-end approach. Different from existing OFDM receivers that first estimate CSI explicitly and then detect/recover the transmitted symbols with the estimated CSI, our deep learning based approach estimates CSI implicitly and recovers the transmitted symbols directly. To address channel distortion, a deep learning model is first trained offline using the data generated from the simulation based on the channel statistics and then used for recovering the online transmitted data directly. From our simulation results, the deep learning based approach has the ability to address channel distortions and detect the transmitted symbols with performance comparable to minimum mean-square error (MMSE) estimator. Furthermore, the deep learning based approach is more robust than conventional methods when fewer training pilots are used, the cyclic prefix (CP) is omitted, and nonlinear clipping noise is presented. In summary, deep learning is a promising tool for channel estimation and signal detection in wireless communications with complicated channel distortions and interferences.
End-to-end learning of communications systems is a fascinating novel concept that has so far only been validated by simulations for block-based transmissions. It allows learning of transmitter and receiver implementations as deep neural networks (NNs) that are optimized for an arbitrary differentiable end-to-end performance metric, e.g., block error rate (BLER). In this paper, we demonstrate that over-the-air transmissions are possible: We build, train, and run a complete communications system solely composed of NNs using unsynchronized off-the-shelf software-defined radios (SDRs) and open-source deep learning (DL) software libraries. We extend the existing ideas towards continuous data transmission which eases their current restriction to short block lengths but also entails the issue of receiver synchronization. We overcome this problem by introducing a frame synchronization module based on another NN. A comparison of the BLER performance of the "learned" system with that of a practical baseline shows competitive performance close to 1 dB, even without extensive hyperparameter tuning. We identify several practical challenges of training such a system over actual channels, in particular the missing channel gradient, and propose a two-step learning procedure based on the idea of transfer learning that circumvents this issue.
Several variants of the Long Short-Term Memory (LSTM) architecture for
recurrent neural networks have been proposed since its inception in 1995. In
recent years, these networks have become the state-of-the-art models for a
variety of machine learning problems. This has led to a renewed interest in
understanding the role and utility of various computational components of
typical LSTM variants. In this paper, we present the first large-scale analysis
of eight LSTM variants on three representative tasks: speech recognition,
handwriting recognition, and polyphonic music modeling. The hyperparameters of
all LSTM variants for each task were optimized separately using random search
and their importance was assessed using the powerful fANOVA framework. In
total, we summarize the results of 5400 experimental runs (about 15 years of
CPU time), which makes our study the largest of its kind on LSTM networks. Our
results show that none of the variants can improve upon the standard LSTM
architecture significantly, and demonstrate the forget gate and the output
activation function to be its most critical components. We further observe that
the studied hyperparameters are virtually independent and derive guidelines for
their efficient adjustment.
In this paper, we propose a novel neural network model called RNN Encoder--Decoder that consists of two recurrent neural networks (RNN). One RNN encodes a sequence of symbols into a fixed-length vector representation, and the other decodes the representation into another sequence of symbols. The encoder and decoder of the proposed model are jointly trained to maximize the conditional probability of a target sequence given a source sequence. The performance of a statistical machine translation system is empirically found to improve by using the conditional probabilities of phrase pairs computed by the RNN Encoder--Decoder as an additional feature in the existing log-linear model. Qualitatively, we show that the proposed model learns a semantically and syntactically meaningful representation of linguistic phrases.
Learning to store information over extended time intervals by recurrent backpropagation takes a very long time, mostly because of insufficient, decaying error backflow. We briefly review Hochreiter's (1991) analysis of this problem, then address it by introducing a novel, efficient, gradient-based method called long short-term memory (LSTM). Truncating the gradient where this does not do harm, LSTM can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units. Multiplicative gate units learn to open and close access to the constant error flow. LSTM is local in space and time; its computational complexity per time step and weight is O(1). Our experiments with artificial data involve local, distributed, real-valued, and noisy pattern representations. In comparisons with real-time recurrent learning, back propagation through time, recurrent cascade correlation, Elman nets, and neural sequence chunking, LSTM leads to many more successful runs, and learns much faster. LSTM also solves complex, artificial long-time-lag tasks that have never been solved by previous recurrent network algorithms.
