Artificial Intelligence of Things (AIoT) represents a technological revolution that integrates the processing of Artificial Intelligence (AI) with the Internet of Things (IoT). This work proposes an AIoT-based Mobile Edge Computing Architecture that enables the decentralization of the on-demand decision-making process in remote environments through hierarchical network management, emphasizing the edge and fog layers. The architecture aims to apply AI in an airborne IoT network infrastructure to increase the efficiency of applications composed of low-power computing devices in terms of processing, communication, and energy. The proposed architecture aims to direct the adequate use of AI models in a decentralized manner on-demand, seeking to improve applications in terms of latency, bandwidth consumption, and energy. The study showed that the NCNN format offers the best energy efficiency and processing speed at the edge layer, processing up to 219 frames in real-time with only 3.57 mAh and 80 ms per image. While fog computing reduced memory usage, it increased latency over 40 meters, highlighting the need for hybrid strategies.

Leveraged by IoT and Industry 4.0 solutions, Wireless Sensor Networks (WSNs) have been proposed as an important alternative for large-scale monitoring applications. Such technology provides sensor nodes with the intelligent and autonomous ability to monitor large areas, create self-organizing structures, detect events and process massive data. In this context, data-driven schemes are increasingly needed. For this, some data clustering techniques (DCTs) are used to tackle common problems in WSNs; however, the vast majority of techniques do not consider the data monitored by the sensors to perform topological changes and provide better network structures. This work addresses an architecture for this type of application and evaluates the impact of different DCTs on network performance and the creation of priority node groups.

Wireless Sensor Networks (WSNs) based on the Long-Range radio modulation (LoRa) can use the LoRaWAN protocol as the medium access layer. However, this protocol only supports a single-hop star topology. As a consequence, devices can not use retransmissions along the network to extend their coverage area or to circumvent signal attenuation with distance, obstacles or interference from other radio sources. This paper proposes a multi-hop LoRaWAN wide-scale WSN based on a scheduled cluster-tree topology. This methodology can expand the spatial coverage of the network, decrease collisions, and improve overall network performance. A multi-hop cluster-tree topology eliminates the need for adjustments of LoRa radio parameters as an attempt to expand the single-hop coverage limitation. Simulation results show that the scheduled cluster-tree topology can scale the network coverage and significantly improve communication and energy consumption performances.

Wireless Sensor Networks (WSNs) have become a fundamental technology for implementing Internet of Things (IoT) solutions and supporting Industry 4.0 applications. This technology presents as main characteristics: high scalability, synchronized communication, and low power consumption. The network topological formation is a key factor in the design of WSNs, as the structure resulting from this process dictates the operation mode in which the network should handle such issues. For this, standards such as IEEE 802.15.4 and ZigBee have composed an adequate protocol stack, providing support for complex network topologies, such as cluster-tree. The objective of this work is to implement cluster-tree topology formation mechanisms driven by the data generated of sensor nodes, called the Bottom-Up Formation (BUF). The underlying idea is to select cluster-heads and build clusters following a bottom-up approach, aiming to define homogeneous tree branches in order to prioritize different data traffic. The results demonstrate that, compared to a state-of-the-art approach, the proposed mechanisms can reduce communication delays by up to 63%, achieve communication success rates of up to 99%, and extend the nodes’ lifetime.

The elastic optical networks have shown promise for future of optical communications supporting high transmission rates data. To meet high transmission rates, you need to solution RMLSA (Routing, Modulation Level, Spectrum Assignment) problem which is to select a route, choose a modulation level and a range of free spectrum. A new algorithm called Circuit Interference Reduction - Minimum Allocation Slots (CIR-MAS) is presented in this work. This algorithm seeks to select a modulation format that is more resistant circuit interferences (established circuits on the networks). Also it selects a route with the minimum allocation slots among the alternative routes, in order to reduce the blockage caused by degradation of circuits. The performance of the CIR-MAS algorithm was evalueted on the NSFNet and EON topologies by comparing with the following algorithms: K-Shortest Path Computation (KS-PC), Modified Dijkstra Path Computation (MD-PC) and K-Shortest Path with Reduction of QoTO (KSP-RQoTO), wich have already been presented in the literature. The CIR-MAS algorithm presented a minimum gain of 40.5% in terms of circuit blocking probability, and in terms of bandwidth blocking probability, it obtained a minimum gain of 37.5% when compared all competing algorithms.

The IEEE 802.15.4 and ZigBee applications are widely used specifications for the design of Wireless Sensor Networks (WSNs), Among the different available network topologies, cluster-tree networks are pointed out as the most suitable to implement large-scale WSNs, becoming an attractive solution for monitoring applications. However, cluster-tree networks may suffer from high end-to-end communication delay and also from network congestion. Thus, the selection of efficient network formation schemes and the definition of adequate communication structures are relevant research issues. In this context, this paper proposes an efficient mechanism to improve convergecast traffic in large-scale cluster-tree wireless sensor networks. The main idea combines a network formation process driven by the data generation characteristics of individual nodes and an efficient allocation scheme of the active periods, in order to improve typical convergecast message streams generated by the sensor nodes. Simulation results show that efficient network configuration approaches can significantly reduce end-to-end communication delays and decrease the overall network congestion.

