arXiv subjects

Pengcheng Zhu

Publications and source records attributed to Pengcheng Zhu.

69 records · Page 4Linked to original sources

arXiv2022

The Complexity of Quantum Circuit Mapping with Fixed Parameters

A quantum circuit must be preprocessed before implementing on NISQ devices due to the connectivity constraint. Quantum circuit mapping (QCM) transforms the circuit into an equivalent one that is compliant with the NISQ device's architecture constraint by adding SWAP gates. The QCM problem asks the minimal number of auxiliary SWAP gates, and is NP-complete. The complexity of QCM with fixed parameters is studied in the paper. We give an exact algorithm for QCM, and show that the algorithm runs in polynomial time if the NISQ device's architecture is fixed. If the number of qubits of the quantum circuit is fixed, we show that the QCM problem is NL-complete by a reduction from the undirected shortest path problem. Moreover, the fixed-parameter complexity of QCM is W[1]-hard when parameterized by the number of qubits of the quantum circuit. We prove the result by a reduction from the clique problem. If taking the depth of the quantum circuits and the coupling graphs as parameters, we show that the QCM problem is still NP-complete over shallow quantum circuits, and planar, bipartite and degree bounded coupling graphs.

quant-ph↗

arXiv2022

Content Popularity Prediction Based on Quantized Federated Bayesian Learning in Fog Radio Access Networks

In this paper, we investigate the content popularity prediction problem in cache-enabled fog radio access networks (F-RANs). In order to predict the content popularity with high accuracy and low complexity, we propose a Gaussian process based regressor to model the content request pattern. Firstly, the relationship between content features and popularity is captured by our proposed model. Then, we utilize Bayesian learning to train the model parameters, which is robust to overfitting. However, Bayesian methods are usually unable to find a closed-form expression of the posterior distribution. To tackle this issue, we apply a stochastic variance reduced gradient Hamiltonian Monte Carlo (SVRG-HMC) method to approximate the posterior distribution. To utilize the computing resources of other fog access points (F-APs) and to reduce the communications overhead, we propose a quantized federated learning (FL) framework combining with Bayesian learning. The quantized federated Bayesian learning framework allows each F-AP to send gradients to the cloud server after quantizing and encoding. It can achieve a tradeoff between prediction accuracy and communications overhead effectively. Simulation results show that the performance of our proposed policy outperforms the existing policies.

cs.LG↗

arXiv2022

Learn2Sing 2.0: Diffusion and Mutual Information-Based Target Speaker SVS by Learning from Singing Teacher

Building a high-quality singing corpus for a person who is not good at singing is non-trivial, thus making it challenging to create a singing voice synthesizer for this person. Learn2Sing is dedicated to synthesizing the singing voice of a speaker without his or her singing data by learning from data recorded by others, i.e., the singing teacher. Inspired by the fact that pitch is the key style factor to distinguish singing from speaking voice, the proposed Learn2Sing 2.0 first generates the preliminary acoustic feature with averaged pitch value in the phone level, which allows the training of this process for different styles, i.e., speaking or singing, share same conditions except for the speaker information. Then, conditioned on the specific style, a diffusion decoder, which is accelerated by a fast sampling algorithm during the inference stage, is adopted to gradually restore the final acoustic feature. During the training, to avoid the information confusion of the speaker embedding and the style embedding, mutual information is employed to restrain the learning of speaker embedding and style embedding. Experiments show that the proposed approach is capable of synthesizing high-quality singing voice for the target speaker without singing data with 10 decoding steps.

cs.SD↗

arXiv2022

Spectral Efficiency of Unicast and Multigroup Multicast Transmission in Cell-free Distributed Massive MIMO Systems

