Owing to provision of a long-range and robust link, LoRa technology has drawn attention for the use in industrial data collection networks. This article proposes a real-time LoRa protocol that can effectively deal with both periodic and aperiodic data. The slots in a frame are logically partitioned such that they are first scheduled for periodic data by a slot scheduling algorithm, and then, the remaining unscheduled slots are used for event-driven aperiodic data. In this logical frame partitioning, the unscheduled slots appear in an interleaved fashion so that aperiodic tasks can transmit data with low delay and fairness while every periodic task still completes data transmission before the beginning of the next period. To deal with the problems of data collision and traffic congestion for aperiodic data, a two-level collision avoidance scheme is proposed that adopts the notion of a contention window and a delay slot. According to simulation, not only can the proposed protocol guarantee the timely delivery of periodic data but it can also deal with aperiodic data with high reliability, fairness, and low delay, compared with other recent protocols.

This study explores a method of detecting smoke plumes effectively as the early sign of a forest fire. Convolutional neural networks (CNNs) have been widely used for forest fire detection; however, they have not been customized or optimized for smoke characteristics. This paper proposes a CNN-based forest smoke detection model featuring novel backbone architecture that can increase detection accuracy and reduce computational load. Since the proposed backbone detects the plume of smoke through different views using kernels of varying sizes, it can better detect smoke plumes of different sizes. By decomposing the traditional square kernel convolution into a depth-wise convolution of the coordinate kernel, it can not only better extract the features of the smoke plume spreading along the vertical dimension but also reduce the computational load. An attention mechanism was applied to allow the model to focus on important information while suppressing less relevant information. The experimental results show that our model outperforms other popular ones by achieving detection accuracy of up to 52.9 average precision (AP) and significantly reduces the number of parameters and giga floating-point operations (GFLOPs) compared to the popular models.