Uncrewed Aerial Vehicles (UAVs) are increasingly used in civilian and industrial applications, making secure low-altitude operations crucial. In dense mmWave environments, accurately classifying low-altitude UAVs as either inside authorized or restricted airspaces remains challenging, requiring models that handle complex propagation and signal variability. This paper proposes a deep learning model, referred to as CoBA, which stands for integrated Convolutional Neural Network (CNN), Bidirectional Long Short-Term Memory (BiLSTM), and Attention which leverages Fifth Generation (5G) millimeter-wave (mmWave) radio measurements to classify UAV operations in authorized and restricted airspaces at low altitude. The proposed CoBA model integrates convolutional, bidirectional recurrent, and attention layers to capture both spatial and temporal patterns in UAV radio measurements. To validate the model, a dedicated dataset is collected using the 5G mmWave network at TalTech, with controlled low altitude UAV flights in authorized and restricted scenarios. The model is evaluated against conventional ML models and a fingerprinting-based benchmark. Experimental results show that CoBA achieves superior accuracy, significantly outperforming all baseline models and demonstrating its potential for reliable and regulated UAV airspace monitoring.

This paper addresses the problem of adaptive reconfigurable intelligent surfaces (RIS) configuration design for user localization in rich-scattering environment (RSE), where electromagnetic waves undergo multiple interactions with dynamic scatterers and RIS elements. We propose an adaptive learning-based localization approach for a distributed RIS-assisted network in a RSE using a bidirectional long-short term memory (biLSTM) model that captures temporal correlations between observations. The proposed approach actively senses the environment using sequential pilot transmissions from the base station (BS), accounting for scattering effects, and adaptively updates the RIS configuration based on prior measurements to eventually accurately estimate and minimize the user localization error. The proposed model comprises two neural sub-networks: Scattering Estimation Network (Bi-SEN), for estimation of scattering in the environment, and Adaptive RIS-Assisted User Localization Network (Bi-ARULN), for RIS configuration and localization. Bayesian optimization is used for hyperparameter tuning of the model. The simulation results demonstrate the effectiveness of the proposed approach, achieving significantly lower localization root mean squared error (RMSE compared to random configuration, prestored codebook look-ups, and adaptive baselines in both single-input-single-output (SISO) and multiple-input-multiple output(MIMO) RIS-assisted networks in RSE. The design is generalized across configurations and scales with RIS size and network dimensions. The results highlight the strong potential of RIS deployment and of the proposed approach to enable reliable location services in RSE.

In sustainable wireless communication networks, it is crucial to assess energy consumption during the scheduling of ultra-massive machine-type communications for the forthcoming 6G radio access network (RAN). We introduce an energy-efficient scheduling algorithm for O-RAN that ensures quality of service (QoS) and reduces complexity. Although 3GPP offers models for energy consumption and efficiency in RAN, these models do not consider the effects of radio access techniques. Therefore, it is essential to determine energy consumption cost of successful packet reception using OMA and NOMA, in line with the 6G sustainability goals and the O-RAN framework. We introduce a mathematical model for energy-efficient scheduler and its analysis in managing ultra-massive machine-type communication (umMTC) devices. The run-time and energy consumption of these algorithms are assessed by calculating and comparing their algorithmic complexity, along with the number of operations and data exchanges. These computations provide a basis for forecasting their run-time and energy consumption. The findings indicate that our proposed scheduling algorithm consumes four times less energy than the benchmark schemes ESAP-NOMA and RAAP-NOMA based on computational complexity, while handling the maximum number of users, and offers a 10% increase in energy efficiency with the maximum NOMA users.

5G radio signal strength-and quality measurements can be used to distinguish between Line of Sight (LOS) and non- Line of Sight (NLOS) channels with good accuracy. Classification results, obtained with the use of Machine Learning (ML) methods, can be used to significantly improve positioning accuracy in complex radio environments. Two measurement datasets were collected, first with 5 G modem and second one with dedicated 5 G measurement receiver. At first, we showed that by implementing ML methods the LOS/NLOS classification accuracy of 96% was achieved. Influence of obtained classification performance on positioning accuracy was simulated next. Those results showed that positioning errors can be decreased up to five times by discarding the NLOS measurements, given that sufficient LOS measurements are available.

