arXiv subjects

Weidong Wang

Publications and source records attributed to Weidong Wang.

48 records · Page 3Linked to original sources

arXiv2020

Machine Learning Framwork for Performance Anomaly in OpenMP Multi-Threaded Systems

Some OpenMP multi-threaded applications increasingly suffer from performance anomaly owning to shared resource contention as well as software- and hardware-related problems. Such performance anomaly can result in failure and inefficiencies, and are among the main challenges in system resiliency. To minimize the impact of performance anomaly, one must quickly and accurately detect and diagnose the performance anomalies that cause the failures. However, it is difficult to identify anomalies in the dynamic and noisy data collected by OpenMP multi-threaded monitoring infrastructures. This paper presents a novel machine learning framework for performance anomaly in OpenMP multi-threaded systems. To evaluate our framework, the NAS Parallel NPB benchmark, EPCC OpenMP micro-benchmark suite, and Jacobi benchmark are used to test the performance of our framework proposed. The experimental results demonstrate that our framework successfully identifies 90.3\% of injected anomalies of OpenMP multi-threaded applications.

cs.DC↗

arXiv2020

Impact and Calibration of Nonlinear Reciprocity Mismatch in Massive MIMO

Time-division-duplexing massive multiple-input multiple-output (MIMO) systems estimate the channel state information (CSI) by leveraging the uplink-downlink channel reciprocity, which is no longer valid when the mismatch arises from the asymmetric uplink and downlink radio frequency (RF) chains. Existing works treat the reciprocity mismatch as constant for simplicity. However, the practical RF chain consists of nonlinear components, which leads to nonlinear reciprocity mismatch. In this work, we examine the impact and the calibration approach of the nonlinear reciprocity mismatch in massive MIMO systems. To evaluate the impact of the nonlinear mismatch, we first derive the closed-form expression of the ergodic achievable rate. Then, we analyze the performance loss caused by the nonlinear mismatch to show that the impact of the mismatch at the base station (BS) side is much larger than that at the user equipment side. Therefore, we propose a calibration method for the BS. During the calibration, polynomial function is applied to approximate the nonlinear mismatch factor, and over-the-air training is employed to estimate the polynomial coefficients. After that, the calibration coefficients are computed by maximizing the downlink achievable rate. Simulation results are presented to verify the analytical results and to show the performance of the proposed calibration approach.

cs.IT↗

arXiv2020

A General Architecture for Behavior Modeling of Nonlinear Power Amplifier using Deep Convolutional Neural Network

Nonlinearity of power amplifier is one of the major limitations to the achievable capacity in wireless transmission systems. Nonlinear impairments are determined by the nonlinear distortions of the power amplifier and modulator imperfections. The Volterra model, several compact Volterra models and neural network models to establish a nonlinear model of power amplifier have all been demonstrated. However, the computational cost of these models increases and their implementation demands more signal processing resources as the signal bandwidth gets wider or the number of carrier aggregation. A completely different approach uses deep convolutional neural network to learn from the training data to figure out the nonlinear distortion. In this work, a low complexity, general architecture based on the deep real-valued convolutional neural network (DRVCNN) is proposed to build the nonlinear behavior of the power amplifier. With each of the multiple inputs equivalent to an input vector, the DRVCNN tensor weights are constructed from training data thanks to the current and historical envelope-dependent terms, I, and Q, which are components of the input. The effectiveness of the general framework in modeling single-carrier and multi-carrier power amplifiers is verified.

eess.SP↗

arXiv2020

Convolutional Neural Network for Behavioral Modeling and Predistortion of Wideband Power Amplifiers

In this paper, we propose a novel behavior model for wideband PAs using a real-valued time-delay convolutional neural network (RVTDCNN). The input data of the model are sorted and arranged as the graph composed of the in-phase and quadrature (I/Q) components and envelope-dependent terms of current and past signals. We design a pre-designed filter using the convolutional layer to extract the basis functions required for the PA forward or reverse modeling. The generated rich basis functions are modeled using a simple fully connected layer. Because of the weight sharing characteristics of the convolutional structure, the strong memory effect does not lead to a obvious increase in the complexity of the model. Meanwhile, the extraction effect of the pre-designed filter also reduces the training complexity of the model. The experimental results show that the performance of the RVTDCNN model is almost the same as the NN models and the multilayer NN models.

