In this work, we present a terrain-aware 3D multi-unmanned aerial vehicles (UAV) cooperative spectrum sensing (CSS) framework that jointly optimizes UAV placement and antenna orientation to maximize detection probability while minimizing hover energy. A bounding volume hierarchy (BVH)-based adaptive scheme is proposed for efficient line-of-sight (LoS) evaluation, and a hierarchical genetic algorithm (GA)–particle swarm optimization (PSO) approch is adopted to address the inherently non-convex bi-objective problem, achieving a balanced exploration of the search space while ensuring LoS connectivity and energy efficiency. Monte Carlo simulations using real terrain data demonstrate up to 37.02% improvement in detection probability and 48.90% reduction in hover energy compared to PSO-only baselines, confirming the practicality and effectiveness of the proposed framework for uneven terrain environments.

Spectrum sensing plays a major role for dynamic spectrum access in Cognitive Radio Networks as we move towards 6G networks characterized by high-frequency operation, ultralow latency requirements, and dense spectrum usage. Most state-of-the-art spectrum sensing techniques suffer from high computational complexity and degraded performance in low Signal-to-Noise Ratio (SNR) environments. To overcome these shortcomings, we propose a Fourier Analysis Network (FAN)-based spectrum sensing approach that specifically addresses the challenge of low detection probability in low-SNR conditions while significantly reducing computational cost. The proposed FAN-based framework integrates Fourier transform properties within the network layers, allowing it to better model periodic and frequency-localized characteristics of wireless signals, such as those found in 5G and 6G-based waveforms. We have evaluated the proposed approach on three datasets that include 5G-NR, DeepSig, and Radar. The experimental results show that the proposed method achieves an improvement in the detection probability of 32.88% and 91.43% compared to the baseline models KAN-SS and PU-DetNet, respectively, at SNR = -10dB. The proposed scheme also outperforms the baseline models in terms of F1-score, with a reduction of 50.60% in FLOPs.

With the rapid expansion of smart applications in 6G-enabled low-altitude economy networks, edge intelligence has emerged as an essential technology to reduce latency and energy consumption by dynamically deploying edge servers close to end users. Unmanned aerial vehicles (UAVs) can each carry a small-sized edge server to offer computation services to multiple end users. In this paper, we consider a non-orthogonal multiple access (NOMA)-based two-layer UAV swarm edge intelligence network, which includes task-generating UAVs (TUAVs) and an edge server UAV (SUAV). SUAV offers extra computation capacity to all TUAVs to process their computational tasks remotely. To minimise the long-term total service delay and the total energy consumption of all TUAVs based on the tasks’ transmission and processing, we propose a proximal policy optimisation (PPO)-based offloading decision and resource allocation joint optimisation algorithm by optimising offloading decisions, communication and computation resources allocation, and transmit power for all TUAVs. Simulations demonstrate that our approach outperforms benchmark solutions in terms of long-term total service delay and total energy consumption of all TUAVs in a UAV swarm system.

Ultra-Reliable and Low-Latency Communication (URLLC) represents a key ingredient of current 5G and future 6G mobile communication networks. The demanding reliability requirement set for URLLC claims for novel techniques that can deliver the necessary level of reliability without sacrificing the overall system capacity. In this context, this work presents and analyses a hybrid transmission scheme for improved reliability in millimetre wave bands with adaptive diversity combining. The proposed scheme is based on a hybrid approach that combines two links, one in the FR1 band (characterised by lower capacity but higher reliability due to more favourable propagation) and one in the FR2 band (offering higher capacity but experiencing a less reliable connectivity). The proposed scheme dynamically adapts the usage of both links in order to exploit the complementary characteristics of both bands (reliability of FR1 bands and capacity of FR2 bands) by switching between FR2-only and joint FR1-FR2 transmission according to the instantaneous channel quality in the main FR2 link. The performance is evaluated under two popular and well-known diversity combining techniques, namely Selection Combining (SC) and Maximal Ratio Combining (MRC). The obtained results demonstrate that the proposed scheme can achieve the same level of reliability as a continuous dual-link transmission scheme but with a much lower level of links usage and without sacrificing (and indeed enhancing) the capacity, thus making it a suitable candidate to deliver URLLC services in a resource-efficient manner.

Radio communication signals are often represented in many practical application scenarios as a spectrogram, which indicates the power observed at several discrete time instants and frequency points within a certain time interval and frequency band, respectively. The concept of Signal Area (SA) was recently introduced in the context of spectrum occupancy measurements as the rectangular region in the time-frequency domain where a signal is believed to be present. An accurate estimation of the original SA for each radio transmission contained in a spectrogram can provide valuable information in many practical application scenarios, such as autonomous spectrum-aware wireless communication systems. In this context, this work proposes new methods for an accurate Signal Area Estimation (SAE) based on the application of the Hough Transform (HT) combined with other techniques from the field of image processing. The performance of the proposed methods is evaluated by means of simulations and experiments. The obtained results show that they can achieve a high level of SAE accuracy. Moreover, an interesting and distinguishing feature of the proposed methods is their ability to not only improve the accuracy of the SAE but also to extract automatically the coordinates and dimensions of each SA detected in a radio spectrogram. This feature can be useful in the automatic processing of radio spectrograms, for example in the context of autonomous spectrum-aware wireless systems.

