LoRaWAN has become a foundational technology in the Internet of Things (IoT) landscape due to its long-range communication and energy efficiency. However, its default Adaptive Data Rate (ADR) mechanism struggles to adapt to dynamic environments and dense network deployments, where co-channel interference and the coupling between transmission parameters limit its ability to ensure reliable and energy-efficient communication. To address these challenges, this paper proposes MARLTC, an ADR mechanism based on Multi-Agent Rein-forcement Learning (MARL) that jointly optimizes spreading factor and transmission power using realistic observable metrics, including recent transmission history, observed signal-to-noise ratios, and distribution of spreading factors in the network. The transmission configuration problem is modeled as a cooperative Markov Game and solved using the Centralized Training and Decentralized Execution (CTDE) paradigm, where pre-trained policies are deployed at end devices to infer suitable transmission parameters with reduced convergence time. The results using a realistic LoRaWAN simulator show that MARLTC achieves up to 67.9% faster convergence, 62.1% higher energy efficiency, and a 5.0% better packet delivery ratio compared to state-of-the-art approaches, highlighting its scalability and responsiveness in dense deployments. The practical feasibility of MARLTC is further validated using physical LoRaWAN hardware, proving that the resulting policies meet the memory and timing constraints of resource-constrained IoT devices.

Air pollution poses a major environmental challenge, raising concerns about human health in urban environments. It leads to diseases such as asthma, exacerbates pulmonary conditions, and creates murky skies, lowering inhabitants’ quality of life. To quantify air pollution, cost-effective IoT (Internet of Things) devices are being deployed in cities, making air quality monitoring available for a wide range of end-users, including public administrations. However, full urban coverage is unfeasible, and awareness of the carbon footprint of IoT deployments is increasing. Therefore, new techniques are needed to maximize the value of IoT networks with reduced infrastructure. To address these challenges, this paper presents an air pollution analytical forecasting solution based on deep-learning/machine-learning techniques to estimate air quality in locations without deployed devices. Different combinations of well-known deep-learning models are compared with machine-learning techniques to determine the best approach for monitoring polluting gases and airborne particles based on well-defined evaluation metrics. Additionally, two new deep-learning techniques, Multipath-CNN-LSTM (M-CNN-LSTM) and Multipath-CNN-BiLSTM (M-CNN-BiLSTM), are proposed to conduct a more exhaustive comparison. Combinations of LSTM (Long Short-Term Memory) techniques give the best results, with different models working best for each pollutant. Specifically, LSTM was optimal for O 3 , and combinations of CNN (Convolutional Neural Networks) and BiLSTM (Bidirectional LSTM) worked best for NO 2 . GRU (Gated Recurrent Unit) was more efficient for PM 2.5 , and BiLSTM performed best for PM 10 . This demonstrates that the best strategy to accurately predict the time evolution of each pollutant’s behavior depends on the selection of the most suitable machine-learning or deep-learning technique. • M-CNN-LSTM and M-CNN-BiLSTM models are proposed for air pollution forecasting. • Forecasting is performed in locations without deployed sensor devices. • LSTM is optimal for O 3 , M-CNN-BiLSTM for NO 2 , GRU for PM 2 . 5 , and BiLSTM for PM 10 .

The rapid expansion of the Internet of Things (IoT) landscape, notably in smart cities, smart metering, environmental monitoring, and smart agriculture, highlights the critical need for seamless and efficient device roaming within Long Range Wide Area Network (LoRaWAN) infrastructures. Although LoRaWAN shows great potential, its deployment faces significant challenges, especially under the constraints of the version 1.0 specifications, to enable efficient device mobility. This study introduces a comprehensive approach to enhance LoRaWAN-supported mobility, using three key advancements: the decoupling of Gateways (GWs) from the core network infrastructures to facilitate flexible deployments, the adoption of an adaptive GW-to-network connection protocol responsive to real-time traffic demands, and the enhancement of GW interoperability to provide dynamic and efficient network configurations. A notable innovation of this research is developing a mechanism that enables packet forwarding through an unidentified GW, eliminating the need for pre-established network agreements. The empirical evaluations of the article detail the effectiveness of the architecture in IoT applications, showcasing significant improvements in LoRaWAN’s roaming capabilities and reporting average latencies of up to 0.39 s in experimental tests. Extensive simulations with up to 100 roaming devices and 100 stationary devices further confirmed scalability, keeping the average roaming-to-stationary latency gap below 0.16 s and the worst-case end-to-end delay below 1.2 s even in highly congested scenarios. This contribution addresses pivotal challenges within the current LoRaWAN networks and sets the background for future advancements in IoT network technologies.