eess.SP↗

arXiv2019

Robust Federated Learning with Noisy Communication

Federated learning is a communication-efficient training process that alternates between local training at the edge devices and averaging the updated local model at the central server. Nevertheless, it is impractical to achieve a perfect acquisition of the local models in wireless communication due to noise, which also brings serious effects on federated learning. To tackle this challenge, we propose a robust design for federated learning to alleviate the effects of noise in this paper. Considering noise in the two aforementioned steps, we first formulate the training problem as a parallel optimization for each node under the expectation-based model and the worst-case model. Due to the non-convexity of the problem, a regularization for the loss function approximation method is proposed to make it tractable. Regarding the worst-case model, we develop a feasible training scheme which utilizes the sampling-based successive convex approximation algorithm to tackle the unavailable maxima or minima noise condition and the non-convex issue of the objective function. Furthermore, the convergence rates of both new designs are analyzed from a theoretical point of view. Finally, the improvement of prediction accuracy and the reduction of loss function are demonstrated via simulations for the proposed designs.

cs.LG↗

arXiv2019

An End-to-End Block Autoencoder For Physical Layer Based On Neural Networks

Deep Learning has been widely applied in the area of image processing and natural language processing. In this paper, we propose an end-to-end communication structure based on autoencoder where the transceiver can be optimized jointly. A neural network roles as a combination of channel encoder and modulator. In order to deal with input sequences parallelly, we introduce block scheme, which means that the autoencoder divides the input sequence into a series of blocks. Each block contains fixed number of bits for encoding and modulating operation. Through training, the proposed system is able to produce the modulated constellation diagram of each block. The simulation results show that our autoencoder performs better than other autoencoder-based systems under additive Gaussian white noise (AWGN) and fading channels. We also prove that the bit error rate (BER) of proposed system can achieve an acceptable range with increasing the number of symbols.

cs.IT↗

arXiv2018

Hybrid Beamforming Design and Performance with Imperfect Phase Shifters in Multiuser Millimeter Wave Systems

Hybrid beamforming (HBF) includes analog beamforming with phase shifted array in RF domain and digital beamforming in baseband domain. Phase shifted array is usually made up with a large amount of phase shifters. Limited to the manufacturing techniques, phase shifters are unavoidable to have phaseshifting error and gain error. In the paper we study the influence of imperfect phase shifters upon the performance of multiuser mmWave massive MIMO systems1 with the wireless channel modeled by Rice fading. We derive the upper bound of the achievable sum rate with imperfect phase shifters, which can be very tight when the number of the antennas at the base station is much larger than the number of users. Results show that there is a performance ceiling due to phase-shifting error and gain error. Then we propose a novel channel estimation and hybrid beamforming method to settle this problem with low training overhead. It performs well when the channel paths are few and the channel Rician K-factor is large, which can be easily satisfied in outdoor mmWave communication environment. We further make a tradeoff between the performance and the training overhead so that the algorithm can work well in more kinds of propagation environment.

cs.IT↗

arXiv2018

TCP Decoupling for Next Generation Communication System

In traditional networks, interfaces of network nodes are duplex. But, emerging communication technologies such as visible light communication, millimeter-wave communications, can only provide a unidirectional interface when cost is limited. It's urgent to find effective solutions to utilize such new unidirectional communication skills. Decoupling implies separating one single resource to two independent resources. This idea can be applied at physical layer, link layer, network layer, even transport layer. TCP decoupling is an end to end solution provided at transport layer. With decoupled TCP, two distinct unidirectional path can be created to meet the requirements of reliable information transfer. However, it is not an easy task to decouple a bidirectional logical path at transport layer. In this paper, we dwell on the idea of TCP decoupling. Advantages of decoupling at transport layer are analyzed also. In addition, an experiment is carried out to figure out how to implement a decouple TCP. Our results show decoupling at transport layer is possible and the modified protocol is available.