Deep Learning has a wide application in the area of natural language processing and image processing due to its strong ability of generalization. In this paper, we propose a novel neural network structure for jointly optimizing the transmitter and receiver in communication physical layer under fading channels. We build up a convolutional autoencoder to simultaneously conduct the role of modulation, equalization and demodulation. The proposed system is able to design different mapping scheme from input bit sequences of arbitrary length to constellation symbols according to different channel environments. The simulation results show that the performance of neural network based system is superior to traditional modulation and equalization methods in terms of time complexity and bit error rate (BER) under fading channels. The proposed system can also be combined with other coding techniques to further improve the performance. Furthermore, the proposed system network is more robust to channel variation than traditional communication methods.
Recent developments in applying deep learning techniques to train end-to-end communication systems have shown great promise in improving the overall performance of the system. However, most of the current methods for applying deep learning to train physical layer characteristics assume the availability of explicit channel model. Training a neural network requires the availability of functional form all the layers in the network to calculate gradients for optimization. The unavailability of gradients in a physical channel forced previous works to adopt simulation based strategies to train the network and then fine tune only the receiver part with actual channel. In this paper, we present a practical method to train an end-to-end communication system without relying on explicit channel models. By utilizing stochastic perturbation techniques, we show that the proposed method can train a deep learning based communication system in real channel without any assumption on channel models.
We consider detection based on deep learning, and show it is possible to train detectors that perform well, without any knowledge of the underlying channel models. Moreover, when the channel model is known, we demonstrate that it is possible to train detectors that do not require channel state information (CSI). In particular, a technique we call sliding bidirectional recurrent neural network (SBRNN) is proposed for detection where, after training, the detector estimates the data in real-time as the signal stream arrives at the receiver. We evaluate this algorithm, as well as other neural network (NN) architectures, using the Poisson channel model, which is applicable to both optical and chemical communication systems. In addition, we also evaluate the performance of this detection method applied to data sent over a chemical communication platform, where the channel model is difficult to model analytically. We show that SBRNN is computationally efficient, and can perform detection under various channel conditions without knowing the underlying channel model. We also demonstrate that the bit error rate (BER) performance of the proposed SBRNN detector is better than that of a Viterbi detector with imperfect CSI as well as that of other NN detectors that have been previously proposed.
This book is for designers and would-be designers of digital communication systems. The general approach of this book is to extract the common principles underlying a range of media and applications and present them in a unified framework. Digital Communication is relevant to the design of a variety of systems, including voice and video digital cellular telephone, digital CATV distribution, wireless LANs, digital subscriber loop, metallic Ethernet, voiceband data modems, and satellite communication systems.
New in this Third Edition:
New material on recent advances in wireless communications, error-control coding, and multi-user communications has been added. As a result, two new chapters have been added, one on the theory of MIMO channels, and the other on diversity techniques for mitigating fading.
Error-control coding has been rewritten to reflect the current state of the art.
Chapters 6 through 9 from the Second Edition have been reorganized and streamlined to highlight pulse-amplitude modulation, becoming the new Chapters 5 through 7.
Readability is increased by relegating many of the more detailed derivations to appendices and exercise solutions, both of which are included in the book.
Exercises, problems, and solutions have been revised and expanded.
Three chapters from the previous edition have been moved to the book’s Web site to make room for new material.
This book is ideal as a first-year graduate textbook, and is essential to many industry professionals. The book is attractive to both audiences through the inclusion of many practical examples and a practical flavor in the choice of topics.
Digital Communication has a Web site at : http://www.ece.gatech.edu/~barry/digital/, where the reader may find additional information from the Second Edition, other supplementary materials, useful links, a problem solutions manual, and errata.