Millimeter wave (mmWave) frequencies have been selected to provide the high data rates and capacity targeted by the next generation (5G) of mobile communication networks. Compared to the sub 6-GHz bands that have been historically exploited by popular communication systems, mmWave bands suffer from adverse radio propagation degradations such as higher path loss and blockage, however they provide significantly larger bandwidths and lower levels of usage than sub 6-GHz bands. Mobile operators can employ Carrier Aggregation (CA) techniques to combine the newly acquired spectrum in mmWave bands with their already allocated spectrum in sub 6-GHz bands. The use of CA as a technique to combine chunks of spectrum in a transparent manner for higher layers of the protocol stack has been widely researched. In this paper, a different focus is considered by exploring the use of CA as a diversity technique in order to further improve the network capacity under various radio propagation scenarios. The obtained results indicate that CA can effectively be exploited as a diversity technique to optimize the overall system performance and increase the network capacity, with the optimum number of component carriers depending on the radio propagation distance.

Spectrum awareness is an essential aspect of wireless communication technology. Wireless communication systems can obtain spectrum awareness information by monitoring the spectrum usage in the frequency and time domains and representing this information as a time-frequency matrix. In many practical cases it is useful to determine the subsets of elements of such matrix where a signal is present (i.e., the signal area). Several signal area (SA) estimation methods with varying performance have been proposed in the literature. However, there is a lack of comparative research that shows how the configuration of such methods affects their relative performance. In this context, this work investigates the impact of two essential configuration aspects for any SA method, namely the threshold used to decide whether each element of the time/frequency grid contains a signal component or just noise, and the frequency/time resolution of the measurements carried out to obtain such data matrix. Several popular threshold decision criteria and a broad range of measurement resolutions are investigated, showing that these two particular aspects play a key role in the optimum configuration and performance of SA estimation methods. Several useful findings and design guidelines are provided as well.

Dynamic Spectrum Access (DSA)/Cognitive Radio (CR) has emerged as an effective paradigm to solve problem of inefficient spectrum utilization. Spectrum sensing is the key in DSA/CR. Spectrum sensing decisions can be utilized to accurately estimate the primary channel activity statistics (PAS) like mean of idle/busy period, Duty cycle etc. Such estimated statistics can be used by CR to improve performance. However, when the secondary user (SU) is mobile, estimating these statistics becomes challenging. Taking this into account, this work provides a thorough review on the estimation of PAS for mobile SUs, considering vehicular scenario. The random way-point based mobility model is adopted for modelling SU mobility. Specifically, this work provides a set of closed form expressions for the estimated statistics under SU mobility as a function of the true PAS, SU velocity, initial distance between PU and SU, PU's protection range (R), SU's sensing range (S).

With the rapid development of intelligent transportation systems, vehicular devices are getting connected with each other. However, this leads to the problem of spectrum scarcity. Dynamic spectrum access (DSA)/cognitive radio (CR) has emerged as an effective solution to solve the problem of inefficient spectrum utilization. Spectrum sensing is the key in DSA/CR system. In cognitive vehicular networks (CVNs), spectrum sensing becomes more complex and challenging and that often leads to a loss in performance detection. Due to the effect of channel fading/shadowing and due to secondary user (SU) mobility, individual SUs may not be able to detect the existence of primary user (PU). In this paper, we propose a weighted cooperative spectrum sensing (weighted-CSS) framework for accurate detection of PU in CVNs. The weights are calculated from the probability of PU being inside the SU's sensing range and SU being outside the PU's protection range (inside probability). The calculated weight for SU indicates the reliability in the signal received by SU. The framework contains two stages. In the first stage, inside probability is calculated at each SU and the inside probability and the energy signal received from PU are sent to a base station (BS). In the second stage, BS assigns a weight to each SU based on the inside probability and makes a decision by combining the information received from SUs. Numerical results indicate that, on an average, the proposed framework performs ≈15% better than the conventional local spectrum sensing.

Primary channel activity statistics play an important role in improving the performance of Dynamic Spectrum Access (DSA) / Cognitive Radio (CR) systems. The statistical information of the idle/busy periods of a primary channel can be estimated based on the outcomes of spectrum sensing. Recent studies have shown that these statistics can be estimated accurately even under Imperfect Spectrum Sensing (ISS) scenarios. Those studies, however, have assumed no constraints on the required sample size of observations of the idle/busy periods in order to provide accurate estimation (i.e., large sample size was assumed to test the accuracy of these statistics estimation methods). In real-world scenario, DSA/CR systems are limited to the hardware design capabilities, which include limited memory capacity, energy consumption and computational capability. As a result, it is very important to find how many samples of the idle/busy periods are required to provide an acceptable level of accuracy for the estimated statistics. Therefore, this work analyses the impact of the sample size on the estimation of the primary channel statistics under ISS and it finds closed-form expressions for the required sample size of the idle/busy periods to achieve a targeted accuracy. In addition, the analytical results achieved in this work are validated by means of simulations and hardware experiments.

