A selective neighbour channels in Wi-Fi networks based on adaptive machine learning techniques

The knowledge of the statistical characterization of the primary user activity can make more efficient the spectrum selection techniques, which are a key technology to improve the Cognitive Radio (CR) systems performance. In order to provide this kind of primary user information, Radio Environment Maps (REM) data bases has been proposed. This paper presents a REM-based Real Time Testbed designed to validate and exploit the dynamicity of the REM data base to improve the performances of the Radio Resource Management (RRM) algorithms in the context of cognitive radio. The implemented testbed is a flexible HW/SW tool able to use spectrum measurements in licensed bands and validate the Cognitive Radio performance of the secondary users, without causing any interference on these primary users. In particular, the RRM strategy under study in the testbed envisages the optimization of the spectrum selection policy when secondary users, with different service profiles, access opportunistically a set of channels during the inactivity periods of primary users. The selection is done taking as input the statistical characterization of the primary user activity, which will be stored in the REM data base, and the secondary user service profiles.

Cognitive Radio (CR) is considered as a promising solution to improve wireless spectrum utilization. Moreover, through optimization frameworks, Radio Resource Management (RRM) procedures are designed to enhance the efficient utilization of resources in CR networks. To face this issue in a holistic perspective, the RRM solutions should take into account not only the optimization task but also their impact on the design task. Thus, in this paper, a system architecture of a novel RRM is proposed for CR networks. In particular, it is fundamental to construct flexible systems to work across heterogeneous systems and help realize their full potential. Therefore, in the proposed system, heterogeneous Primary Users (PUs) with multiple features and variable CR demands are considered. A Radio Environment Map (REM) is used to obtain the required PU features, which are exploited to improve the adaptability in CR networks and, thus, to design an efficient Cognitive RRM. The functionalities of the network elements are described in detail, along with the message exchange to carry out the resource management.

Recent advances in cognitive radio have identified the small-cells among the most promising future wireless networking scenarios. Utilizing radio context information, smallcells should perform the most optimal radio resource management (RRM) to maximize performances and minimize inter-cell interference. Radio Environmental Maps (REM) data: empirical propagation models, active transmitters’ locations, upto-date interference levels, statistical channels occupancies, are especially beneficial in these scenarios. The proposed demonstration aims to showcase the benefits of using REM information in the small-cell optimization. The demonstration utilizes a modular/flexible REM prototype, performing a realtime REM data acquisition, processing and inference as input to an enhanced small-cell optimization.

Unmanned Aerial Vehicles (UAV), also known as drones, are receiving increasing attention as enablers for many emerging technologies and applications, a trend likely to continue in the next future. In this regard, using Unmanned Aerial Base Stations (UABSs), i.e. base stations carried by UAVs, is one of the most promising means to offer coverage and capacity in 5G applications to those users that are not being served by terrestrial base stations. In this paper, we propose a novel approach for trajectory design and Radio Resource Management (RRM) in UAV-aided networks using information retrieved from precise Radio Environmental Map (REM) based on Ray Launching (RL) simulations for RF propagation and narrow band estimations. Furthermore, we consider different possible models for antennas to be installed on multiple UABSs as well as proper RRM strategies which are able to take advantage of REM inputs. Simulation results will show the performance achieved by the system for the different approaches and it will compare them with the previous use of statistical models.

Chapter 11 - Network Support: The Radio Environment Map

This paper focuses on femtocell self-organization through radio environment map (REM) based power control. Different downlink power control strategies with REM have been studied with realistic LTE system evaluation. We have shown that downlink power control can be an effective means for femtocell self-organization. With the help of REM information, the performance between macrocell and femtocell is better balanced. Our study suggests that REM-based power control is useful in practical femtocell network planning, in which a careful balance should be struck in maintaining satisfactory network performance and user QoS (quality of service) in heterogeneous networks.

This paper presents the indirect methods for constructing radio environment maps (REMs), which utilize known model information, to first estimate the primary transmitter parameters and then generate REMs. Two indirect methods under lognormal shadowing are presented and compared. The better of these two methods is further investigated in different scenarios. These scenarios include different number of sensors, varied size of measurements, several shadowing spread values, different percentages of error in path-loss exponent, and the effect of the number of moving sensors and their speeds to the REM quality. The results show that performance is enhanced as the number of sensors and the size of measurements increase, whereas clear degradation in REM quality is shown when shadowing spread increases or the model parameters are not well calibrated. Also, as the number of moving sensors or their speeds increase, the REM performance becomes less effective.

