- Advanced MIMO Systems Optimization
- Cooperative Communication and Network Coding
- Advanced Wireless Network Optimization
- Advanced Wireless Communication Techniques
- Wireless Communication Networks Research
- Advanced Wireless Communication Technologies
- Full-Duplex Wireless Communications
- Energy Harvesting in Wireless Networks
- Millimeter-Wave Propagation and Modeling
- IoT and Edge/Fog Computing
- Wireless Communication Security Techniques
- Cognitive Radio Networks and Spectrum Sensing
- Error Correcting Code Techniques
- Satellite Communication Systems
- Wireless Signal Modulation Classification
- Cloud Computing and Resource Management
- IoT Networks and Protocols
- Advanced Computational Techniques and Applications
- Wireless Networks and Protocols
- Photonic and Optical Devices
- Caching and Content Delivery
- PAPR reduction in OFDM
- Quantum Information and Cryptography
- Recommender Systems and Techniques
- Antenna Design and Analysis
University of Science and Technology of China
2015-2024
Tsinghua University
2011-2024
Suzhou University of Science and Technology
2022-2024
Tencent (China)
2020-2024
Arizona State University
2018-2024
China Southern Power Grid (China)
2022-2023
Huaiyin Institute of Technology
2023
Chinese Academy of Sciences
1999-2022
Anhui Institute of Optics and Fine Mechanics
2021-2022
Institute of Technology of Cambodia
2022
The distributed wireless communication system (DWCS) is a new architecture for access with antennas, processors, and control. With the capacity can be expanded through dense frequency reuse, transmission power greatly decreased. processors control, works like software or network radio, so different standards coexist, increased by coprocessing of signals to from multiple antennas.
Multi-task learning (MTL) has been successfully applied to many recommendation applications. However, MTL models often suffer from performance degeneration with negative transfer due the complex and competing task correlation in real-world recommender systems. Moreover, through extensive experiments across SOTA models, we have observed an interesting seesaw phenomenon that of one is improved by hurting some other tasks. To address these issues, propose a Progressive Layered Extraction (PLE)...
In this paper, we investigate the optimal power allocation strategy that aims at maximizing capacity in OFDM- based cognitive radio systems. We show traditional water-filling algorithm applied general OFDM systems needs to be modified due per subchannel constraints such An iterative partitioned is proposed and proved on convex optimization theory.
Optimal power allocation for orthogonal frequency division multiplexing (OFDM) wiretap channels with Gaussian channel inputs has already been studied in some previous works from an information theoretical viewpoint. However, these results are not sufficient practical system design. One reason is that discrete inputs, such as quadrature amplitude modulation (QAM) signals, instead of deployed current wireless systems to maintain moderate peak transmission and receiver complexity. In this...
In this paper, reinforcement learning (RL) based power control algorithm in underlay D2D communication is studied. The approach we use regards as a multi-agents system, and achieved by maximizing system capacity while maintaining the requirement of quality service(QoS) from cellular users. We propose two RL methods for users, i.e., team-Q distributed-Q learning. former centralized method which only one Q-value table needs to be maintained, latter enables users learn independently reduces...
In this paper, we design the uplink transceiver for a new multiuser (MU) massive spatial modulation (SM) multiple-input-multiple-output (SM-MIMO) system over frequency-selective fading channels, where base station (BS) is equipped with antennas, and user equipment (UE) has multiple transmit antennas (TAs) but only one radio-frequency (RF) chain. UE data bits to BS simultaneously by cyclic-prefix single-carrier (CP-SC) SM. For MU detection (MUD), construct low-complexity generalized...
The rapid growth of Internet-of-things (IoT) and artificial intelligence applications have called forth a new computing paradigm--edge computing. In this paper, we study the suitability deploying FPGAs for edge from perspectives throughput sensitivity to workload size, architectural adaptiveness algorithm characteristics, energy efficiency. This goal is accomplished by conducting comparison experiments on an Intel Arria 10 GX1150 FPGA Nvidia Tesla K40m GPU. experiment results imply that key...
3D room layout estimation by a single panorama using deep neural networks has made great progress. However, previous approaches can not obtain efficient geometry awareness of with the only latitude boundaries or horizon-depth. We present that horizondepth along height omnidirectionalgeometry in both horizontal and vertical directions. In addition, we propose planar-geometry aware loss function normals gradients to supervise planeness walls turning corners. an network, LGT-Net, for...
In this article, a non-through metal-insulator-metal (MIM) waveguide that can excite fivefold Fano resonances is reported. The are obtained by the interaction between modes excited square split-ring resonator (SSRC) and bus waveguide. After detailed analysis of transmission characteristics magnetic field strength structure using finite element method (FEM), it was found independent tuning resonance wavelength transmittance be achieved adjusting geometric parameters SSRC. addition, after...
A new joint frame synchronization and carrier frequency offset estimation scheme in orthogonal division multiplexing (OFDM) systems is proposed this paper, where both can be performed by using only ONE training symbol. Frame acquisition are simultaneously the scheme. Reliable obtained even low SNR. The maximum range of up to one half total signal bandwidth. same symbol also utilized Fine Adjustment, which estimates remaining after with higher accuracy. performance comparison Adjustment...