In this paper, a real-time implementation of Support Vector Machines (SVM) — Real-Time tsSVM Anomaly Detection (RTtsSVM-AD) algorithm is proposed.Here, real-time abnormality detection is referencing to the ability of the algorithm to detect true gait anomaly occurrence during the swing phase of ongoing step. Anomaly detection is presented with ”earliness” measure. For comparative research, eight different human gait deviations were simulated by two healthy volunteers. Corresponding gyroscope angular velocities, from the sensor placed on the forefoot, were recorded. F1 score, true positive rate (TPR), false positive rate (FPR) and ”earliness” values were estimated and analyzed.Real-time classification results, where classification is performed during the ongoing step, are different from regular classification results, where classification is performed after the full step. Thus, they can not be compared directly.Achieved results prove the concept, that it is possible to detect anomalies in real-time during the swing phase of a step with RTtsSVM-AD algorithm. Best F1 scores for first person’s gait recordings were 57%, 53% and 52% for Steppage, Parkinsonian and Ataxic gait types respectively. For the second person’s gait recordings, best F1 scores were 65%, 58% and 50% for Slap, Steppage and Hemiplegic gait types respectively.RTtsSVM-AD algorithm would be developed further and could be used as a base method for comparison with other algorithms.

Device-to-device (D2D) enables direct communication between two-user equipment (UEs) with or without the involvement of a base station (BS). D2D communication is a vital paradigm for designing a reliable public safety network (PSN) and supports several services in the sixth generation (6G) such as target monitoring, emergency search and rescue, etc. This paper presents the measurement campaign to characterize the performance of ProSe direct discovery in terms of connectivity. Furthermore, we implement and analyse two supervised learning models: multiple linear regression (MLR) and multiple non-linear regression (MNLR) using the least squares method to predict the probability of direct discovery in PS out-of-coverage scenarios. Experimental data collected in real-time heterogeneous environments has been used to train both models. The comparative analysis indicates that MNLR model is 6 % more efficient compared to MLR model to predict the discovery probability. The impact of this work is that it is possible to deploy the equipment to build reliable connectivity for efficient D2D networks in out-of-coverage emergency scenarios, and can bring intelligence to the UE level to adopt the optimal parameters to establish a stable D2D link.

Network slicing in 5G RAN enables building logical networks on top of physical infrastructure. In case of RAN spectrum resources are divided into small bandwidth parts then, the network function in core network, namely Network Slice Admission Control Function (NSACF) enables to limit the count of user equipments (UE) and packet data unit (PDU) sessions in a slice. This paper presents negative results and a practical analysis of how many UEs and PDU sessions can be allowed to use if available bandwidth is fixed and tries to evaluate the optimal count of subslices at this bandwidth. In case of slice resource overutilization there are 2 options: increase resources or move UEs to another slice. In this paper, the feasibility of the other option is evaluated in case of fixed overall bandwidth where resource increase is not possible. Results show that infrastructure can afford as few slices as possible and slicing does not help if theoretical capacity is exceeded.

Smart body area networks (SmartBAN) is an emerging wireless body area networks (WBAN) standard proposed by the European Telecommunications Standards Institute (ETSI). This paper first examines the potential of SmartBAN medium access control (MAC) layer with scheduled access to support a myriad of WBAN applications, having diverse data rate requirements. Extra scheduled access slots can be allocated to high date rate sensor nodes for managing their data rate requirements. High data rate sensor nodes can also be re-assigned to use the available time slots of low data rate sensor nodes in Inter-Beacon Interval (IBI) by the central hub. But these two schemes incorporate different physical (PHY) and MAC layer overheads related to frame transmission, frame acknowledgement and slot re-assignment. This redundant overhead transmission results in high overhead energy consumption and reduced effective throughput. Therefore, an innovative and flexible enhanced throughput and reduced overhead (FETRO) MAC protocol for scheduled access is proposed in this article. In the proposed scheme, the sensor node data rate requirements are considered while assigning the scheduled access slot duration by allowing minimal changes in the base-line standard implementation. This infers the provision of scheduled access slots with variable slot durations within an IBI. We also evaluate the existing techniques of extra slot allocation and slot re-assignment in SmartBAN as well as the proposed FETRO MAC protocol with variable slot length. The proposed FETRO MAC scheme results in optimizing both the overall throughput and normalized overhead energy consumption per kilo bits per second (Kbps). Additionally, the impact of various WBAN channel models over these throughput management approaches is also investigated. The proposed FETRO MAC protocol with variable slot duration gives an average reduction of 65.5 and 59.16 percent, respectively, in the hub and nodes normalized overhead energy consumption per Kbps outcomes, as compared to the de-facto SmartBAN MAC scheduling strategies.