In future networks, Radio Resource Management (RRM) could benefit from Geo-Localized Measurements (GLM) thanks to the Minimization of Drive Testing (MDT) feature introduced in Long Term Evolution (LTE). Such measurements can be processed by the network and be used to optimize its performance. The purpose of this paper <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">a</sup> is to use GLM to significantly improve scheduling. We introduce the concept of forecast scheduler for users in high mobility that exploits GLM. It is assumed that a Radio Environment Map (REM) can provide interpolated Signal to Interference plus Noise Ratio (SINR) values along the user trajectories. The diversity in the mean SINR values of the users during a time interval of several seconds allows to achieve a significant performance gain. The forecast scheduling is formulated as a convex optimization problem namely the maximization of an α-fair utility function of the cumulated rates of the users along their trajectories. Numerical results for thee different mobility scenarios illustrate the important performance gain achievable by the forecast scheduler.

This paper addresses the problem of scheduling a distributed set of RF sensors for the creation of a multichannel Radio Environment Map (REM). We propose a method for tracking the spatial distribution of the activity patterns on each channel via Kaiman filtering. The activity patterns on each channel are then used to inform schedule creation by balancing the achievable expected uncertainty across all channels by using the concepts of Permissible Consecutive Observation Loss (PCOL) and Least Consecutive Observation (LCO). For a stationary wireless sensor network, where each sensor can scan one channel at a time, we create an adaptive stochastic channel sensing order that balances the spatial uncertainty of the estimate of the REM for each of the channels of interest. Herein, the algorithm's effectiveness is demonstrated through simulation. It will also be demonstrated on actual spectrum measurements under the DARPA RadioMap, for which this algorithm was developed.

In this paper, we discuss a method to construct a radio environment map (REM) in an environment with multiple primary users (PUs). The REM provides statistical information about the PU activity at each location. It enables the secondary user to access the licensed band dynamically. We derive the measurement architecture based on the “crowd-sourcing” scheme to gather large-scale measurement data with inexpensive sensor nodes. The PU detection or identification scheme is key part of the REM construction then there are multiple PUs in the environment. However, it is difficult to identify multiple PUs in an individual user terminal. Therefore, the PU detection or identification problem is solved at the REM servers using the Hidden Markov Model (HMM), which is a time-series-based machine learning technique. The proposed HMM method classifies the measurement data depending on the combined state of each transmitter, which can be either active or idle. The results show that the proposed method exhibits better performance than the existing unsupervised clustering method.

The need for extra spectrum to support future 5G networks and the fact that a large amount of spectrum below 6 GHz is allocated for different radar systems have motivated regulatory bodies and wireless researchers to investigate the feasibility of Dynamic Spectrum Access (DSA) in radar bands. Moreover, the next generation of infrastructures and devices will need to be more flexible to support a much wider range of application services, operate over much wider bandwidths, and exploit multiple radio access technologies. Radio Environment Maps (REMs) can be utilized to enhance the awareness of network entities of their operational radio environment. In the context of potential spectrum sharing in radar bands, REMs can be used to determine spectrum usage and propagation patterns of different incumbent radar systems operating in a given area. This information in turn can be utilized to establish different radar systems protection requirements in a particular band, and can also help to mitigate any interference between incumbent radar systems and Secondary Users (SUs). In this paper, based on spectrum measurement campaigns in Finland and Ireland, we explore the functional architecture of REMs for potential spectrum sharing in radar bands. We argue that the unique operating principle and signal characteristics of this type of incumbent needs to be considered in implementation of the sensor network, REM communication protocol, and REM architecture. We also discuss the potential of REM to facilitate the use of more advanced interference cancellation or avoidance technologies such as temporal sharing that can significantly reduce exclusion zones.

