Scholarlink

<p>The proposed paradigm offers a conversational AI solution to the management of Wireless Sensor Networks in the framework of smart agri-ecology, even though there are unresolved challenges in network lifetime and security and energy management. By using conversational AI at the sink level in dynamic sensor node selection, duty cycling, and anomaly resolution, this solution overcomes the requirement for complex mathematical modeling at sensor nodes. It achieves this by the combined solution of a paradigm-shifting sensor reconfiguration strategy against attacks on sensor nodes and a risk aware network energy management strategy coping with both the network security issues and the energy efficiency. This solution ensures robustness to dynamic traffic and adversarial events with minor impact on the energy of the sensors, contrary to the current solutions relying on static heuristics. Simulation results demonstrate the improvement of network security and network sustainability of about 20% and 30% respectively, with the improvement of the accuracy of anomaly detection. The conversational intelligence relaxes the sensor nodes from computation burden by operating at sink or cloud-assisted edge. In this method, adaptive decision-making is enabled while preserving the lightweight operation of WSNs. Organized prompts and network summaries in a few words are used; such systems can be deployed over low-bandwidth connections with very low communication costs, pretty suitable for real-world operations.</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

Conversational intelligence framework for secure and energy-efficient wireless sensor networks

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月7日

查看全文
  • Sauvik Bal,
  • Lopa Mandal

摘要

The proposed paradigm offers a conversational AI solution to the management of Wireless Sensor Networks in the framework of smart agri-ecology, even though there are unresolved challenges in network lifetime and security and energy management. By using conversational AI at the sink level in dynamic sensor node selection, duty cycling, and anomaly resolution, this solution overcomes the requirement for complex mathematical modeling at sensor nodes. It achieves this by the combined solution of a paradigm-shifting sensor reconfiguration strategy against attacks on sensor nodes and a risk aware network energy management strategy coping with both the network security issues and the energy efficiency. This solution ensures robustness to dynamic traffic and adversarial events with minor impact on the energy of the sensors, contrary to the current solutions relying on static heuristics. Simulation results demonstrate the improvement of network security and network sustainability of about 20% and 30% respectively, with the improvement of the accuracy of anomaly detection. The conversational intelligence relaxes the sensor nodes from computation burden by operating at sink or cloud-assisted edge. In this method, adaptive decision-making is enabled while preserving the lightweight operation of WSNs. Organized prompts and network summaries in a few words are used; such systems can be deployed over low-bandwidth connections with very low communication costs, pretty suitable for real-world operations.