In this paper, we consider a joint unicast and multi-group multicast cell-free distributed massive multiple-input multiple-output (MIMO) system, while accounting for co-pilot assignment strategy based channel estimation, pilot contamination and different precoding schemes. Under the co-pilot assignment strategy, we derive the minimum-mean-square error (MMSE) channel state information (CSI) estimation for unicast and multicast users. Given the acquired CSI, the closed-form expressions for downlink achievable rates with maximum ratio transmission (MRT), zero-forcing (ZF) and MMSE beamforming are derived. Based on these expressions, we propose an efficient power allocation scheme by solving a multi-objective optimization problem (MOOP) between maximizing the minimum spectral efficiency (SE) of multicast users and maximizing the average SE of unicast users with non-dominated sorting genetic algorithm II (NSGA-II). Moreover, the MOOP is converted into a deep learning (DL) problem and solved by an unsupervised learning method to further promote computational efficiency. Numerical results verify the accuracy of the derived closed-form expressions and the effectiveness of the joint unicast and multigroup multicast transmission scheme in cell-free distributed massive MIMO systems. The SE analysis under various system parameters and the trade-off regions between these two conflicting optimization objectives offers numerous flexibilities for system optimization.

cs.IT↗

arXiv2022

VISinger: Variational Inference with Adversarial Learning for End-to-End Singing Voice Synthesis

In this paper, we propose VISinger, a complete end-to-end high-quality singing voice synthesis (SVS) system that directly generates audio waveform from lyrics and musical score. Our approach is inspired by VITS, which adopts VAE-based posterior encoder augmented with normalizing flow-based prior encoder and adversarial decoder to realize complete end-to-end speech generation. VISinger follows the main architecture of VITS, but makes substantial improvements to the prior encoder based on the characteristics of singing. First, instead of using phoneme-level mean and variance of acoustic features, we introduce a length regulator and a frame prior network to get the frame-level mean and variance on acoustic features, modeling the rich acoustic variation in singing. Second, we further introduce an F0 predictor to guide the frame prior network, leading to stabler singing performance. Finally, to improve the singing rhythm, we modify the duration predictor to specifically predict the phoneme to note duration ratio, helped with singing note normalization. Experiments on a professional Mandarin singing corpus show that VISinger significantly outperforms FastSpeech+Neural-Vocoder two-stage approach and the oracle VITS; ablation study demonstrates the effectiveness of different contributions.

eess.AS↗

arXiv2022

One-shot Voice Conversion For Style Transfer Based On Speaker Adaptation

One-shot style transfer is a challenging task, since training on one utterance makes model extremely easy to over-fit to training data and causes low speaker similarity and lack of expressiveness. In this paper, we build on the recognition-synthesis framework and propose a one-shot voice conversion approach for style transfer based on speaker adaptation. First, a speaker normalization module is adopted to remove speaker-related information in bottleneck features extracted by ASR. Second, we adopt weight regularization in the adaptation process to prevent over-fitting caused by using only one utterance from target speaker as training data. Finally, to comprehensively decouple the speech factors, i.e., content, speaker, style, and transfer source style to the target, a prosody module is used to extract prosody representation. Experiments show that our approach is superior to the state-of-the-art one-shot VC systems in terms of style and speaker similarity; additionally, our approach also maintains good speech quality.

eess.AS↗

arXiv2022

Opencpop: A High-Quality Open Source Chinese Popular Song Corpus for Singing Voice Synthesis

This paper introduces Opencpop, a publicly available high-quality Mandarin singing corpus designed for singing voice synthesis (SVS). The corpus consists of 100 popular Mandarin songs performed by a female professional singer. Audio files are recorded with studio quality at a sampling rate of 44,100 Hz and the corresponding lyrics and musical scores are provided. All singing recordings have been phonetically annotated with phoneme boundaries and syllable (note) boundaries. To demonstrate the reliability of the released data and to provide a baseline for future research, we built baseline deep neural network-based SVS models and evaluated them with both objective metrics and subjective mean opinion score (MOS) measure. Experimental results show that the best SVS model trained on our database achieves 3.70 MOS, indicating the reliability of the provided corpus. Opencpop is released to the open-source community WeNet, and the corpus, as well as synthesized demos, can be found on the project homepage.

cs.SD↗

arXiv2022

On Performance Loss of DOA Measurement Using Massive MIMO Receiver with Mixed-ADCs