In this paper, we propose an inter-cell interference (ICI) minimization scheme for uplink transmission in the narrowband internet of things (NB-IoT) systems. We first establish the theoretical ICI problem formulation and propose its corresponding solution for the orthogonal multiple access (OMA) NB-IoT system. Based on the theoretical formulation, we design a cooperative radio resource scheduler that reduces the impact of ICI and allocates transmit powers to reduce the energy consumption to the scheduled users in a multi-cell scenario. We compare the performance of the proposed scheme with that of some benchmark OMA schedulers. The results show that the proposed technique significantly reduces the impact of ICI and hence is more suitable for the massive connectivity of the NB-IoT system. For example, the users operating under the proposed approach experience up to 50% reduced energy consumption when compared to the best channel quality indicator (CQI) scheme. Furthermore, 30% and 35% improvements in terms of achieved user's data rates are obtained as compared to the MaxMin and round-robin schemes, respectively.

NarrowBand Internet of Things (NB-IoT) is an emerging cellular IoT technology that offers attractive features for deploying low-power wide-area networks suitable for implementing massive machine-type communications. NB-IoT features include, e.g., extended coverage and deep penetration for massive connectivity, longer battery-life, appropriate throughput, and desired latency at lower bandwidth. Regarding the device energy consumption, NB-IoT is mostly underestimated for its control and signaling overheads, which calls for a better understanding of the energy consumption profiling of an NB-IoT radio transceiver. With this aim, this work presents a thorough investigation of the energy consumption profiling of the radio resource control (RRC) communication protocol between an NB-IoT radio transceiver and a cellular base station. Using two different commercial off the shelf NB-IoT boards and two mobile network operators (MNOs) NB-IoT test networks operational at Tallinn University of Technology, Estonia, we propose an empirical baseline energy consumption model. Based on comprehensive analyses of the profile traces from the widely used BG96 NB-IoT module operating in various states of the RRC protocol, our results indicate that the proposed model accurately depicts the baseline energy consumption of an NB-IoT radio transceiver while operating at different coverage class levels. The evaluation errors of our proposed model vary between 0.33% and 15.38%.

In an effort towards designing a batteryless Internet of Things (IoT) sensor node that is powered by miniaturized energy-harvesting source(s), we combine the techniques of transient computing, approximate computing, data and energy predictions so as to handle the unpredictable power shortages of the miniaturized energy harvesting sources and reduce the overall power consumption of the IoT node. To evaluate the feasibility of our proposed approach, we build upon and extend an existing platform that consists of a peer-to-peer network (a sender node and a receiver node) where each of these nodes combines a Texas Instruments' FRAM-based micro-controller with a low cost, low power radio module and exchanging its data through SimpliciTI protocol. Our results illustrate that combining transient computing, approximate computing, data and energy predictions adds up their individual benefits to achieve an overall better utilization of the harvested energy of the node. Our results show that out of the total 60 transmissions that were due in an interval of 5 hours, for sending the temperature data from sender node to the receiver node every 5 minutes, a total of 32 transmissions were avoided, leading to a saving of more than 50% of the radio transmissions in the sender node.

The high mobility of Unmanned Aerial Vehicles (UAVs) and their capability to rapidly deploy Aerial Base Stations (ABS) in areas where the terrestrial network becomes unavailable is a key enabler for Public Safety Networks. In our work we introduce a model in order to identify Line of Sight (LoS) and Non-Line of Sight (NLoS) conditions for User Equipments (UEs) that attempt a connection to an ABS through the Physical Random Access Channel (PRACH) based on Convolutional Neural Networks (CNNs). Our method limits the number of antennas employed with respect to other methods that were developed for traditional approaches, while achieving higher than 80% accuracy for SNR of -20 dB. Finally, we study the impact of UAV’s height on the accuracy of our method and we compare it with typical computationally efficient methods based on the delay spread with and without the aid of beamforming.

