Jiaying Zhang

Jiaying Zhang

Auto-generated profile, based on publicly available information

Jiaying Zhang

College of Information Science and Electronic Engineering, Zhejiang University, Hangzhou, China

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

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

Representative Publications

View All Publications
3 November 2025 in IEEE Communications Letters
Verified
Journal Article
Closed access
JZJiaying Zhang
Jiaying Zhang
QHQiushuo Hou
Qiushuo Hou
XSXin Su
Xin Su
Guanding YuGuanding Yu
Guanding Yu
Beamforming is crucial in wireless communication given its ability to significantly suppress interference and enhance capacity in multi-user multi-input multi-output (MU-MIMO) system. Traditional mathematical optimization methods generally achieve optimal beamforming but at the cost of high computational complexity. Recently, artificial intelligence (AI)-based models have emerged as effective techniques to overcome the high complexity in beamforming design. However, those models are poorly calibrated, generally tending to be overconfident. Due to this inherent poor calibration, the AI-based beamformer does not come with coverage guarantee with respect to the optimal solution by the optimization methods during deployment. To tackle this issue, we propose a conformal-prediction-based method to calibrate the prediction output produced by AI-based models, with formal coverage guarantee. Experimental results on the MU-MIMO system demonstrate the effectiveness of the proposed method.

27 January 2025 in IEEE Communications Letters
Verified
Journal Article
Closed access
JZJiaying Zhang
Jiaying Zhang
QHQiushuo Hou
Qiushuo Hou
Guanding YuGuanding Yu
Guanding Yu
In recent years, machine learning (ML) tasks have been widely deployed at the edge of wireless networks, e.g., autonomous cars and tactile robots. However, the impairments of wireless channels between devices, such as fading and noise, deteriorate the effectiveness of ML inference tasks. In this work, we propose an efficient framework with an adaptive power control mechanism, which considers the constraint of the limited energy budget of edge devices. To guarantee the inference performance of ML tasks that are transmitted through wireless channels, we design hypernetworks with meta-parameters. The hypernetwork takes the context, such as the network condition, as the input and outputs the parameters of the power control network and artificial intelligence (AI) model. The training loss is designed by minimizing the trade-off between inference performance and energy consumption. Simulation results verify the effectiveness of the proposed adaptive inference framework on energy saving while ensuring the accuracy of inferring ML tasks.

24 October 2024 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
JZJiaying Zhang
Jiaying Zhang
QHQiushuo Hou
Qiushuo Hou
HGHuiguo Gao
Huiguo Gao
Guanding YuGuanding Yu
Guanding Yu
Sidelink relay technique is an effective technique for coverage enhancement and data rate improvement, which has recently achieved success in diversified wireless applications. The main bottleneck of applying sidelink to multiple-user multiple-input multiple-output (MU-MIMO) system is how to overcome interference among user equipments (UEs). In this paper, we consider an MU-MIMO relay network with both downlink and uplink UEs, where sidelink reuses the same resources of uplink UEs. To maximize the sum rate of downlink UEs, we propose a novel joint resource block allocation and beamforming design algorithm for the considered MU-MIMO relay network. Specifically, we first formulate an optimization problem with the constraint on the sum-rate degradation of uplink UEs. However, the formulated mixed integer nonlinear programming problem is NP-hard. In order to solve this problem, we first relax the discrete variables and then transform the continuous problem into a convex one based on the weighted minimum mean square error (WMMSE) algorithm. Meanwhile, we prove the equivalence between the aforementioned two problems. Finally, we utilize the block coordinate descent method to solve the transformed problem and obtain the resource block allocation and beamforming design. Simulation results validate the effectiveness of the proposed algorithm and demonstrate the importance of joint optimization.