The maximum-likelihood detection (MLD) algorithm is the optimal scheme for vertical Bell Laboratories layered space-time (V-BLAST) systems. However, it generally infeasible to implement MLD in practical systems because of its high computational complexity. In this paper, we used two parallel interference cancellation techniques, including precancellation and postcancellation, simplify into a class computationally efficient suboptimal algorithms, herein called generalized (GPIC) algorithm....
Digital filters are widely used in signal processing and communication systems. In some cases, the reliability of those systems is critical, fault tolerant filter implementations needed. Over years, many techniques that exploit filters' structure properties to achieve tolerance have been proposed. As technology scales, it enables more complex incorporate filters. systems, common operate parallel, for example, by applying same different input signals. Recently, a simple technique exploits...
High-throughput and low-latency Convolutional Neural Network (CNN) inference is increasingly important for many cloud- edge-computing applications. FPGA-based acceleration of CNN has demonstrated various benefits compared to other high-performance devices such as GPGPUs. Current FPGA CNN-acceleration solutions are based on a single design, which limited by the available resources an FPGA. In addition, they can only accelerate conventional 2D neural networks. To address these limitations, we...
Heterogeneous network (HetNet) is considered as a prime way to solve the limits of system capacity and broadband service coverage in traditional network. However, deployments small cells with varied sizes make topology more complicated. Self-organizing (SON) technology, aiming reduce operational costs, significant technology HetNet. One common use cases improve handover performance. In this paper, optimization algorithm based on enhanced mobility state estimation (EMSE) proposed. Considering...
Facing the development of future 5G, emerging technologies such as Internet things, big data, cloud computing, and artificial intelligence is enhancing an explosive growth in data traffic. Radical changes communication theory implement technologies, wireless communications networks have entered a new era. Among them, (WBD) has tremendous value, (AI) gives unthinkable possibilities. However, application groups, lack sound theoretical foundation mathematical methods regarded real challenge...
It is an effective scheme to the phase retrieval for axial intensity derivative computing. In this paper, we demonstrate a method estimating and improving calculation accuracy in transport of equation (TIE) from multiple measurements. The takes both higher-order derivatives noise into account, minimizes impact detecting noise. simulation results that proposed can effectively reduce error
In this paper we consider data transmission in a decode-and-forward (DF) relay-assisted network which the relay is energy harvesting (EH) powered while base station (BS) power-grid powered. Our purpose to maximize BS's efficiency (EE) making full use of relay's renewable and satisfying specific average throughput requirements. contrast existing literature on system only considers radio power, take static circuit power into account as well. We formulate EE optimization problem prove that BS...
Sparse coded multiple access (SCMA) has received considerable attention recently as a candidate scheme for 5G cellular systems. This letter proposes an improved uplink SCMA where instead of the multi-dimensional constellation (MDC) mapping and fixed number nonzero elements per sparse codeword, varying symbols from each user are directly mapped onto resource elements. We show that proposed outperforms MDC-SCMA in terms both error rate performance receiver complexity when load is high.
Non‐orthogonal multiple access (NOMA) has recently attracted a lot of attention as promising technology to be used in the fifth generation (5G) cellular networks. In this study, authors investigate uplink NOMA transmissions cooperative network presence frequency selective multipath fading channels. Specifically, two users communicate with base station simultaneously existence half‐duplex relay employing decode‐and‐forward scheme assist far user. At transmitting nodes, power allocation...
Introducing relay node to Device-to-Device (D2D) technology can extend the cell coverage range and improve system performance, selection is a key point of relay-assisted D2D communication. In this paper, problem discussed for communication underlaying cellular network. We propose two-stages resource allocation joint method The maximizes total throughput uplink (UL) link, it guarantee quality service (QoS) two links simultaneously. Through method, optimal UE its transmission power be...
Identifying a short segment in long video that semantically matches text query is challenging task has important application potentials language-based search, browsing, and navigation. Typical retrieval systems respond to with either whole or pre-defined segment, but it localize undefined segments untrimmed unsegmented videos where exhaustively searching over all possible intractable. The outstanding challenge the representation of must account for different levels granularity temporal...
In this paper, we propose a nonorthogonal downlink multiple access scheme, where the base station assigns bit streams to different users from an adjustable hierarchical quadrature-amplitude-modulation (QAM) constellation and all deploy simple non-iterative receiver with same complexity as in orthogonal (OMA) system. We analyze achievable information rates of proposed termed QAMA, show that it possesses similar rate region hybrid OMA MUST approach 5G standard, while latter relies on...
Soft errors pose a reliability threat to modern electronic circuits. This makes protection against soft requirement for many applications. Communications and signal processing systems are no exceptions this trend. For some applications, an interesting option is use algorithmic-based fault tolerance (ABFT) techniques that try exploit the algorithmic properties detect correct errors. Signal communication applications well suited ABFT. One example fast Fourier transforms (FFTs) key building...