Ye Xue

Ye Xue

Auto-generated profile, based on publicly available information

Ye Xue

School of Intelligent Systems Engineering, Shenzhen Campus of Sun Yat sen University, Shenzhen, China

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 April 2022 in IEEE Internet of Things Journal
Verified
Journal Article
Closed access
YXYe Xue
Ye Xue
LSLiqun Su
Liqun Su
VLVincent K. N. Lau
Vincent K. N. Lau
Federated learning (FL) is a machine learning framework, where multiple distributed edge Internet of Things (IoT) devices collaboratively train a model under the orchestration of a central server while keeping the training data distributed on the IoT devices. FL can mitigate the privacy risks and costs from data collection in traditional centralized machine learning. However, the deployment of standard FL is hindered by the expense of the communication of the gradients from the devices to the server. Hence, many gradient compression methods have been proposed to reduce the communication cost. However, the existing methods ignore the structural correlations of the gradients and, therefore, lead to a large compression loss which will decelerate the training convergence. Moreover, many of the existing compression schemes do not enable over-the-air aggregation and, hence, require huge communication resources. In this work, we propose a gradient compression scheme, named FedOComp, which leverages the correlations of the stochastic gradients in FL systems for efficient compression of the high-dimension gradients with over-the-air aggregation. The proposed design can achieve a smaller deceleration of the training convergence compared to other gradient compression methods since the compression kernel exploits the structural correlations of the gradients. It also directly enables over-the-air aggregation to save communication resources. The derived convergence analysis and simulation results further illustrate that under the same power cost, the proposed scheme has a much faster convergence rate and higher test accuracy compared to existing baselines.

8 June 2023 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
YXYe Xue
Ye Xue
VLVincent Lau
Vincent Lau
Low-rank model compression is a widely used technique for reducing the computational load when training machine learning models. However, existing methods often rely on relaxing the low-rank constraint of the model weights using a regularized nuclear norm penalty, which requires an appropriate hyperparameter that can be difficult to determine in practice. Furthermore, existing compression techniques are not directly applicable to efficient over-the-air (OTA) aggregation in federated learning (FL) systems for distributed Internet-of-Things (IoT) scenarios. In this paper, we propose a novel manifold optimization formulation for low-rank model compression in FL that does not relax the low-rank constraint. Our optimization is conducted directly over the low-rank manifold, guaranteeing that the model is exactly low-rank. We also introduce a consensus penalty in the optimization formulation to support OTA aggregation. Based on our optimization formulation, we propose an alternating Riemannian optimization algorithm with a precoder that enables efficient OTA aggregation of low-rank local models without sacrificing training performance. Additionally, we provide convergence analysis in terms of key system parameters and conduct extensive experiments with real-world datasets to demonstrate the effectiveness of our proposed Riemannian low-rank model compression scheme compared to various state-of-the-art baselines.

2 November 2020 in IEEE Transactions on Wireless Communications
Verified
Journal Article
Closed access
YXYe Xue
Ye Xue
YSYifei Shen
Yifei Shen
VLVincent K. N. Lau
Vincent K. N. Lau
JZJun Zhang
Jun Zhang
KLKhaled B. Letaief
Khaled B. Letaief
Massive MIMO has been regarded as a key enabling technique for 5G and beyond networks. Nevertheless, its performance is limited by the large overhead needed to obtain the high-dimensional channel information. To reduce the huge training overhead associated with conventional pilotaided designs, we propose a novel blind data detection method by leveraging the channel sparsity and data concentration properties. Specifically, we propose a novel ℓ3-norm-based formulation to recover the data without channel estimation. We prove that the global optimal solution to the proposed formulation can be made arbitrarily close to the transmitted data up to a phase-permutation ambiguity. We then propose an efficient parameter-free algorithm to solve the ℓ3-norm problem and resolve the phase-permutation ambiguity. We also derive the convergence rate in terms of key system parameters such as the number of transmitters and receivers, the channel noise power, and the channel sparsity level. Numerical experiments will show that the proposed scheme has superior performance with low computational complexity.