Radio Environment Map (REM) is a map which indicates the radio signal strength (RSS) over a geographical region. With the help of REM, Cognitive Radio (CR) users can opportunistically access the licensed spectrum. Distributed cooperative REM estimation is vulnerable to malicious sensors that submits false sensing reports. In this paper, we develop a secure distributed scheme to estimate the REM in hierarchical wireless CR networks. We formulate the estimation process as a LS problem with two ii-norm constraints using the basis pursuit approach. Reputation factors are introduced to further improve the estimation accuracy. Our scheme enables joint valid estimation result and malicious sensor identification. The performance of the proposed scheme is confirmed by extensive simulation studies.

Electroencephalography (EEG)-based control has emerged as a promising solution for intuitive and non-invasive control of assistive robotic devices, particularly for individuals with severe limb loss or neuromuscular impairments where muscle signals are unavailable. This paper presents the design and development of a low-cost, lightweight 3D-printed prosthetic arm controlled exclusively by EEG signals. The proposed system decodes user movement intention directly from brain activity using signal processing, sensor fusion, and adaptive machine learning techniques. EEG signals are acquired through a non—invasive multi-channel headset, conditioned to suppress artifacts, and transformed into meaningful features representing motor imagery and movement intention. To improve reliability and reduce noise sensitivity inherent in EEG signals, multiple signal fusion and filtering techniques—including averaging, weighted averaging, Principal Component Analysis (PCA), and Kalman filtering—are investigated. Experimental results demonstrate that Kalman filter-based fusion significantly enhances signal stability and control accuracy in real-time prosthetic operation. The prosthetic arm employs a modular 3D-printed mechanical structure with servo-driven actuation, ensuring affordability and customization. Additionally, a future bidirectional feedback framework based on deep learning is proposed to enhance user adaptation and sensory perception. This work contributes toward accessible, intelligent, and fully brain-controlled prosthetic systems for medical rehabilitation and assistive healthcare.

In the IoT era, 5G will enable various IoT services such as broadband access everywhere, high user and devices mobility, and connectivity of massive number of devices. Radio environment map (REM) can be applied to improve the utilization of radio resources for the access control of IoT devices by allocating them reasonable wireless spectrum resources. However, the primary problem of constructing REM is how to collect the large scale of data. Mobile crowd sensing (MCS), leveraging the smart devices carried by ordinary people to collect information, is an effective solution for collecting the radio environment information for building the REM. In this paper, we build a REM collecting prototype system based on MCS to collect the data required by the radio environment information. However, limited by the budget of the platform, it is hard to recruit enough participants to join the sensing task to collect the radio environment information. This will make the radio environment information of the sensing area incomplete, which cannot describe the radio information accuracy. Considering that the Kriging algorithm has been widely used in geostatistics principle for spatial interpolation for Kriging giving the best unbiased estimate with minimized variance, we utilize the Kriging interpolation algorithm to infer complete radio environment information from collected sample radio environment information data. The interpolation performance is analyzed based on the collected sample radio environment information data. We demonstrate experiments to analyze the Kriging interpolation algorithm interpolation results and error and compared them with the nearest neighbor (NN) and the inverse distance weighting (IDW) interpolation algorithms. Experiment results show that the Kriging algorithm can be applied to infer radio environment information data based on the collected sample data and the Kriging interpolation has the least interpolation error.

The aim of this paper is to preserve location privacy of crowdsourced-based spectrum sensing agents using geo-indistinguishability. We considered database-driven dynamic spectrum access, where a radio environment map provides spectrum availability information for dynamic spectrum access management. Moreover, we assumed crowdsourced-based spectrum sensing, where a pool of allocated mobile users, called crowdsourced-based spectrum sensing agents, sense the spectrum and report their actual location and the received signal strength to the spectrum manager that constructs a radio environment map. This discloses location information of crowdsourced-based spectrum sensing agents and violates their location privacy. Consequently, crowdsourced-based spectrum sensing agents could be discouraged to participate in spectrum sensing. In our paper, to solve the problem of location disclosure, we adopted planar Laplacian mechanism, where each crowdsourced-based spectrum sensing agent reports an obfuscated location instead of its actual location, which achieves geo-indistinguishability. Our simulation results were based on real-world CRAWDAD dataset. Our results showed that with a moderate privacy level, location privacy of crowdsourced-based spectrum sensing agents was preserved while the effect of introduced location noise on the accuracy of radio environment map was insignificant.