Recent advancements in cellular technologies allow discrete computing devices embedded with sensors to communicate over long distances with low-cost and low-energy consumption. This could drastically impact the future internet of thing (IoT) ecosystem. In this regard, Third Generation Partnership Project (3GPP) has introduced a new cellular-based technology called Narrowband Internet of Things (NB-IoT) which is one of the potential technologies for enabling IoT application in vehicular, health-care, industry 4.0, etc. However, NB-IoT technology is still in its infancy and so far it is unclear whether it is sufficiently reliable to complement or replace existing short-range and cellular technologies to enable such use-case scenarios. Therefore, this paper presents a preliminary investigation of the coverage of NB-IoT in three different scenarios i.e., outdoor, indoor, and underground with empirical measurements, one of the key performance indicators in which operators are most interested. The measurements were conducted on Tallinn University of Technology (TalTech) campus and nearby residential areas within a range of 700 m. The obtained results indicate that NB-IoT is able to provide good connectivity to meet the IoT application requirements in outdoor and indoor environments. However, for an underground scenario with the beam enabled for 700 m, it is only possible to provide connectivity to the devices up to 400 m. Furthermore, it is also observed that NB-IoT is able to provide connectivity to devices with a received signal strength indicator (RSSI) value as low as −105 dBm as compared to −95 dBm as in long-term evolution (LTE).

A new backoff scheme named Unique Backoff Algorithm (UBA) has been introduced in this paper for IEEE 802.15.6 Wireless Body Area Networks (WBANs). This scheme tries to eliminate the collision among nodes by assigning unique backoff values. UBA removes the large delays during transmissions between different nodes and hence increases overall system performance. Furthermore, the concept of Backup Slot ( BS) has also been introduced in the superframe to accommodate one failure of highest priority in the current superframe. For evaluation, proposed scheme has been simulated in comparison with the Binary Exponential Backoff (BEB) using different priorities. Results show that our scheme outperforms BEB in terms of throughput and superframe efficiency.

During a disaster incident, existing network infrastructures might either be damaged or overloaded. However, disaster operations require real-time and low latency data communication, especially live video and audio streaming, as well as in integration with social networks. Thus, deployed tactical networks face several challenges to connect intervening teams. To date, there is still a lack of routing standards for such networks. Indeed, in such harsh environment, medical rescue teams, firefighters, military, police and even victims need to be steadily connected to a distant command center (CC) which conducts the rescue operations. In our recent works, a new routing protocol called ORACE-Net was proposed, evaluated and compared to the existing approaches. ORACE-Net aims to create a tactical Internet of Humans (IoH) emergency network. ORACE-Net is implemented on mobile devices and evaluated within realistic conditions. The complete proposed solution consists of the CROW2, an Internet of Humans-based platform for disaster relief and emergency communication. In this paper we describe the setup and the demonstration of the CROW2platform based on the our conducted experimentation.

This paper addresses a reliability concern of the emerging wearable applications under dynamic and realistic conditions. We propose a new joint throughput and channel aware (TCA) dynamic scheduling algorithm for IEEE 802.15.6 standard to enhance the system performance while exploiting m-periodic scheduled access mechanism. First, various mobility patterns are generated with special emphasis on space and time varying links and their performance which are most vulnerable under dynamic environment. A deterministic pathloss values (as an estimate of the channel) are obtained from a set of training sequence which is generated using motion capture system and bio-mechanical modeling. Consequently, signal to noise (SNR), bit error rate (BER) and packet error rate (PER) are calculated. The TCA algorithm during the first phase uses this estimated PER to select the potential nodes for a time slot. Whereas in the second phase, based on the nodes priority and the data packets availability among the potential candidates, finally a slot is assigned to one node. This process is repeated by the coordinating node until the end of the super frame. This scheme has a significant gain over a reference scheme (i.e., without above adaptation). On average, 20-to-55 percent extra packets are received, along with 1-to-5 joules of energy savings though at the cost of higher delay ranging from 20-to-200 ms while operating at low power levels (i.e., 0 dBm, -5 dBm, -10 dBm).

