Yijie Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Yijie Chen has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Wenchao Xia
Jiangsu Key Laboratory of Wireless Communications and IoT, Nanjing University of Posts and Telecommunications, Nanjing, China
4 shared publications - Hongbo Zhu
Nanjing University of Posts and Telecommunications
2 shared publications - Jun Zhang
Nanjing University of Posts and Telecommunications
2 shared publications - Changwei Zhang
Purple Mountain Laboratories, Nanjing, China
1 shared publication - Cheng Zhang
Pervasive Communications Center, Purple Mountain Laboratories, Nanjing, China
1 shared publication - Gan Zheng
University of Warwick
1 shared publication
Latest Publications
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15, 5432-5436
Anti-Jamming Hybrid Beamforming for Millimeter-Wave Multi-Subarray Systems
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15, 5432-5436
Anti-Jamming Hybrid Beamforming for Millimeter-Wave Multi-Subarray Systems
- 11 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (11) , 18125-18136
Joint Precoding and Fronthaul Compression for Cell-Free MIMO With Hybrid Topology
Cell-free multiple-input multiple-output (MIMO) generally uses a star topology for superior communication but faces high costs due to long cables. An economical alternative, the stripe topology, is suitable for specific deployments but cannot meet user demands in densely populated areas due to limited fronthaul capacity. To address these limitations, we propose a hybrid network structure combining stripe and star topologies, ensuring system performance while reducing deployment costs. In such a network, joint precoding and fronthaul compression is considered to maximize system sum-rate and an alternating optimization (AO) algorithm is proposed. However, the AO algorithm involves an iterative process and complex matrix calculations, making it unsuitable for practical applications. To deal with this issue, we propose a low-complexity iterative gradient descent (IGD) algorithm with simple matrix operations. To further reduce online computational complexity, we propose a novel deep unfolding neural network (DUNN) scheme, which is interpretable and scalable, based on the IGD algorithm. Simulation results show that the hybrid topology significantly improves system capacity compared to the stripe-only topology. Additionally, the DUNN achieves a tradeoff between the achievable sum-rate performance and the corresponding computational complexity.
- 11 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (11) , 18125-18136
Joint Precoding and Fronthaul Compression for Cell-Free MIMO With Hybrid Topology
Cell-free multiple-input multiple-output (MIMO) generally uses a star topology for superior communication but faces high costs due to long cables. An economical alternative, the stripe topology, is suitable for specific deployments but cannot meet user demands in densely populated areas due to limited fronthaul capacity. To address these limitations, we propose a hybrid network structure combining stripe and star topologies, ensuring system performance while reducing deployment costs. In such a network, joint precoding and fronthaul compression is considered to maximize system sum-rate and an alternating optimization (AO) algorithm is proposed. However, the AO algorithm involves an iterative process and complex matrix calculations, making it unsuitable for practical applications. To deal with this issue, we propose a low-complexity iterative gradient descent (IGD) algorithm with simple matrix operations. To further reduce online computational complexity, we propose a novel deep unfolding neural network (DUNN) scheme, which is interpretable and scalable, based on the IGD algorithm. Simulation results show that the hybrid topology significantly improves system capacity compared to the stripe-only topology. Additionally, the DUNN achieves a tradeoff between the achievable sum-rate performance and the corresponding computational complexity.
- 10 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network Science and Engineering
- Vol. 12 (3) , 1804-1817
Model-Driven Deep Learning-Based Optimization of Downlink Precoding and Fronthaul Compression in Cell-Free MIMO Systems
Cell-free multiple-input multiple-output (MIMO) system is a promising solution for wireless communications. However, the potential in improving system performance is restricted due to the capacity-limited fronthaul links. To tackle this problem, we utilize the compress-and forward approach, which leverages various compression and encoding techniques to reduce the fronthaul overhead. Thus, jointly optimizing precoding and fronthaul compression presents a significant challenge. However, current optimization algorithms frequently entail high computational complexity because of their iterative processes, rendering impractical for real-time use. To overcome this challenge, we present a model-driven deep learning-based framework to characterize the structure of precoding vectors and the compression matrix via low-dimensional parameters, leveraging the uplink-downlink duality as the expert knowledge. A deep neural network can learn these low-dimensional “key features” from channel state information. Simulation results indicate that our approach achieves performance comparable to the optimal algorithm and substantially reduces complexity.
