IoT provides ubiquitous connectivity and pervasive intelligence. Challenges arise from the data-intensive applications pertaining to IoT. The major contribution of this article is a new asymptotically optimal, fully decentralized, real-time framework which seamlessly integrates wireless computation offloading and fog computing in IoT networks with random traffic variations over space and time. Other enabling techniques are discussed to address practical challenges, such as outdated incentive for the cooperation of multiple service providers, massive access requests within limited system bandwidth, and computing acceleration. The new framework is able to provide ubiquitous computing for continuously increasing IoT services. Numerical experiments indicate that ubiquitous fog computing can substantially improve throughput, reduce computing latency, and cut off signaling overhead, without compromising the optimality of network operations.

中文翻译：

---

随时随地的协作计算：无所不在的雾服务

物联网提供了无处不在的连接性和无处不在的智能。与物联网有关的数据密集型应用带来了挑战。本文的主要贡献是一个新的渐近最优，完全去中心化的实时框架，该框架无缝集成了IoT网络中的无线计算分载和雾计算，并随时间和空间变化了流量。讨论了解决其他实际问题的其他使能技术，例如对多个服务提供商合作的过时激励，有限系统带宽内的大量访问请求以及计算加速。新框架能够为不断增长的物联网服务提供无处不在的计算。数值实验表明，普遍存在的雾计算可以显着提高吞吐量，减少计算延迟，