High hardware cost and high power consumption of massive multiple-input and multiple output (MIMO) are two challenges for the future wireless communications including beyond fifth generation (B5G) and sixth generation (6G). Adopting the low-resolution analog-to-digital converter (ADC) is viewed as a promising solution. Additionally, the direction of arrival (DOA) estimation is an indispensable technology for beam alignment and tracking in massive MIMO systems. Thus, in this paper, the performance of DOA estimation with mixed-ADC structure is firstly investigated. The Cramer-Rao lower bound (CRLB) for this architecture is derived based on the additive quantization noise model. Eventually, a performance loss factor and the associated energy efficiency factor is defined for analysis in detail. Simulation results show that the mixed-ADC architecture can strike a good balance among performance loss, circuit cost and energy efficiency. More importantly, just a few bits (up to 4 bits) of low-resolution ADCs can achieve a satisfactory performance for DOA measurement.

cs.IT↗

arXiv2021

A Variation-Aware Quantum Circuit Mapping Approach Based on Multi-agent Cooperation

The quantum circuit mapping approach is an indispensable part of the software stack for the noisy intermediatescale quantum (NISQ) device. It has a significant impact on the reliability of computational tasks on NISQ devices. To improve the overall fidelity of physical circuits, we propose a quantum circuit mapping method based on multi-agent cooperation. This approach considers the Spatio-temporal variation of quantum operation quality on the NISQ device when inserting ancillary operation. It consists of two core components: the qubit placement algorithm and the qubit routing method. The qubit placement algorithm exploits the iterated local search framework to find a desirable initial mapping for the reduced symmetric form of the original circuit. The qubit routing method generates the physical circuit through multi-agent communication and collaboration. Each agent inserts the ancillary gates independently according to its environment state. The quality of the physical circuit evolves according to an information-exchanging mechanism between agents, which combines the local search and global search. To experiment on the benchmark circuits (with hundreds of quantum gates) beyond the capacity of current NISQ devices, we build a noisy simulator with gate error 10x lower than that of the latest NISQ device of IBM. The experimental results confirm the performance of our approach in improving circuit fidelity. Compared with the stateof-the-art method, our method can improve the success rate by 25.86% on average and 95.42% at maximum.

quant-ph↗

arXiv2021

Spectral Efficiency Analysis of Cell-free Distributed Massive MIMO Systems with Imperfect Covariance Matrix

In this paper, the impacts of imperfect channel covariance matrix on the spectral efficiency (SE) of cell-free distributed massive multiple-input multiple-output (MIMO) systems are analyzed. We propose to estimate the channel covariance matrix by alternately using the assigned pilots and their phase-shifted pilots in different coherent blocks, which improves the accuracy of channel estimation with imperfect covariance matrix and reduces pilot overhead. Under this scheme, the closed-form expressions of SE with maximum ratio combination (MRC) and zero-forcing (ZF) receivers are derived, which enables us to select key parameters for better system performance. Simulation results verify the effectiveness of the proposed channel estimation method and the accuracy of the derived closed-form expressions. When more coherent blocks are used to estimate the covariance matrix, we can get better system performance. Moreover, some insightful conclusions are arrived at from the SE comparisons between different receiving schemes (ZF and MRC) and different pilot allocation schemes (orthogonal pilot and pilot reuse).

cs.IT↗

arXiv2021

Privacy-preserving Channel Estimation in Cell-free Hybrid Massive MIMO Systems

We consider a cell-free hybrid massive multiple-input multiple-output (MIMO) system with $K$ users and $M$ access points (APs), each with $N_a$ antennas and $N_r< N_a$ radio frequency (RF) chains. When $K\ll M{N_a}$, efficient uplink channel estimation and data detection with reduced number of pilots can be performed based on low-rank matrix completion. However, such a scheme requires the central processing unit (CPU) to collect received signals from all APs, which may enable the CPU to infer the private information of user locations. We therefore develop and analyze privacy-preserving channel estimation schemes under the framework of differential privacy (DP). As the key ingredient of the channel estimator, two joint differentially private noisy matrix completion algorithms based respectively on Frank-Wolfe iteration and singular value decomposition are presented. We provide an analysis on the tradeoff between the privacy and the channel estimation error. In particular, we show that the estimation error can be mitigated while maintaining the same privacy level by increasing the payload size with fixed pilot size; and the scaling laws of both the privacy-induced and privacy-independent error components in terms of payload size are characterized. Simulation results are provided to further demonstrate the tradeoff between privacy and channel estimation performance.

cs.IT↗

arXiv2019

Joint Sparse Beamforming and Power Control for Large-scale DAS with Network-Assisted Full Duplex