Even though the initial design of LoRa did not specifically aim at provisioning mobile devices, onboard IoT devices are paving the way for safer, more efficient, and ultimately, autonomous transportation. The LoRaWAN Adaptive Data Rate (ADR) mechanism enables the minimization of energy consumption and the optimization of data rates and airtime values. However, it has been demonstrated to have limited adaptability in mobile devices. In this paper, we propose a proactive ADR mechanism for LoRaWAN based on trajectory estimation and well-known machine-learning methods used to forecast the signal-to-noise ratio (SNR). This enables the efficient selection of the most suitable transmission parameters beforehand. We compare our approach with existing mechanisms in a realistic simulator and several urban environments. Results prove that our method achieves a proper balance between minimizing energy consumption and satisfying performance constraints over time even in highly dynamic and unpredictable environments.

The development of Low-Power Wide-Area Networks (LPWAN) has significantly boosted the uptake of Internet of Things technologies, fostering their adoption in such domains as Smart Cities and Industry 4.0. In the context of LPWANs, LoRa is one of the most promising developments, experiencing massive growth in recent years. This situation has led to the steadily growing trend of redundant deployments across locations, or scenarios, where it would be more natural to be able to leverage (e.g., rent) LoRa infrastructure belonging to a third party. Key to reversing this unsustainable trend is roaming approaches. Although the LoRa specifications define mechanisms for roaming, gaps and challenges make roaming difficult to deploy and operate in real-world scenarios. This paper analyzes these concerns and proposes a novel architecture compliant with the LoRa specification to design and accommodate roaming services in a consistent, efficient, and scalable way.

Flow-guided electromagnetic nanonetworks will enable innovative medical applications for monitoring, information gathering, and data transmission inside the human body. These nanonetworks will have to operate under extreme computational and powering-related constraints, and in very hostile environments inside human vascular systems. Under these circumstances, successful transmissions between in-body nanonodes and an on-body nanorouter rarely occur, thus requiring new approaches to improve the network throughput in this scenario. Along this view, in classical flow-guided nanonetworks the nanonodes are envisioned to transmit packets if they have enough energy for the transmission, regardless of their vicinity to the nanorouter. In this paper, we propose a nanorouter awareness model that can provide significant throughput gains compared to the baseline based on blind transmissions, facilitating the roll-out of nanocommunication-supported medical applications.

In this paper, we propose a flow-guided nano-network deployed inside of a human vascular system. The network consists of bio-sensors reporting medical events, small nano-nodes working as information carriers, and nano-routers communicating with external systems. In order to evaluate the performance and feasibility, a preliminary analytical model for this type of flow-guided nano-network is derived and validated through simulations. Results reveal that the network can meet medical requirements in terms of reported events per minute with a feasible number of nano-nodes.

This paper evaluates the performance of the Cooperative Awareness Message (CAM) dissemination stated in the European Standard EN 302 637-2 in risky sharp bends and winding roads. We propose a novel triggering condition based on the dynamic estimation of the road radius, used as a risk metric. So as lower the radius, the higher the beaconing rate. As a case study, two real road sections, with different lengths and angles, have been simulated to prove that both a better awareness and responsiveness are achieved in the vehicles, to later ensure a proper application layer functioning. Finally, congestion constraint is also tested to check that no significant interferences are found in the described behavior.

Implicit Error Correction (IEC) is a near Video-on-Demand (nVoD) scheme that trades bandwidth utilization for initial playback delay to potentially support an infinite number of users. Additionally, it provides error protection without any further bandwidth increase by exploiting the implicit redundancy of nVoD protocols, using linear combinations of the segments transmitted in a given time slot. However, IEC packet loss protection is weaker at the beginning of the playback due to the lack of implicit redundancy and lower decoding efficiency, resulting in worse subjective playback quality. In tackling this issue, this paper contributes with an extension of the original nVoD architecture, enhancing its performance by adding a new element namely, subchannels. These subdivisions of the original channels do not provide further packet loss protection but significantly improve the decoding efficiency, which in turn increases playback quality, especially at the beginning. Even for very high packet loss probabilities, subchannels are designed to obtain higher decoding efficiency which results in greater packet loss protection than that provided by IEC. The proposed scheme is especially useful in wireless cooperative networks using techniques such as network coding, as content transmissions can be split into different subchannels in order to maximize network efficiency.