- 10 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network Science and Engineering
- Vol. 12 (3) , 1804-1817
Model-Driven Deep Learning-Based Optimization of Downlink Precoding and Fronthaul Compression in Cell-Free MIMO Systems
Cell-free multiple-input multiple-output (MIMO) system is a promising solution for wireless communications. However, the potential in improving system performance is restricted due to the capacity-limited fronthaul links. To tackle this problem, we utilize the compress-and forward approach, which leverages various compression and encoding techniques to reduce the fronthaul overhead. Thus, jointly optimizing precoding and fronthaul compression presents a significant challenge. However, current optimization algorithms frequently entail high computational complexity because of their iterative processes, rendering impractical for real-time use. To overcome this challenge, we present a model-driven deep learning-based framework to characterize the structure of precoding vectors and the compression matrix via low-dimensional parameters, leveraging the uplink-downlink duality as the expert knowledge. A deep neural network can learn these low-dimensional “key features” from channel state information. Simulation results indicate that our approach achieves performance comparable to the optimal algorithm and substantially reduces complexity.
- Open Access
- 1 January 2025
- Published by Institution of Engineering and Technology (IET) in Electronics Letters
- Vol. 61 (1)
Downlink Power Control for Cell‐Free Massive MIMO System With Radio Stripes
Radio stripes present a promising fronthaul topology for cell‐free massive multiple‐input multiple‐output systems, enabling seamless connections between the access points (APs) and central processing unit through a single shared cable. In contrast to the traditional star topology, the stripes structure offers a substantial reduction in deployment and maintenance costs, thereby emerging as a more cost‐effective and practical solution. However, the distributed arrangement of APs within the radio stripes potentially poses a threat to the overall system performance, particularly if some APs are situated far away from users. To address this issue, we introduce an AP switching technique aimed at enhancing the spectral efficiency of the system. To further improve the system performance, we have formulated a weighted sum‐rate maximization problem that takes into account the quality of service of users, the transmit power of APs, and the interference among users. To tackle this problem, a power control algorithm that optimizes the transmit power vector of the active APs has been proposed. Simulation results demonstrate that our proposed method not only enhances the system throughput but also improves the energy efficiency.
- 1 January 2025
- Published by Institution of Engineering and Technology (IET) in Electronics Letters
- Vol. 61 (1)
Open AccessDownlink Power Control for Cell‐Free Massive MIMO System With Radio Stripes
Radio stripes present a promising fronthaul topology for cell‐free massive multiple‐input multiple‐output systems, enabling seamless connections between the access points (APs) and central processing unit through a single shared cable. In contrast to the traditional star topology, the stripes structure offers a substantial reduction in deployment and maintenance costs, thereby emerging as a more cost‐effective and practical solution. However, the distributed arrangement of APs within the radio stripes potentially poses a threat to the overall system performance, particularly if some APs are situated far away from users. To address this issue, we introduce an AP switching technique aimed at enhancing the spectral efficiency of the system. To further improve the system performance, we have formulated a weighted sum‐rate maximization problem that takes into account the quality of service of users, the transmit power of APs, and the interference among users. To tackle this problem, a power control algorithm that optimizes the transmit power vector of the active APs has been proposed. Simulation results demonstrate that our proposed method not only enhances the system throughput but also improves the energy efficiency.
- 27 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 13 (5) , 1359-1363
Joint Learning of Channel Estimation and Beamforming for Cell-Free Massive MIMO Systems
In cell-free massive multiple-input multiple-output (MIMO) systems, beamforming is a crucial technique to enhance system performance. Nevertheless, practical challenges arise in obtaining accurate downlink channel state information (CSI), which plays a crucial role in effective beamforming design. Moreover, the limited fronthaul capacity hinders system performance improvement. In this letter, we propose a model-driven deep learning-based framework to obtain the optimal beamforming solution with unknown downlink CSI, comprising the neural network (NN) module and the beamforming recovery (BR) module. Specifically, the NN module jointly trains two neural networks with different functions to obtain downlink CSI and key features extracted from the beamforming matrix. According to the output of the NN module, the BR module recovers the beamforming matrix during the deployment phase. The simulation results demonstrate that the performance of the proposed method is close to that of the numerical algorithm and outperforms benchmark approaches, but the computational complexity is significantly reduced.