Providing efficient routing service over tactical multi-hop ad-hoc networks is a crucial building block in wireless communication networks especially during a disaster relief. To date, there is still a lack of routing standards for such networks. Indeed, in such harsh environment, medical rescue teams, firefighters, military, police and even victims need to be steadily connected to a distant command center (CC) which conducts the rescue operations. In this paper, we propose a new multi-hop routing approach called ORACE-Net. The proposed protocol uses advertisement packets to establish routes from deployed nodes towards the CC (i.e. Direct Route Establishment). Then, it utilizes the data packets to establish reverse routes (from the CC to all nodes in the network). We implemented and evaluated our approach in realistic scenario using tactical and on-body mobile nodes. Our experiments include also an Internet of Thing (IoT) platform and a real-time dynamic topology website which are used for analyzing the behavior of the protocol. The experimental results show that our protocol increases the mobile nodes connectivity and packet delivery rate. Also, it reduces the average round trip time delay for the on-body nodes compared to the tactical deployed base stations.

In this paper, a new multi-hop routing protocol (called ORACE-Net) for disaster and emergency networks is proposed. The proposed hierarchical protocol creates an ad-hoc network through body-to-body (B2B) communication between rescue members and the command center. The on-body coordinators establish forward routes towards a command center, whereas, the command center builds reverse routes towards these nodes. The Routing tables are optimized based on real-time end-to-end Link Quality Estimation (LQE) metrics (i.e., end-to-end signal strength level and end-to-end hop count). We evaluate our proposed protocol with other widely used protocols in the disaster context which are covered by the routing classes (i.e., Reactive, Proactive and Geographic-based). The evaluation is based on a realistic disaster mobility trace. The results show that the proposed protocol outperforms the other studied protocols in terms of packet reception rate and energy consumption. The proposed ORACE-Net protocol increases the body-to-body network lifetime and reliability.

With the recent advent of Internet of Humans (IoH), wireless body-to-body networks (WBBNs) are emerging as the fundamental part of this new paradigm. In particular with reference to newly emerging applications, the research trends on data routing and dissemination strategies have gained a great interest in WBBN. In this paper, we present the performance evaluation of the clustered and distributed data dissemination approaches in tactical WBBN. We used a realistic radio-link and bio-mechanical mobility model for on-body motions, and group mobility model for WBBN to effectively realize rescue and emergency management application scenario. In this regard, we are using the newly proposed IEEE 802.15.6 standard targeted for body area networks. Extensive (IEEE 802.15.6 standard compliance) network level, packet oriented simulations are conducted in WSNet simulator. During the simulations, various payloads, frequencies (narrow-band) and modulation techniques are exploited. We based our performance evaluation on relevant metrics according to the operational requirements for tactical networks such as packet reception ratio, latency, energy consumption and hop count. The results showed a trade-offs between clustered-based and distributed-based dissemination approaches. With regards to packet delay, distributed approach provided the best performance. However, in terms of average packet reception ratio (PRR), clustered-based approach achieves up to 97% reception and remained the best strategy. Whereas, almost double energy is consumed while using clustered approach, though this is due to multi-standard coordinator which is considered as more powerful device. Finally, the results of hop count are almost comparable in both schemes.

Critical and public safety operations require realtime data transfer from the incident area(s) to the distant operations command center going through the evacuation and medical support areas. Any delay in communication may cause significant loss. In some cases, it is anticipated that the existing communication infrastructures can be damaged or out-of-service. It is thus required to deploy tactical ad-hoc networks to cover the operation zones. Routing data over the deployed network is a significant challenge with consideration to the operations conditions. In this paper we evaluate the performance of multi-hop routing protocols while using different wireless technologies in an urban critical and emergency scenario. Using a realistic mobility model, mobile ad hoc, geographic based and data centric routing protocols are evaluated with different communication technologies (i.e. WiFi IEEE 802.11; WSN IEEE 802.15.4; WBAN IEEE 802.15.6). It is concluded that, WiFi IEEE 802.11 is the best wireless technology with regards to the packet reception ratio and the energy consumption. Whereas, in terms of delay, WBAN IEEE 802.15.6 is the most efficient. With regards to the routing protocols, assuming that the location information is available, geographical based routing protocol with WiFi IEEE 802.11 performed much better compared to the others routing protocols. In case where the location information is unavailable, gradient based routing protocol with WBAN IEEE 802.15.6 seems the best combination.

