Self-Aware Fog Computing in Private and Secure Spheres
Self-Aware Fog Computing in Private and Secure Spheres
复制标题
私有和安全领域的自我感知雾计算
DOI：
10.1007/978-3-319-57639-8_5
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
E. Ounapuu
中科院分区：
文献类型：
--
作者：
Kalle Tammemäe;A. Jantsch;A. Kuusik;J. Preden;E. Ounapuu
在实时健康分析、智能城市、军事传感系统和其他方面，大数据分析是通过引入适当的传感和驱动系统来实现的。下一代传感和致动系统或物联网时代的引入得到了以下因素的推动：经济实惠的低功耗32位微控制器结合了适当电源的低成本高效传感器、移动和本地数据收集(本地大数据)能力、使用机器学习的自适应行为和进化基于模型的行为等。云计算在服务器层面提供大数据处理和致动能力，而Mist计算则在网络的最边缘提供数据处理和致动能力。雾计算在边缘网关的中间提供了相同的能力。雾计算是许多应用的推动者，这些应用是用其他方法无法实现的，例如智能城市，城市街道适应城市发生的变化，社会智能房屋，室内环境管理与居民健康监测相集成，或者军事传感系统，从原始数据自动推断情况信息，并传递给信息消费者。虽然这些有远见的应用程序承诺改变我们的环境以及我们与环境交互的方式，但我们在实施这些系统时面临着严峻的挑战，如节点和路由器之间数据交换的可靠性、功率分配、决策质量等。
In real-time health analytics, smart cities, military sensing systems, and others, big data analytics is enabled by the introduction of appropriate sensing and actuation systems. The introduction of next generation of sensing and actuation systems or the Internet of Things era has been facilitated by affordable low-power 32-bit microcontrollers combined with low-cost and effective sensors with appropriate power supplies, mobile and local data collection (local big data) capabilities, adaptive behavior using machine learning and evolving model-based behavior, etc. While Cloud computing offers big data processing and actuation capability at the server level, mist computing offers data processing and actuation capability at the very edge of the network. Fog computing offers the same capability in the middle at edge gateways. Mist computing is an enabler for many applications, which cannot be realized with alternative methods, such as smart cities, where city streets adapt to the changes happening in the city, socially intelligent houses where indoor environment management is integrated with inhabitants health monitoring, or military sensing systems where situational information is automatically deduced from raw data and delivered to the information consumers. While these visionary applications promise to change our environment and the way we interact with the environment, we face serious challenges in implementing these systems, such as reliability of data exchange between nodes and routers, power distribution, quality of decision-making, etc.