.png){kind=icon}
返回
Toward Intelligent Resource Allocation on Task-Oriented Semantic Communication
.png){kind=icon}
PRE
AI
DOI：10.1109/MWC.008.2200504.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Task-oriented semantic communication (TOSC) has significant advantages in reducing the amount of data transmission and alleviating the scarcity of spectrum resources. Unlike traditional communication, the resource allocation in semantic communication is tightly linked to target intelligent tasks and specific interaction requirements. In this article, the intelligent resource allocation in a task-oriented manner is investigated. To further improve spectrum utilization and energy sustainability, a communication network combining energy harvesting (EH), cognitive radio (CR), and non-orthogonal multiple access (NOMA) is considered. This article proposes a semantic-aware resource allocation scheme in the EH-CR-NOMA scenario, where the quality of experience (QoE) is adopted as the evaluation metric. To achieve the preferential occupation of resources by data with richer semantic information, a joint optimization problem of the transmit power, time slot division factor, and semantic compression ratio of the semantic communication user is formulated. With the goal of maximizing the long-term QoE of TOSC, a two-tier deep reinforcement learning framework is designed to solve the semantic-aware resource allocation problem. By striking a trade-off between semantic rate and semantic fidelity, the proposed scheme can better satisfy user intentions.
Keyword:
Measurement
NOMA
Semantics
Reinforcement learning
Resource management
Quality of experience
Energy harvesting
期刊
IF:
11.5
论文数:
2.8K
被引数:
1.3W
机构
引用论文
分享收藏Resource Allocation for Optimizing Energy Efficiency in NOMA-based Fog UAV Wireless Networks基于NOMA的雾无人机无线网络能效优化资源分配.png)
IEEE NETWORK
IEEE NETWORKIF6.3
PREAI
分享收藏分享收藏Semantic Communications With Artificial Intelligence Tasks: Reducing Bandwidth Requirements and Improving Artificial Intelligence Task Performance与人工智能任务的语义通信: 降低带宽需求并提高人工智能任务性能
PREAI
分享收藏No-Pain No-Gain: DRL Assisted Optimization in Energy-Constrained CR-NOMA Networks无痛苦无增益: 能量受限cr-noma网络中的DRL辅助优化
OAAI
OAAI分享收藏分享收藏
查看更多内容