24 February 2020 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
CZChuan Zhang
Chuan Zhang
JJJiejun Jin
Jiejun Jin
YXYe Xue
Ye Xue
XTXiaosi Tan
Xiaosi Tan
CSChristoph Studer
Christoph Studer
XYXiaohu You
Xiaohu You
Massive multi-user (MU) multiple-input multiple-output (MIMO) is an enabling technology for next-generation wireless systems as it provides significant improvements in terms of spectral efficiency and per-user peak data rates compared to existing, small-scale MIMO. The presence of increasing antennas at the infrastructure base-station results, however, in high computational complexity for data detection, which requires low-complexity algorithms and efficient hardware designs. Descent search (DS) algorithms have been proposed to reduce complexity of data detection and enable efficient hardware implementations. However, their error-rate performance is not always satisfactory for channels that exhibit correlation or for systems with small base-station to user antenna ratios. In this paper, we propose a linear inequality constraint quadratic programming (LICQP) model, which enables the design of a constrained DS (CDS) detector that outperforms existing DS algorithms. To further improve the performance of CDS in the presence of challenging channel conditions, we propose a universal pre-conditioner. Theoretical analysis and numerical results indicate that CDS with pre-conditioning reduces the complexity over the state-of-the-art (SOA) by 60%. We develop a reference Xilinx Virtex-7 FPGA design, which demonstrates that our implementation achieves 1.75 × higher throughput (35 Mbps in a 128 antenna, 8 user system) at comparable hardware resource utilization.

25 September 2023
Verified
Conference Paper
Closed access
XNXinzhi Ning
Xinzhi Ning
SZShutao Zhang
Shutao Zhang
YXYe Xue
Ye Xue
Xi ZhengXi Zheng
Xi Zheng
QSQingjiang Shi
Qingjiang Shi
TCTsung-Hui Chang
Tsung-Hui Chang
The design of an effective codebook is crucial for the fifth generation (5G) systems, particularly in the context of massive multiple-input-multiple-output (MIMO) and millimeter wave communication. However, existing codebook design methods that rely on the instantaneous channel impulse response (CIR) face challenges in adapting to dynamic channels that undergo small-scale fading changes. To address this issue, this paper proposes a novel two-stage codebook design framework based solely on the reference signal receiving power (RSRP), which provides large-scale statistical information about environment attenuation. The proposed approach involves two stages: first, the angular power spectrum (APS) is estimated through sparse Bayesian learning (SBL), which uses the statistical relationship between APS and RSRP. Second, an iterative clustering and optimization algorithm is used to determine the optimal codebook based on the estimated APS. Experimental results demonstrate that the proposed codebook design framework with a smaller number of beams can adapt to localized environment well and achieves better performance than the classical DFT codebook.

1 July 2021 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
YXYe Xue
Ye Xue
VLVincent K. N. Lau
Vincent K. N. Lau
SCSongfu Cai
Songfu Cai
Sparse coding is a class of unsupervised methods for learning a sparse representation of the input data in the form of a linear combination of a dictionary and a sparse code. This learning framework has led to state-of-the-art results in various image and video processing tasks. However, classical methods learn the dictionary and the sparse code based on alternative optimizations, usually without theoretical guarantees for either optimality or convergence due to the non-convexity of the problem. Recent works on sparse coding with a complete dictionary provide strong theoretical guarantees thanks to the development of non-convex optimization. However, initial non-convex approaches [14] learn the dictionary in the sparse coding problem sequentially in an atom-by-atom manner, which leads to a long execution time. More recent works [57] seek to directly learn the entire dictionary at once, which substantially reduces the execution time. However, the associated recovery performance is degraded with a finite number of data samples. In this paper, we propose an efficient sparse coding scheme with a two-stage optimization. The proposed scheme leverages the global and local Riemannian geometry of the two-stage optimization problem and facilitates fast implementation for a superb dictionary recovery performance by a finite number of samples without atom-by-atom calculation. We further prove that the proposed scheme can exactly recover any atom in the target dictionary with a finite number of samples if it is adopted to recover one atom of the dictionary. An application on wireless sensor data compression is also proposed. Experiments on both synthetic and real-world data verify the efficiency and robustness of the proposed scheme. The codes to reproduce the experimental results are available online at https://github.com/yokoxue/HRP.