An Enhanced Reliable Access Scheme for Massive IoT Applications in Ubiquitous IoT Systems

Abstract

As a key infrastructure for intelligent manufacturing, the Industrial Internet of Things (IIoT) can connect people, massive intelligent terminals, automated control equipment, and convergence nodes to achieve a high integration of industrial systems with advanced computing, sensor technology, fifth-generation technology (5G), and the Internet. How to realize the robust and reliable connection of massive IoT applications has become the research focus of industry and academia. However, the massive IoT industrial equipment, diverse business types, complex networking mode, and broadcast characteristics of wireless media have posed serious challenges to providing stable and reliable connections in industrial communication scenarios. In order to overcome this problem, the paper proposes an enhanced reliable access scheme (ERAS) for massive IoT applications in software-defined industrial heterogeneous networks. Based on the idea of software-defined networking (SDN), the ERAS is designed from two perspectives of the control plane and the data plane. Based on the idea of greedy algorithms, ERAS uses priority, interference set, deadline, and other information to find the local optimal solution and then obtain the global optimum. First, low-priority periodic data is accessed to the channel based on the earliest deadline. Then, high-priority data (such as control and interactive services) judges the channel occupancy status according to the access flow table, and if there is an idle channel in the current time slot, it is normally accessed. Otherwise, the subslots reserved with high priority in the control channel are used to preempt the low-priority channel. This preemption mechanism can ensure reliable and real-time connections for massive IIoT. Simulation results show that the ERAS performs better than traditional methods in terms of delay, deadline guarantee rate, and packet loss rate.

Evaluation of a High-Accuracy Indoor-Positioning System with Wi-Fi Time of Flight (ToF) and Deep Learning

“…Additionally, minibatch SGD improves computing efciency. We use the estimate of moments of gradients to hasten convergence and slow down the quick decay of learning rates [42,43]. First, calculate g t,i that is the gradient of the cost function w.r.t.…”