We present and discuss several novel applications of deep learning (DL) for the physical layer. By interpreting a communications system as an autoencoder, we develop a fundamental new way to think about communications system design as an end-to-end reconstruction task that seeks to jointly optimize transmitter and receiver components in a single process. We show how this idea can be extended to networks of multiple transmitters and receivers and present the concept of radio transformer networks (RTNs) as a means to incorporate expert domain knowledge in the machine learning (ML) model. Lastly, we demonstrate the application of convolutional neural networks (CNNs) on raw IQ samples for modulation classification which achieves competitive accuracy with respect to traditional schemes relying on expert features. The paper is concluded with a discussion of open challenges and areas for future investigation.
Multivariate stochastic optimization plays a major role in the analysis and control of many engineering systems. In almost all real-world optimization problems, it is necessary to use a mathematical algorithm that iteratively seeks out the solution because an analytical (closed-form) solution is rarely available. In this spirit, the "simultaneous perturbation stochastic approximation (SPSA)" method for difficult multivariate optimization problems has been developed. SPSA has recently attracted considerable international attention in areas such as statistical parameter estimation, feedback control, simulation-based optimization, signal and image processing, and experimental design. The essential feature of SPSA - which accounts for its power and relative ease of implementation - is the underlying gradient approximation that requires only two measurements of the objective function regardless of the dimension of the optimization problem. This feature allows for a significant decrease in the cost of optimization, especially in problems with a large number of variables to be optimized.
We introduce Adam, an algorithm for first-order gradient-based optimization
of stochastic objective functions. The method is straightforward to implement
and is based an adaptive estimates of lower-order moments of the gradients. The
method is computationally efficient, has little memory requirements and is well
suited for problems that are large in terms of data and/or parameters. The
method is also ap- propriate for non-stationary objectives and problems with
very noisy and/or sparse gradients. The method exhibits invariance to diagonal
rescaling of the gradients by adapting to the geometry of the objective
function. The hyper-parameters have intuitive interpretations and typically
require little tuning. Some connections to related algorithms, on which Adam
was inspired, are discussed. We also analyze the theoretical convergence
properties of the algorithm and provide a regret bound on the convergence rate
that is comparable to the best known results under the online convex
optimization framework. We demonstrate that Adam works well in practice when
experimentally compared to other stochastic optimization methods.
Generative Adversarial Nets [8] were recently introduced as a novel way to
train generative models. In this work we introduce the conditional version of
generative adversarial nets, which can be constructed by simply feeding the
data, y, we wish to condition on to both the generator and discriminator. We
show that this model can generate MNIST digits conditioned on class labels. We
also illustrate how this model could be used to learn a multi-modal model, and
provide preliminary examples of an application to image tagging in which we
demonstrate how this approach can generate descriptive tags which are not part
of training labels.
The past decade has seen many advances in physical layer wireless communication theory and their implementation in wireless systems. This textbook takes a unified view of the fundamentals of wireless communication and explains the web of concepts underpinning these advances at a level accessible to an audience with a basic background in probability and digital communication. Topics covered include MIMO (multi-input, multi-output) communication, space-time coding, opportunistic communication, OFDM and CDMA. The concepts are illustrated using many examples from real wireless systems such as GSM, IS-95 (CDMA), IS-856 (1 x EV-DO), Flash OFDM and UWB (ultra-wideband). Particular emphasis is placed on the interplay between concepts and their implementation in real systems. An abundant supply of exercises and figures reinforce the material in the text. This book is intended for use on graduate courses in electrical and computer engineering and will also be of great interest to practising engineers.
The statistical characteristics of the fields and signals in the reception of radio frequencies by a moving vehicle are deduced from a scattering propagation model. The model assumes that the field incident on the receiver antenna is composed of randomly phased azimuthal plane waves of arbitrary azimuth angles. Amplitude and phase distributions and spatial correlations of fields and signals are deduced, and a simple direct relationship is established between the signal amplitude spectrum and the product of the incident plane waves' angular distribution and the azimuthal antenna gain.