- 27 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 13 (5) , 1359-1363
Joint Learning of Channel Estimation and Beamforming for Cell-Free Massive MIMO Systems
In cell-free massive multiple-input multiple-output (MIMO) systems, beamforming is a crucial technique to enhance system performance. Nevertheless, practical challenges arise in obtaining accurate downlink channel state information (CSI), which plays a crucial role in effective beamforming design. Moreover, the limited fronthaul capacity hinders system performance improvement. In this letter, we propose a model-driven deep learning-based framework to obtain the optimal beamforming solution with unknown downlink CSI, comprising the neural network (NN) module and the beamforming recovery (BR) module. Specifically, the NN module jointly trains two neural networks with different functions to obtain downlink CSI and key features extracted from the beamforming matrix. According to the output of the NN module, the BR module recovers the beamforming matrix during the deployment phase. The simulation results demonstrate that the performance of the proposed method is close to that of the numerical algorithm and outperforms benchmark approaches, but the computational complexity is significantly reduced.
- 10 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 891-895
Simulation and test of elastic characteristics of fixed cylinder hydro-pneumatic suspension
As a typical structure of the hydro-pneumatic suspension of medium and heavy tracked vehicles, the fixed cylinder hydro-pneumatic suspension has superior nonlinear characteristics compared with the traditional suspension, which can significantly improve the cushioning performance of the vehicle, so it has been widely concerned by the industry. Most domestic and foreign scholars have studied the mechanical properties of products, but the methods and means are relatively simple, and the transmission ratio of linkage mechanism is mostly simplified linearly, so the calculation accuracy is not high. This paper focuses on the study of the elastic characteristics of fixed cylinder hydro-pneumatic suspension, using ideal gas equation of state theory to describe the internal change law, derived the calculation and analytical formula of the elastic characteristics, and combined with multi-body dynamics software Recurdyn to establish a parametric MBD (Multi body dynamics) model of fixed cylinder hydro-pneumatic suspension. The geometric nonlinear description of the transmission ratio of the linkage mechanism can be realized, and the elastic characteristic curves of the hydro-pneumatic suspension under different parameters can be obtained. The accuracy of the model is verified by bench test, which can effectively guide the design of the product. It lays a foundation for exploring the influence law of different parameters on hydro-pneumatic suspension characteristics and multi-parameter tuning and optimization of vehicle performance.
- 10 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 891-895
Simulation and test of elastic characteristics of fixed cylinder hydro-pneumatic suspension
As a typical structure of the hydro-pneumatic suspension of medium and heavy tracked vehicles, the fixed cylinder hydro-pneumatic suspension has superior nonlinear characteristics compared with the traditional suspension, which can significantly improve the cushioning performance of the vehicle, so it has been widely concerned by the industry. Most domestic and foreign scholars have studied the mechanical properties of products, but the methods and means are relatively simple, and the transmission ratio of linkage mechanism is mostly simplified linearly, so the calculation accuracy is not high. This paper focuses on the study of the elastic characteristics of fixed cylinder hydro-pneumatic suspension, using ideal gas equation of state theory to describe the internal change law, derived the calculation and analytical formula of the elastic characteristics, and combined with multi-body dynamics software Recurdyn to establish a parametric MBD (Multi body dynamics) model of fixed cylinder hydro-pneumatic suspension. The geometric nonlinear description of the transmission ratio of the linkage mechanism can be realized, and the elastic characteristic curves of the hydro-pneumatic suspension under different parameters can be obtained. The accuracy of the model is verified by bench test, which can effectively guide the design of the product. It lays a foundation for exploring the influence law of different parameters on hydro-pneumatic suspension characteristics and multi-parameter tuning and optimization of vehicle performance.