相关参考文献
注意：仅列出部分参考文献，下载原文获取全部文献信息。
Article
Engineering, Electrical & Electronic
Chen Dong et al.
Summary: Traditional communication systems, which neglect the semantic aspects and meaning of messages, have limitations in terms of communication efficiency and intent-sharing. In this paper, a novel semantic communication system is proposed, which incorporates essential semantic information exploration, transmission, and recovery for more efficient communication. The system is characterized by the flow of intelligence through the propagation of the model and introduces the concept of semantic slice-models (SeSM) to enable flexible model-resembling based on different requirements. A layer-based semantic communication system for images (LSCI) is built to demonstrate feasibility, and a semantic metric called semantic service quality (SS) is proposed for evaluation. The experiments show significant bitrate reduction and energy savings compared to existing methods.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Wenyu Zhang et al.
Summary: This paper proposes a predictive and adaptive deep coding (PADC) framework that achieves flexible code rate optimization with a given target transmission quality requirement. By using a variable code length enabled DeepJSCC model, an Oracle Network model, and a CR optimizer, PADC can minimize bandwidth consumption while guaranteeing the PSNR constraint for each image data in wireless image transmission tasks.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Computer Science, Hardware & Architecture
Xuewen Luo et al.
Summary: This article provides an overview of the latest deep learning and end-to-end communication-based semantic communications, and discusses the open issues that need to be addressed.
IEEE WIRELESS COMMUNICATIONS
(2022)
Article
Engineering, Multidisciplinary
Ping Zhang et al.
Summary: The sixth generation (6G) mobile networks will reshape the world by offering instant, efficient, and intelligent hyper-connectivity. To achieve this vision, a new route towards wisdom-evolutionary and primitive-concise network (WePCN) is proposed, which involves semantic representation frameworks and intelligent semantic communication architectures.
Article
Engineering, Electrical & Electronic
Jialong Xu et al.
Summary: By utilizing attention mechanisms, we propose a novel DL based JSCC method that can efficiently operate under different SNR levels. The proposed method requires less storage and is more robust in the presence of channel mismatch compared to existing methods.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2022)
Article
Engineering, Electrical & Electronic
Jincheng Dai et al.
Summary: This paper proposes a high-efficiency deep joint source-channel coding method that can adapt to the source distribution under nonlinear transform and potentially support future semantic communications.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2022)
Article
Engineering, Electrical & Electronic
Yining Wang et al.
Summary: This paper proposes a semantic communication framework for textual data transmission. The framework extracts semantic information from textual data at a base station (BS) and transmits it to each user. A knowledge graph (KG) is used to model the semantic information, and a graph-to-text generation model is employed by each user to recover the original text. To evaluate the performance of the framework, a metric of semantic similarity (MSS) is introduced. An optimization problem is formulated to maximize the total MSS by jointly optimizing resource allocation and determining the partial semantic information to be transmitted. The proposed proximal-policy-optimization-based reinforcement learning (RL) algorithm integrates an attention network to evaluate the importance of each triple in the semantic information. Simulation results demonstrate the effectiveness of the proposed framework in reducing data transmission requirements and improving the total MSS.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2022)
Proceedings Paper
Computer Science, Information Systems
Lei Yan et al.
Summary: This paper investigates the problem of semantic-aware resource allocation in a multi-cell multi-task network. A novel measure of semantic entropy is proposed to quantify the semantic information for different tasks. A quality-of-experience (QoE) model is also proposed, and the resource allocation problem is formulated in terms of the number of transmitted semantic symbols, channel assignment, and power allocation. Simulation results demonstrate the effectiveness and superiority of the proposed method on the overall QoE.
2022 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM 2022)
(2022)
Proceedings Paper
Acoustics
Mingyu Yang et al.
Summary: We propose a novel adaptive deep joint source-channel coding (JSCC) scheme for wireless image transmission. The scheme supports multiple rates using a single deep neural network (DNN) model and dynamically controls the rate based on channel condition and image contents. The scheme introduces a policy network to optimize the tradeoff between rate and signal quality, and achieves similar performance to an optimized model specifically trained for a fixed target rate.
2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP)
(2022)
Proceedings Paper
Computer Science, Information Systems
Danlan Huang et al.
Summary: This paper introduces an image semantic coding approach based on GANs, achieving visually pleasing reconstruction and semantic preservation at extremely low bitrates through rate-perception-distortion optimization. The proposed coarse-to-fine image semantic coding model in multimedia semantic communication system outperforms existing methods.
2021 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM)
(2021)
Proceedings Paper
Telecommunications
Zhenzi Weng et al.
Summary: The researchers proposed a semantic communication system named DeepSC-S, leveraging breakthroughs in deep learning, aimed at minimizing errors at the semantic level when recovering transmitted speech signals. They designed an end-to-end system with an attention mechanism utilizing squeeze-and-excitation networks to learn and extract essential speech information. Simulation results showed that the system is more robust to channel variations and outperforms traditional communication systems in low signal-to-noise environments.
IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS (ICC 2021)
(2021)
Proceedings Paper
Acoustics
Mingze Ding et al.
Summary: This paper proposes a novel deep joint source-channel coding scheme based on autoencoders, where the decoder can adaptively decode transmitted images based on estimated SNR and exhibits impressive adaptability and robustness to noise in SNR estimation. This is the first deep JSCC scheme focusing on adaptability for different SNRs and applicable to multi-user scenarios.
2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021)
(2021)
Article
Engineering, Electrical & Electronic
Huiqiang Xie et al.
Summary: The research presents a deep learning based semantic communication system, DeepSC, for text transmission. By maximizing system capacity and minimizing semantic errors, DeepSC recovers the meaning of sentences rather than bit or symbol errors in traditional communications. Transfer learning is used to ensure adaptability to different communication environments and accelerate the model training process.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Huiqiang Xie et al.
2020 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM)
(2020)
Article
Computer Science, Information Systems
Silvija Kokalj-Filipovic et al.
BELL LABS TECHNICAL JOURNAL
(2012)
Article
Computer Science, Artificial Intelligence
Z Wang et al.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2004)
