增强学习机制下的工业物联网可信信息覆盖协同控制策略研究

Coverage is one of the most important criteria which can measure the quality of service and performance of the Industrial Internet of Things (IIoT), therefore, the coverage control is the crucial issue in the research topics of IIoT. There are several typical shortcomings in the existing related coverage control theories and methods. For example, the widely used disk coverage model is too idealistic to be applied in practical scenarios. Most of the current coverage strategies ignore the heterogeneity and spatial correlation of the monitored environmental parameters and haven’t considered the unknown sensing region and network environment. In order to overcome the existing drawbacks in the related works, considering the requirements of the industrial monitoring application, this project focuses on the coverage control problem in the industrial internet of things based on the novel confident information coverage model and reinforcement learning. The major investigations of this project can be summarized as follows: (1)We study how to energy efficiently deploy a number of nodes in the industrial region of interest by devising penalty and reward mechanisms based on the irregular rule cell learning automata model to satisfy the confident information coverage; (2) We propose some confident information coverage node scheduling protocols based on the dynamic irregular cell learning automata model to prolong the network lifetime; (3) We will evaluate and cross-verify the performance of the proposed models and algorithms by experimental simulations and a practical network prototype. The proposed theories and methods in this project can dramatically save network deployment cost, prolong network lifetime, improve network coverage and quality of service. The research of this project will remarkably extend the current coverage control theory and methodology framework in industrial internet of things, which can effectively guide the IIoT-based practical industrial environment monitoring applications.

覆盖是度量工业物联网服务质量和网络性能的关键指标之一，覆盖控制问题是工业物联网研究中的关键问题。现有覆盖控制理论和方法中存在节点覆盖模型过于理想化、忽略环境变量差异性及空间相关特性、未充分考虑环境信息未知性等不足。为了克服上述不足，本项目紧密契合工业区域环境安全监测应用重大需求，基于可信信息覆盖模型和增强学习机制，深入研究工业物联网区域覆盖控制问题。研究内容包括：(1)基于不规则元胞学习自动机模型，构建奖惩机制，研究可信信息覆盖节点部署策略，提升网络覆盖率；(2)基于动态不规则元胞学习自动机模型，研究可信信息覆盖节点调度方法，延长网络寿命；(3)研究并设计覆盖控制理论方法交叉验证真实平台。本项目的研究成果可降低网络部署成本，延长网络寿命，提升网络覆盖性能和网络服务质量。本项目的研究对拓展现有工业物联网覆盖控制研究理论框架和方法体系具有重要理论价值和实践意义。

在工业物联网研究中,网络覆盖是度量工业物联网服务质量和网络性能的重要依据之一,网络覆盖控制问题是工业物联网研究中的关键问题。现有的覆盖控制理论及方法中存在诸如节点覆盖模型偏理想,未考虑空间相关性问题,未对环境信息未知性进行有效处理等问题,缺乏对环境信息的全面掌握。为克服上述缺陷,针对工业区域环境安全监测方面的关键需求,本项目利用可信信息覆盖模型和增强学习机制,深入研究了工业物联网区域网络覆盖控制。研究内容包括:(1)研究了可信信息覆盖节点部署策略，通过不规则元胞学习自动机模型，构建了奖惩机制;(2)基于动态不规则元胞学习自动机模型，提出了可信信息覆盖节点调度的方法，延长了网络寿命;(3)研究并设计了覆盖控制理论方法交叉验证真实平台。本项目的研究成果降低了网络部署成本，延长了网络寿命，提升了网络覆盖性能和网络服务质量。本项目的研究拓展了现有工业物联网覆盖控制研究的理论框架和方法体系，为工业物联网的发展提供了重要理论价值和实践指导意义。

内容获取失败，请点击重试