Location-based services (LBSs) play a vital role in many Internet applications. Privacy is a mandatory aspect of these tasks, including protecting the user's sensitive information from malicious entities. Location privacy-preserving mechanisms (LPPMs) were designed to ensure privacy for LBS users, and several strategies have emerged, such as cloaking mechanisms. Likewise, several attacks appeared to threaten the user's privacy, being based on ground-related aspects. Therefore, we must investigate new strategies to enhance privacy protection mechanisms. Unmanned Aerial Vehicles (UAVs) can provide assisted coverage to ground users in different tasks, including the support of LPPMs. However, the existing strategies rely on some unfeasible premises, and this collaboration needs to be adequately explored. Therefore, in this study, we propose Sky Cloaking, which promotes the opportunistic connection between ground users and UAVs in such a way the UAVs manage the user's query, creating a cloaking region and hampering the success of an attacker. Through a comprehensive evaluation, we demonstrated that SkyCloaking can ensure high levels of location privacy to LBS users with a slight impact on the communication channel, overcoming existing strategies. In the best scenarios, SkyCloaking protected more than 80% of the user trajectory, mitigating the exploitation of users' sensitive information.

Unmanned aerial vehicles (UAVs)–also known as drones or Unmanned Aircraft–have found diverse applications in various fields owing to their significant advantages, including fast mobility and rapid deployment. UAVs are crucial in aerial networks, providing increased coverage and on-demand connectivity as mobile nodes. In recent years, UAVs have made room to leverage the Internet of Things (IoT) to the sky, enhancing air-to-ground communication and pointing toward the next generation of UAV networks. As this expansion is still in its early stages, there are several aerial network terminologies, each with similarities and differences, depending on the deployment domain and the services they offer. However, studies have yet to discuss these different terminologies consistently. Key aspects have yet to be thoroughly explored, such as the anticipated requirements for deploying these networks and how they relate to the various terminologies. This work systematically analyzes the existing terminologies of UAV networks, considering their requirements and applications, shedding light on their intersections and differences. Furthermore, we present the demands for the next generation of UAV networks and discuss how they impact the design of UAV-related applications, aiding in the design of new protocols, tools, and technologies for both industry and academia. Lastly, we highlight the emerging trends and challenges associated with deploying and integrating these networks.

Location Privacy Protection Mechanisms (LPPMs) have been designed to enhance privacy in the Internet of Drones (IoD), however, they present suitable privacy levels only in specific network conditions. Also, they can lead to a lack of Quality of Service (QoS) if applied in unfavorable conditions. Thus, the dynamic assignment of the most suitable LPPM, given the IoD conditions, is a significant challenge. Reinforcement Learning (RL) represents a useful concept to handle this problem, given its exploratory characteristics and being able to enhance the knowledge about network dynamics. In this study, we propose IoDAPM, an RL-based approach for the Dynamic Assignment of Protection Mechanisms in IoD. Through simulations, we extensively trained the RL-based model, exploring the possible IoD network conditions. With this model, we carried out a comparative evaluation of existing LPPMs. The results highlighted that IoDAPM outperforms the compared mechanisms considering the QoS, providing enhanced performance regarding location privacy, energy efficiency, and flight delay.

In Unmanned Aerial Vehicles (UAVs), the real-time detection and identification of unauthorized UAVs is a significant challenge to be appropriately addressed. Currently, supervised-based learning models (e.g., Deep Neural Networks) can detect the presence of authorized UAVs with reasonable accuracy. Still, they can not handle properly the wide range of unknown signals in the airspace, mainly their categorization. Clustering techniques (e.g., DBSCAN) can be applied to identify and classify unfamiliar signals. However, the uncertainty regarding the nature of unknown sounds can lead to a large dimensional problem, hampering the performance of these techniques. Given these issues, we proposed DissIdent, a dissimilarity-based method for identifying unknown drones. Our approach takes advantage of the dissimilarity concept, in which a function of proximity maps extensive and multi-dimensional problems to a binary problem. DissIdent can identify patterns from different features through an intelligent workflow, mitigating the trade-off between the traceability and accuracy of massive multi-class problems. We carried out an extensive evaluation of DissIdent, comparing it with eight different approaches. The results pointed out DissIdent as a robust approach to detection and identification tasks, overcoming the compared methods. DissIdent addressed accuracy rates higher than 93% in all scenarios, presenting a concise detection and identification of unauthorized drones.

The Internet of Drones (IoD) is a network paradigm that allows drones to perform several services, gathering and sharing location-based information, representing a piece of the next generation of the Internet of Things (IoT). Location privacy is a paramount requirement in IoD. A few Location Privacy Protection Mechanisms (LPPMs) are designed for IoD. Unfortunately, they are not energy-aware approaches, essential for IoD protocols since drones have power limitations. We present the MixRide, an energy-aware LPPM for IoD to overcome these issues. It provides location privacy through the aerial-grounded vehicle collaboration, where the drones take a ride with grounded vehicles, changing their pseudonyms while saving energy. A comparative experimental evaluation pointed out that MixRide can provide location privacy to the IoD at the same level as the state-of-the-art LPPM while improving the drone's power consumption. Our results also provide new insights on the trade-off between the delay caused by the ride and the drone's energy power consumption.