By empowering the Internet of Things (IoT) with nanoscale communications, thousands of small devices, called nano-nodes, are enabled to monitor complex environments in a nonintrusive way, giving rise to the Internet of Nano-Things (IoNT) paradigm. It is in one of these very complex and critical environments, the human cardiovascular system, where IoNT might stand out. However, the supervision of human health poses significant challenges: first, to reduce costs and allow in vivo monitoring nano-sensor networks, the number of deployed nano-routers must be limited, which could cause nano-nodes to undergo long out-of-coverage periods. Lastly, nano-nodes rely on harvested energy from the medium, which severely reduces the number of potential transmissions. To overcome these two problems, smart policies that direct devices on how to proceed with sensed events are crucial. In this line, we propose a generic Markov decision process (MDP) model which can be exploited to derive optimal transmission policies that are easily employed by nano-nodes. These policies maximize nano-node throughput and cope with the energy and coverage problems. We have also run a set of simulations to validate our proposal and to compare it to other alternative policies. Results reveal that: 1) our policy systematically outperforms the rest of policies by a large margin and 2) precision in placing nano-routers and average distance between them strongly affect the nano-network performance. The Python code for the MDP model and simulations has been programmed to be easily adaptable to researchers' needs, making it easier for future IoNT works to add intelligence to the network.

LoRa is an extremely flexible low-power wide-area technology that enables each IoT node to individually adjust its transmission parameters. Consequently, the average per-node throughput of LoRa-based networks has been mathematically formulated and the optimal network-level configuration derived. For end nodes to update their transmission parameters, this centrally computed global configuration must then be disseminated by LoRa gateways. Unfortunately, the regional limitations imposed on the usage of ISM bands - especially those related to the maximum utilization of the band - pose a potential handicap to this parameter dissemination. To solve this problem, a set of tools from the machine learning field have been used. Precisely, the updating process has been formulated as a reinforcement learning (RL) problem whose solution prescribes optimal disseminating policies. The use of these policies together with the optimal network configuration has been extensively analyzed and compared to other well-established alternatives. Results show an increase of up to 147% in the accumulated per-node throughput when our RL-based approach is employed.

Low-power wide-area network (LPWAN) technologies enable Internet of Things (IoT) devices to efficiently and robustly communicate over long distances, thus making them especially suited for industrial environments. However, the stringent regulations on the usage of certain industrial, scientific, and medical bands in many countries in which LPWAN operate limit the amount of time IoT motes can occupy the shared bands. This is particularly challenging in industrial scenarios, where not being able to report some detected events might result in the failure of critical assets. To alleviate this, and by mathematically modeling LPWAN-based IoT motes, we have derived optimal transmission policies that maximize the number of reported events (prioritized by their importance) while still complying with current regulations. The proposed solution has been customized for two widely known LPWAN technologies: 1) LoRa and 2) Sigfox. Analytical results reveal that our solution is feasible and performs remarkably close to the theoretical limit for a wide range of network activity patterns.

IoT is rapidly expanding to all productive sectors; however, as thousands of devices are envisioned to be deployed, a careful study of the radio spectrum (and its interferences) must be considered to predict and improve the network performance. In this sense, radio-channel characterization is essential; in particular, path-loss characterization allows to mathematically predict the propagation behavior and interferences of IoT devices with precision. In this paper, we show that by means of a set of mathematical corrections and measuring procedures, we can soundly employ the received signal strength indicator (RSSI), available in most IoT motes, to derive accurate path-loss models of industrial infrastructures. Experimental results reveal that the proposed scheme substantially improves the characterization accuracy of a real smart grid environment. This, in turn, is proven to have a measurable positive impact on the planning stage of an IoT network, whose life-of-time and cost analysis can be better anticipated.