In this paper, we studied the joint sparse beamforming and power control for network-assisted full duplex (NAFD) in large-scale distributed antenna system (L-DAS), where the remote antenna units either be operating in half-duplex mode or full-duplex mode are all connected to the central processing unit (CPU) via high-speed backhaul links. With joint processing at CPU, NAFD could achieve truly flexible duplex, including flexible half-duplex and full-duplex. Cross-link interference and the finite-capacity backhaul are the main problems of NAFD in L-DAS. To solve these problems, we aim to maximize the aggregated rate of uplink and downlink subject to quality of service constraints and backhaul constraints. Two approaches have been proposed to solve the optimization problem, where the first approach converts the object function to the difference of two convex functions with semi-definite relax (SDR) and then an iterative SDR-block coordinate descent method is applied to solve the problem. The second method is based on sequential parametric convex approximation. Simulation results have shown that the proposed algorithms yield a higher rate gain compared to the traditional time-division duplex scheme.

cs.IT↗

arXiv2019

Power Minimization for Wireless Backhaul Based Ultra-Dense Cache-enabled C-RAN

This correspondence paper investigates joint design of small base station (SBS) clustering, multicast beamforming for access and backhaul links, as well as frequency allocation in backhaul transmission to minimize the total power consumption for wireless backhaul based ultra-dense cache-enabled cloud radio access network (C-RAN). To solve this nontrivial problem, we develop a low-complexity algorithm, which is a combination of smoothed ${\ell _0}{\text{-norm}}$ approximation and convex-concave procedure. Simulation results show that the proposed algorithm converges fast and greatly reduces the backhaul traffic.

eess.SP↗

arXiv2019

Performance of Network-Assisted Full-Duplex for Cell-Free Massive MIMO

Network assisted full-duplex (NAFD) is a spatial-division duplex technique for future wireless networks with cell-free massive multiple-input multiple-output (CF massive MIMO) network, where a large number of remote antenna units (RAUs), either using half-duplex or full-duplex, jointly support truly flexible duplex including time-division duplex, frequency-division duplex and full duplex on demand of uplink and downlink traffic by using network MIMO methods. With NAFD, bi-directional data rates of the wireless network could be increased and end-to-end delay could be reduced. In this paper, the spectral efficiency of NAFD communications in CF massive MIMO network with imperfect channel state information (CSI) is investigated under spatial correlated channels. Based on large dimensional random matrix theory, the deterministic equivalents for the uplink sum-rate with minimum-mean-square-error (MMSE) receiver as well as the downlink sum-rate with zero-forcing (ZF) and regularized zero-forcing (RZF) beamforming are derived. Numerical results show that under various environmental settings, the deterministic equivalents are accurate in both a large-scale system and system with a finite number of antennas. It is also shown that with the downlink-to-uplink interference cancellation, the uplink spectral efficiency of CF massive MIMO with NAFD could be improved. The spectral efficiencies of NAFD with different duplex configurations such as in-band full-duplex, and half-duplex are compared. With the same total numbers of transmit and receive antennas, NAFD with half-duplex RAUs offers a higher spectral efficiency. To alleviate the uplink-to-downlink interference, a novel genetic algorithm based user scheduling strategy (GAS) is proposed. Simulation results show that the achievable downlink sum-rate by using the GAS is greatly improved compared to that by using the random user scheduling.

cs.IT↗

arXiv2019

Deep Learning Based Pilot Design for Multi-user Distributed Massive MIMO Systems

This letter proposes a deep learning based pilot design scheme to minimize the sum mean square error (MSE) of channel estimation for multi-user distributed massive multiple-input multiple-output (MIMO) systems. The pilot signal of each user is expressed as a weighted superposition of orthonormal pilot sequence basis, where the power assigned to each pilot sequence is the corresponding weight. A multi-layer fully connected deep neural network (DNN) is designed to optimize the power allocated to each pilot sequence to minimize the sum MSE, which takes the channel large-scale fading coefficients as input and outputs the pilot power allocation vector. The loss function of the DNN is defined as the sum MSE, and we leverage the unsupervised learning strategy to train the DNN. Simulation results show that the proposed scheme achieves better sum MSE performance than other methods with low complexity.

eess.SP↗