The recent advancement of drone technologies and communication protocols allows us to envision a robust and dynamic mobile vehicular network paradigm called the Internet of Drones (IoD). In this environment, drones will perform several location-based services (LBSs) for users, awakening the interest of malicious entities whose intention is to hamper the service. Hence, drones need high protection regarding their localization in LBSs. However, there is a lack of Location Privacy Protection Mechanisms (LPPMs) for an IoD scenario. Dummy-based LPPMs provide proper location privacy in traditional mobile networks, mainly for sparse configurations. The design of this mechanism for IoD can overcome this deficiency. This study proposes a novel dummy-based LPPM for the IoD, called TDG, that focuses on the IoD topology characteristics regardless of near drones being the first approach presented in this context. Through extensive experiments, we show that TDG can provide proper location privacy for drones in sparse configurations, reducing the use of the wireless communication channel. TDG can protect the real drone’s trajectory up to more than 90% of the time, leaking less than 25% of the drone’s real coordinates.

The Internet of Drones (IoD) refers to a robust mobile network with well-defined airways where drones perform interoperable services, enhancing the deployment of Smart Cities. Automatic Drone Identification (ADI) is a protection mechanism to detect and avoid malicious drones, where different techniques have been used, such as acoustic signals. In this field, the sound generated by propellers and motors has particular characteristics, being a potential aspect to explore; however, this investigation is still missing. This study examines the use of rhythm-based descriptors as input features to ADI, based on the hypothesis that the acoustic signal generated by different drones has different rhythmic properties. Aiming to explore and validate our approach, we formulate an ADI methodology using rhythm-based features. We use a freely available drone audio dataset, comparing our results with a baseline study. As a result, our classification model improves 3.47% the baseline binary classification and 2.97% the multiclass classification, reaching accuracy rates of 0.9985 and 0.9591, respectively. Although the improvements are narrow, they point out that acoustic features have a great potential to enhance ADI mainly in dense drone-based environments, where drone identification is an essential task.

The Internet of Drones (IoD) is a novel network paradigm that presents a unique mobile network scenario with particular characteristics. Hence, privacy is a mandatory aspect to be assured, but there is a lack of studies regarding location privacy protection in IoD. Mix Zones is a location privacy protection mechanism well explored in terrestrial mobile networks, however, its investigation in IoD is still missing. In this work, we propose a novel Mix Zones-based approach, called MixDrones, that changes the airway of a drone besides its pseudonym. Hence, we advance the state of the art of location privacy protection mechanisms for IoD, in which MixDrones is the first approach proposed in this context. We carried out experiments through simulations comparing our approach with the traditional mechanism regarding anonymization coverage and resilience. We also evaluate the possibility of drone collision occurrence. The results pointed out that MixDrones provided a better location privacy protection than the traditional mechanism, anonymizing a large number of drones and being resilient through a trajectory-based de-anonymization attack, with less than 25% of trajectories being de-anonymized in all scenarios. Moreover, MixDrones mitigated the side effects of airway change, presenting low rates of airways change competition situations.

This paper proposes the MannAccess, a novel low cost assistive educational tool of digital image for Visually Impaired (VI), aiding the teaching-learning process of visual content. It consists of an assistive environment composed of interactive software and a refreshable pin display with a novel 3-axis pin activation mechanism, decreasing its development cost substantially. The MannAccess allows the integration with different image recognition methods using a proposed image intermediary representation. In order to accomplish evaluations, we developed a prototype of MannAccess and integrated it with MannAR, an automata image recognition method. We carried out experiments with VI students which pointed out that our tool provided proper accessibility, usability, and user experience. In addition, we accomplished a monetary cost evaluation, indicating that MannAccess had the most accessible monetary cost compared to related devices. In a nutshell, MannAccess showed that it is possible to develop a low-cost assistive technology to aid the VI visual content education, integrating different image recognition methods in a single assistive tool.

This work introduces a Brazilian Speech Database (BrSD), a novel dataset freely available created to support the development of speech-based recognition tasks. As far as we know, this is the first Portuguese language based database with these characteristics created and made available to the research community. We also describe experiments accomplished on BrSD exploring its different possibilities of classification tasks, i.e., age group and gender classification. We use four well-known acoustic features extracted directly from the audio signal and one texture-based feature extracted from a visual representation of the audio signal, the spectrogram. We considered three different classification scenarios: each feature individually, early fusion of the features, and late fusion of the features. Experiments were conducted using Support Vector Machine (SVM) and Multi-layer Perceptron (MLP) classifiers. The obtained results showed that SVM classifier achieved the best recognition rates both in early and late fusion scenarios. The best recognition rates achieved were 91.25%, 88.75%, and 80.25% for gender, age group, and age-gender classification tasks, respectively.