The smart grid (SG) represents the present and the future of electric power grids. Wireless networks (WNs) are a key technology for this updating process, empowering electric networks to be easily managed in a flexible and cost-effective way. However, as with any wireless technology and especially in a harsh multipath environment such as the SG, reckoning with an extensive channel characterization is crucial. This paper presents an exhaustive characterization of two representative scenarios, both in the transport and distribution segment of an SG: a 400-kV outdoor substation and a main power room. The 2.4-GHz ISM band-in which most wireless communication technologies applied to SG operate-is examined in depth to characterize the radio-propagation phenomena. Large and small-scale fading, RMS delay, coherence bandwidth, and electromagnetic interferences are studied under different line-of-sight, polarizations, and frequency conditions to derive empirical models allowing to estimate the aforementioned parameters in other SGs of a similar nature. Finally, to further evaluate and compare our channel characterization, a WN was simulated in a substation to examine the channel modeling impact on its suitable design and operation.

Wireless Mesh Sensor Networks (WMSNs) have recently received a great deal of attention in the scientific and developers communities due to the significant advances of this technology in the wireless communication field. The main reason was that competing WMSN approaches that emerged in the last few years provide mesh capabilities (e.g., robustness, scalability, multi-hop mesh routing, or energy efficiency, among others) to conventional WSN-based applications, encouraging researchers and end users to adopt new perspectives and solutions. Unfortunately, each one of these approaches lacks some (or many) of the aforementioned mesh capabilities, not assuring, a priori, the feasibility and, especially, the long-term stability of WMSN applications. The IEEE 802.15.5 standard and, in particular, its Asynchronous Energy Saving (ASES) mode was conceived to fill this gap since it integrates, in a single solution, most of these capabilities, guaranteeing, among other benefits, a long network lifetime. However, the ASES mode has no built-in mechanisms to mitigate the negative effects of hidden terminals, which sharply degrades the network performance. This fact leads us to conclude that any current WMSN approach is non-exempt of some problem, which prevents the definitive boost of this technology in the consumer market. Under these circumstances, our contribution to the WMSN field in this paper is a twofold proposal addressed to alleviating the hidden terminal problems in a scenario running under the most relevant design premises of ASES mode. Therefore, we first formulate a multi-objective optimization and then solve it by using Goal Programming. Both mathematical techniques are applied to obtain the best solution that simultaneously minimizes the aggregate message collision time due to hidden terminals and maximizes the network lifetime. Secondly, we propose the design of a MULti-channel TIme-scheduled algorithm for the HIdden Terminal problem avoidance (MULTI-HIT) which appropriately exploits the available bandwidth and accomplishes a straightforward coordination between any sender–receiver pair. Finally, the analysis and simulation experiments are presented and their results carefully discussed, demonstrating the effectiveness of both proposals in the WMSN arena.

Wireless personal area networks (WPAN) are wireless sensor networks (WSN) designed for allowing a low-cost, low-power and short-range wireless communication. The IEEE 802.15.4 is the WPAN standard which satisfies these features and facilitates diverse self-configuring operation processes for adapting to different environments and applications. One of these processes is responsible for resolving the loss of coverage of a device, namely an orphan device. It has associated a long procedure when the device has to return to the wireless network, incurring in a high delay and energy consumption. This behavior implies that the Orphan device does not perform its assigned tasks during this delay increasing simultaneously its power consumption. We have designed a new algorithm, called optimized orphan algorithm (OOA), which offers a significant better delay and energy consumption performance. The OOA algorithm is implemented on real devices and evaluated under different scenarios. In this paper, the results obtained are presented and discussed as well.

IEEE 802.15.4 is a wireless technology aimed at the transmission of information at low rates and short distances. Therefore, the expected traffic carried by these networks is one whose transmission rates are low, i.e. information captured by sensors. The extension of this technology to higher data rates is a challenge that gives us the opportunity to support additional services. This paper deals with this issue and evaluates the feasibility of transmitting MPEG-4 video information by computer simulation. Another subject to take into consideration is the computation of the human quality perception when the video streaming is received. To achieve these objectives, the GTS slots of the IEEE 802.15.4 MAC are implemented in the currently available simulation tools. The results obtained are also presented and discussed.

Software technologies to implement distributed multimedia services currently require intensive network and computing resources. In many environments, the design of these services on top of a general-purpose middleware layer is the usual choice. CORBA, Java RMI and .NET are major representatives of this approach. This paper describes and compares the three aforementioned middleware alternatives by evaluating the performance of the implementation of A/V Streaming services on top. From this comparison, several key points and implementation details are highlighted.