Scholarlink

<p>Confronted by the global challenge of ageing and the paucity of elderly care resources in rural areas, a novel smart elderly care ecosystem model integrating digital twins and reinforcement learning is proposed. A digital twin platform for elderly care in rural areas has been constructed, combining multi-source IoT and edge computing nodes. This has enabled the resolution of low-bandwidth data fusion issues. The core innovation of this study lies in the deep integration of backscatter communication and multi-agent deep reinforcement learning (MADDPG) into a cohesive system. This integration empowers the dynamic resource scheduling engine to perform real-time service optimisation and emergency response while simultaneously adapting to hardware energy constraints. Specifically, the backscatter communication module provides energy-efficient data links, the status of which is fed as real-time state inputs to the reinforcement learning agents. In return, the MADDPG algorithm output dynamically guides the adjustment of communication parameters. In the pilot areas of Shandong, a significant reduction in the response time to medical emergencies was observed, from 34.7&#xa0;s to 8.9&#xa0;s. Furthermore, the service continuity in disaster situations was recorded at 89.2%, indicating a notable enhancement in the effectiveness of emergency response systems. The device’s backscatter communication technology has been demonstrated to ensure a battery life of 68&#xa0;h when subjected to ± 15% voltage fluctuations, with a 3.2% loss of communication packets. The community resilience index demonstrated a marked increase from 0.72 to 0.94 over a 12-month period, while user satisfaction levels attained 9.6. The cost-benefit analysis of the platform indicates a 2.8-year investment payback period, attributable to reduced medical expenses and enhanced service efficiency. The present study validates the technical solution’s efficacy in addressing the digital divide in rural elderly care, thereby establishing a replicable “technology - policy - community” model for global ageing governance.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Backscatter communication and multi-agent reinforcement learning enable low-power digital twin system for rural elderly care

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月29日

查看全文
  • Xiaoyan Guo

摘要

Confronted by the global challenge of ageing and the paucity of elderly care resources in rural areas, a novel smart elderly care ecosystem model integrating digital twins and reinforcement learning is proposed. A digital twin platform for elderly care in rural areas has been constructed, combining multi-source IoT and edge computing nodes. This has enabled the resolution of low-bandwidth data fusion issues. The core innovation of this study lies in the deep integration of backscatter communication and multi-agent deep reinforcement learning (MADDPG) into a cohesive system. This integration empowers the dynamic resource scheduling engine to perform real-time service optimisation and emergency response while simultaneously adapting to hardware energy constraints. Specifically, the backscatter communication module provides energy-efficient data links, the status of which is fed as real-time state inputs to the reinforcement learning agents. In return, the MADDPG algorithm output dynamically guides the adjustment of communication parameters. In the pilot areas of Shandong, a significant reduction in the response time to medical emergencies was observed, from 34.7 s to 8.9 s. Furthermore, the service continuity in disaster situations was recorded at 89.2%, indicating a notable enhancement in the effectiveness of emergency response systems. The device’s backscatter communication technology has been demonstrated to ensure a battery life of 68 h when subjected to ± 15% voltage fluctuations, with a 3.2% loss of communication packets. The community resilience index demonstrated a marked increase from 0.72 to 0.94 over a 12-month period, while user satisfaction levels attained 9.6. The cost-benefit analysis of the platform indicates a 2.8-year investment payback period, attributable to reduced medical expenses and enhanced service efficiency. The present study validates the technical solution’s efficacy in addressing the digital divide in rural elderly care, thereby establishing a replicable “technology - policy - community” model for global ageing governance.