cs.NI↗

arXiv2016

Stopping Condition for Greedy Block Sparse Signal Recovery

For greedy block sparse recovery where the sparsity level is unknown, we derive a stopping condition to stop the iteration process. Focused on the block orthogonal matching pursuit (BOMP) algorithm, we model the energy of residual signals at each iteration from a probabilistic perspective. At the iteration when the last supporting block is detected, the resulting energy of residual signals is supposed to suffer an obvious decrease. Based on this, we stop the iteration process when the energy of residual signals is below a given threshold. Compared with other approaches, our derived condition works well for the BOMP recovery. What is more, we promote our approach to the interference cancellation based BOMP (ICBOMP) recovery in paper [1]. Simulation results show that our derived condition can save many unnecessary iterations and at the same time guarantees a favorable recovery accuracy, both for the BOMP and ICBOMP recoveries.

cs.IT↗

arXiv2016

An Intelligent QoS Identification for Untrustworthy Web Services Via Two-phase Neural Networks

QoS identification for untrustworthy Web services is critical in QoS management in the service computing since the performance of untrustworthy Web services may result in QoS downgrade. The key issue is to intelligently learn the characteristics of trustworthy Web services from different QoS levels, then to identify the untrustworthy ones according to the characteristics of QoS metrics. As one of the intelligent identification approaches, deep neural network has emerged as a powerful technique in recent years. In this paper, we propose a novel two-phase neural network model to identify the untrustworthy Web services. In the first phase, Web services are collected from the published QoS dataset. Then, we design a feedforward neural network model to build the classifier for Web services with different QoS levels. In the second phase, we employ a probabilistic neural network (PNN) model to identify the untrustworthy Web services from each classification. The experimental results show the proposed approach has 90.5% identification ratio far higher than other competing approaches.

cs.SE↗

arXiv2015

HBTM: A Heartbeat-based Behavior Detection Mechanism for POSIX Threads and OpenMP Applications

Extreme-scale computing involves hundreds of millions of threads with multi-level parallelism running on large-scale hierarchical and heterogeneous hardware. In POSIX threads and OpenMP applications, some key behaviors occurring in runtime such as thread failure, busy waiting, and exit need to be accurately and timely detected. However, for the most of these applications, there are lack of unified and efficient detection mechanisms to do this. In this paper, a heartbeat-based behavior detection mechanism for POSIX threads (Pthreads) and OpenMP applications (HBTM) is proposed. In the design, two types of implementations are conducted, centralized and decentralized respectively. In both implementations, unified API has been designed to guarantee the generality of the mechanism. Meanwhile, a ring-based detection algorithm is designed to ease the burden of the centra thread at runtime. To evaluate the mechanism, the NAS Parallel Benchmarks (NPB) are used to test the performance of the HBTM. The experimental results show that the HBTM supports detection of behaviors of POSIX threads and OpenMP applications while acquiring a short latency and near 1% overhead.

cs.DC↗

arXiv2013

Low complexity resource allocation for load minimization in OFDMA wireless networks

To cope with the ever increasing demand for bandwidth, future wireless networks will be designed with reuse distance equal to one. This scenario requires the implementation of techniques able to manage the strong multiple access interference each cell generates towards its neighbor cells. In particular, low complexity and reduced feedback are important requirements for practical algorithms. In this paper we study an allocation problem for OFDMA networks formulated with the objective of minimizing the load of each cell in the system subject to the constraint that each user meets its target rate. We decompose resource allocation into two sub-problems: channel allocation under deterministic power assignment and continuous power assignment optimization. Channel allocation is formulated as the problem of finding the maximum weighted independent set (MWIS) in graph theory. In addition, we propose a minimal weighted-degree greedy (MWDG) algorithm of which the approximation factor is analyzed. For power allocation, an iterative power reassignment algorithm (DPRA) is proposed. The control information requested to perform the allocation is limited and the computational burden is shared between the base station and the user equipments. Simulations have been carried out under constant bit rate traffic model and the results have been compared with other allocation schemes of similar complexity. MWDG has excellent performance and outperforms all other techniques.

cs.IT↗