In this paper a realistic performance evaluation of the IEEE 802.15.6 Medium Access Control (MAC) protocol for wireless body sensor networks (WBSNs) is presented. The accuracy of the evaluation rely on the enhanced mobility and radio link models, which are based on real-time motion capture mobility traces. Bio-mechanical modeling is used to capture 'walking', 'stand-sit' and 'running' patterns for comprehensive mobility modeling. There are multiple contributions in this paper. First, dynamic (i.e., space and time-varying) mobility models are presented which provides dynamic distances and hence more accurate pathloss models in comparison to IEEE 802.15.6 channel models. Second, accurate radio-link modeling is presented which computes on-line Signal-to-Noise-Ratio (SNR), Bit-Error- Rate (BER) and Packet-Error-Rate (PER). Third, a comprehensive analysis of the IEEE 802.15.6 PHY and MAC layers parameters are explored. The rescue and critical applications based specific constraints are used to analyze the IEEE 802.15.6 standard. Finally, few configurations among huge set of possibilities are selected for performance evaluation. Three metrics i.e., Packet Delivery Radio (PDR), energy consumption and latency are considered. At higher transmission power (i.e., 0 dBm), most of the medical applications data rates constraints are satisfied, whereas, at lower transmit power with high data rates and higher frequency, the results does not meet the non-medical applications requirements. However, further optimization by dynamically adjusting the number of slots and its duration at the MAC layer can improve the throughput which can help significantly to meet the application requirements.

This paper provides an overview of the project Critical and Rescue Operations using Wearable Wireless sensor networks (CROW2) and discusses the related research challenges and upcoming objectives. The project objective is to provide ubiquitous wireless communication and monitoring systems enabling life-critical and rescue operations. In this context, it is envisioned that wearable Wireless Sensor Networks (WSNs) will play an important role, and opportunistic exploitation of both On-Body and Body-to-Body communications will enable beyond state-of-the-art and cross-layer communication architecture for wearable WSNs.

In this paper accurate mobility and radio link modeling for intra-body communication is presented. The proposed enhanced mobility and radio link models introduce space and time variations to existing IEEE 802.15.6 channel models which can be used for better performance evaluation of new and existing algorithms and protocols. The contributions of this paper are in many folds. First, an accurate pathloss model is developed using real-time motion capture trace which provide dynamic distances for the space-time variations and temporal correlation in time. Further, the accurate mobility model is used to distinguish Line of Sight (LOS), Non Line Of Sight (NLOS) and periodic (or time varying) radio links. Second, a comprehensive radio link modeling approach is proposed, including the computations of SNR (Signal to Noise Ratio), BER (Bit Error Rate) and PER (Packet Error Rate) which are accurately evaluated after achieving realistic pathloss across all the links locations. Finally, we assess the impact of the proposed enhanced models on the performance evaluation of the IEEE 802.15.6 MAC protocol. The obtained results show between 10% to 30% lower packet delivery (depending on the considered mobility scenarios and physical layer parameters) due to accurate modeling.

This paper investigates the issues related to the link quality estimation and routing performance in highly dynamic WBSNs, where the inherent body mobility introduces frequent time and space variations in the observed link reliability and network connectivity. A new Opportunistic and Mobility Aware Routing (OMAR) protocol is proposed, which optimizes the routing process by taking benefit from captured long-term and short-term link behavior. Extensive simulations are performed using realistic WBSNs mobility and channel modeling, and representative MAC constraints. The obtained results show that the proposed OMAR approach outperforms conventional routing protocols in terms of energy consumption and routing path length, while ensuring a stable packet delivery ratio.

This paper presents a self organized asynchronous medium access control (MAC) protocol for wireless body area sensor (WBASN). The protocol is optimized in terms of latency and energy under variable traffic. A body sensor network (BSN) exhibits a wide range of traffic variations based on different physiological data emanating from the monitored patient. For example, electrocardiogram data rate is multiple times more in comparison with body temperature rate. In this context, we exploit the traffic characteristics being observed at each sensor node and propose a novel technique for latency-energy optimization at the MAC layer. The protocol relies on dynamic adaptation of wake-up interval based on a traffic status register bank. The proposed technique allows the wake-up interval to converge to a steady state for variable traffic rates, which results in optimized energy consumption and reduced delay during the communication. A comparison with other energy efficient protocols is presented. The results show that our protocol outperforms the other protocols in terms of energy as well as latency under the variable traffic of WBASN.