Indoor Positioning Systems (IPS) constitutes a relevant research issue mainly due to the difficulties of employing Global Positioning System (GPS) technology in closed environments. Localization algorithms based on scene analysis technique, which map the environment signal levels at design time, present good results, being the main IEEE 802.11-based solution. In environments where this type of infrastructure is not available or the requirements for low-cost and power consumption limit the employment of such networks, other technologies should be used. Within this context, this work proposes a self-positioning system implemented in an IEEE 802.15.4 network. Considering a real scenario application, the proposed system presented relevant results, obtaining an accuracy of 0.6 m for the position estimation.

Cluster-tree topologies are commonly employed in large-scale Wireless Sensor Networks applications where nodes are randomly deployed. This topology presents several challenges related to connectivity and message schedulability. As a consequence, effective methods for network formation and operation are required. This work proposes a combination of methods of neighborhood discovery, network formation and message scheduling to deal with this problem. This proposal aims to build a flexible network, with better performance and is also able to adapt for different scenarios. Simulation results highlight significant improvements of the network performance, when compared to the use of LEACH protocol.

Topology formation in wireless sensor networks is usually done assuming just either the geographical proximity between nodes or the signal strength of communication. In this paper, a heuristic called DbCTF is proposed to guide the formation of cluster-tree networks, which also considers data clustering techniques. The use of DbCTF allows the setup of a data-based topology in the cluster-tree, and also the prioritisation of monitored regions in which relevant events may be occurring. The performance of DbCTF has been compared with a state-of-the-art algorithm, for the specific case of a classical WSN laboratory experiment. The simulation assessment revealed that the cluster-tree formed by DbCTF was able to reduce by more than 20% the average communication delay of message streams conveying critical data, and was also able to increase by more than 35% the average lifetime of the network.

The use of Wireless Sensor Network (WSN) based technologies is an attractive solution for large-scale sensing applications (wide area deployment), such as environmental monitoring, precision agriculture and industrial automation. IEEE 802.15.4/ZigBee standards are the most used communication protocols for WSN technologies, where the clustertree topology is pointed out as a suitable topology to support the implementation of large-scale WSNs. These networks are usually scheduled to prioritise convergecast (upstream) traffic generated from sensor nodes toward the sink node. However, this scheduling pattern results in higher delays for control messages (downstream traffic). Within this context, this paper proposes a Hybrid Beacon Scheduling (Fast-HyBeS) scheme to enable the periodic reconfiguration of cluster-tree WSNs. The underlying idea is to periodically schedule a downstream opportunity window, to allow a faster dissemination of control messages. This opportunity window follows a top-down scheduling approach that prioritises the downstream traffic. Simulation results show that the use of Fast-HyBeS can significantly decrease the end-to-end communication delay for control messages, when compared to the use of static convergecast scheduling schemes. Moreover, the simulation results also highlight that the Fast-HyBeS has a negligible impact upon end-to-end communication delays of the monitoring traffic.

The use of Wireless Sensor Networks (WSNs) is considerably increasing, mainly due to their inherent low cost and wide range of applications. The formation of a WSN is driven by a request-based procedure that associates sensor nodes to a PAN (Personal Area Network) coordinator. In the IEEE 802.15.4 standard, the communication exchanges between a node willing to join the network and the PAN coordinator follows a CSMA/CA protocol. In this paper, we propose a set of specific improvements for the enhancement of the nodes joining procedure. The goal is to reduce the time spent in the association process, minimizing the time spent to deploy the network. A simulation assessment of the proposed set of improvements was performed using the OMNeT++/Castalia simulator, comparing their usage with the default mechanism defined by the IEEE 802.15.4 standard. The achieved results demonstrate that an adequate definition of the parameters of the CSMA/CA protocol significantly reduces the duration of the association procedure.

The IEEE 802.15.4/ZigBee standards support the implementation of cluster-tree networks, which are a suitable topology to deploy wide-scale networks. In this type of topology, an important issue is the configuration of the superframe duration that is allocated to each of the network clusters. In this paper, we propose an allocation scheme for setting-up these superframe duration values. The proposed scheme allocates adequate duration values for each cluster coordinator, based on the message load originated from its child nodes (including child coordinators and its descendants). The target of the proposed allocation scheme is to improve the network throughput and to reduce network congestion and dropped messages. Its main advantages are: 1) to have a balanced allocation of network resources to the different clusters; 2) the message generation at the application layer does not need to be synchronised with the beacon arrival; and 3) the message periods do not need to be multiple of the beacon interval.

In this paper we propose a novel approach, named Real-Time Alternative-Route Definition (RT-ARounD), to support real-time message streams over large-scale wireless sensor networks. Starting from a standard cluster-tree topology, where each cluster has active and inactive periods, the proposal selects a set of nodes to build a mesh topology which is used during the inactive periods of the respective clusters to transmit messages with real-time requirements. This new topology is built considering shortest path routes between source and sink nodes. In that way, the number of hops that a real-time message needs to transverse can be drastically reduced, thereby improving the real-time responsiveness of the overall network. Moreover, the use of the inactive period guarantees that the normal cluster-tree operation remains unaffected.

Brazil has one of the largest private healthcare markets in the world. However, it appears that many indicators of our health insurance companies (including consultations, tests, hospitalizations, etc.) are well above the established international standards. This indicates that many appointments and tests are being carried out without need, generating unnecessary costs on businesses and making the service offered more expensive. In this paper, computational intelligence techniques were used to model the behavior of medical reviewers - professionals who assess whether medical requests should be allowed or not. For the generation of knowledge, a database from a nonprofitable health insurance company containing more than one million records collected since the year 2007 was used. Promising experimental results are presented, indicating that the techniques used can support medical reviewers.