The coherence of two mobile-radio signals of different frequencies is shown to depend on the statistical distribution of the relative time delays in the arrival of the component waves, and the coherent bandwidth is shown to be the inverse of the spread in time delays.
Wherever possible theoretical predictions are compared with the experimental results. There is sufficient agreement to indicate the validity of the approach. Agreement improves if allowance is made for the nonstationary character of mobile-radio signals.
We present a new method for accelerating matrix multiplication asymptotically. This work builds on recent ideas of Volker Strassen, by using a basic trilinear form which is not a matrix product. We make novel use of the Salem-Spencer Theorem, which gives a fairly dense set of integers with no three-term arithmetic progression. Our resulting matrix exponent is 2.376.
A low-density parity-check code is a code specified by a parity-check matrix with the following properties: each column contains a small fixed number j geq 3 of l's and each row contains a small fixed number k > j of l's. The typical minimum distance of these codes increases linearly with block length for a fixed rate and fixed j . When used with maximum likelihood decoding on a sufficiently quiet binary-input symmetric channel, the typical probability of decoding error decreases exponentially with block length for a fixed rate and fixed j . A simple but nonoptimum decoding scheme operating directly from the channel a posteriori probabilities is described. Both the equipment complexity and the data-handling capacity in bits per second of this decoder increase approximately linearly with block length. For j > 3 and a sufficiently low rate, the probability of error using this decoder on a binary symmetric channel is shown to decrease at least exponentially with a root of the block length. Some experimental results show that the actual probability of decoding error is much smaller than this theoretical bound.
The statistical properties of radio propagation between a mobile unit and a base-station terminal are derived. The power spectrum of the transmission coefficient in the multipath medium is used to determine probability distributions of amplitude and phase, correlations of fields versus time and space at mobile and base stations, level-crossing rates and durations of fades, and random frequency modulation. Duality between the power spectrum and density of time delays is shown. The correlations versus frequency and the coherence bandwidth then follow from the density of time delays. The performance of standard diversity systems is then predicted. There is a review of results previously presented by Clarke [1], however, the derivations given herein utilize expressions of the power spectrum rather than expressions of the component waves. The power-spectral approach, used throughout, allows direct application of previous statistical analyses, particularly those of Rice [2].
Technological evolution and the ever-increasing demand for
higher-quality services give broadcasters a strong incentive to
completely digitize their broadcasting networks. This digitization,
which is already well advanced in many program production areas and
transmission links, now has to be extended to complete the last link in
the broadcast chain; i.e., from broadcast transmitter to consumer
receivers. It is therefore necessary to develop wholly new techniques
for the broadcasting of digitally coded TV programmes. Thus an efficient
baseband digital coding must be combined with a robust digital
modulation and channel coding scheme that can meet the requirements of
every mode of broadcast reception. This article presents the research
work related to the coded orthogonal frequency division multiplex
(COFDM) technology, which has now been completed in the field of digital
radio (DAB), and which is under progress in the field of digital
terrestrial TV
This article is an introduction to the simultaneous perturbation stochastic approximation (SPSA) algorithm for stochastic optimization of multivariate systems. Optimization algorithms play a critical role in the design, analysis, and control of most engineering systems and are in widespread use in the work of APL and other organizations: The future, in fact, will be full of [optimization] algorithms. They are becoming part of almost everything. They are moving up the complexity chain to make entire companies more efficient. They also are moving down the chain as computers spread. (USA Today, 31 Dec 1997) Before presenting the SPSA algorithm, we provide some general background on the stochastic optimization context of interest here
An introduction to deep learning for the physical layer
- T Shea
- J Hoydis
Digital Communication Receivers: Synchronization, Channel Estimation, and Signal Processing
Academic Press Library in Signal Processing
- J K Tugnait