A pro-active spectrum usage prediction is a key technique in decision making and spectrum selection for dynamic spectrum access systems. This work focuses on the estimation of the duty cycle (DC) metric to reflect spectrum usage. The prediction is formulated as a time-series regression problem. Deep neural networks (DNNs) is selected to obtain accurate predictions of channel usage. Namely, Multilayer perceptron (MLP), Long short term memory (LSTM) and a hybrid model based on convolutional neural network followed by an LSTM (CNN-LSTM) layer are selected. The hyper-parameters selection has been optimised utilising both grid search and multi-stage grid search. Moreover, in many cases, the spectrum usage is measured on a smaller time scale from the actual required one. Hence, down-sampling and averaging is required. Averaging operation results in flattening the data and losing essential features to assist DNN to predict the channel usage. We show what is the minimum required time resolution to have a pro-active prediction system. Then, we propose utilising feature engineering to improve prediction accuracy. All the proposed DNNs approaches are trained on real-life measurements. The experimental evaluation demonstrated a high potential of DNNs to learn from previous spectrum usage and accurately predict the spectrum usage. Moreover, adding input features significantly assists the system to achieve accurate predictions in a pro-active manner.

Cognitive Radio (CR) is aimed at increasing the efficiency of spectrum utilisation by allowing unlicensed users to opportunistically access licensed spectrum bands during the inactivity periods of the licensed users. CR systems can benefit from an accurate knowledge of the spectrum occupancy patterns and their statistical properties. This statistical information can be obtained by periodically monitoring (sensing) the idle/busy state of the licensed channels. However, a reliable estimation of the primary activity statistics may require long observations times. This work proposes efficient methods to reduce the observation time required to produce a reliable estimation of the primary activity statistics. Furthermore, a method enabling CR users to quantify the accuracy of the estimated statistics is also proposed. Compared to other existing approaches, the proposed methods can provide accurate estimations of the primary activity statistics in significantly shorter observation times, thus allowing CR users to quickly adapt to new unknown operating channels.

Statistical information of primary channels has received considerable research interest in the recent years. This is due to the important role that these statistics play in improving the performance of Dynamic Spectrum Access (DSA)/Cognitive Radio (CR) systems. Although a DSA/CR system has no initial knowledge about the statistical information of the primary channels, these statistics can be estimated from the observations of spectrum sensing. However, spectrum sensing is not perfect in the real world and sensing errors are likely to occur during DSA/CR operation, which in turn leads to incorrect estimation of primary channel statistics as well. As a result, several attempts have arisen to reconstruct the estimated periods of the primary channel occupancy patterns which are affected by the sensing errors, in order to provide more accurate estimation for the statistical information. However, all the reconstruction methods available in the literature assume the perfect knowledge of the primary users' minimum occupancy time. In this context, this work proposes the first reconstruction method that does not require any prior knowledge about the primary channel activity and inactivity patterns while achieving almost the same performance achieved by the latest reconstruction methods available in the literature, making it significantly attractive and feasible in practical implementation scenarios.

Cooperative spectrum sensing (CSS) is a technique where multiple cognitive radio users cooperate among themselves to make binary decisions about the presence of a primary user. The single cognitive user often faces the hidden terminal problem. However, CSS tackles this problem by sending local sensing‐based decisions to the fusion centre. A major drawback of conventional energy detection is the poor performance at low SNR regime. In this work, likelihood ratio statistics is considered as a test‐statistic due to its highest statistical power. An improved likelihood ratio statistic‐based CSS scheme is proposed by considering several past sensing events. The proposed scheme mitigates the poor detection at low SNR regime and misdetections arising due to sudden drops in signal energy. Furthermore, the generalised Byzantine attack is taken into account considering a security aspect. The proposed scheme is also shown to outperform Anderson Darling‐based malicious user detection in CSS at a low SNR regime. The proposed scheme is verified and validated over empirical spectrum data. The performance improvement is at the cost of computational time, which in practice is very low and is justified by the significant performance improvements of the proposed scheme at low SNR regime.

The majority of practical studies and analyses in the context of the Internet of Things (IoT) have been carried out assuming that data packet generation follows theoretical models (typically a Poisson process with exponentially distributed packet interarrival times) without previous experimental validation and supporting evidence. In contrast to this approach, this article proposes a novel experimental and mathematical framework to determine statistical models for IoT data traffic. Based on empirical data generated by common smart home devices (e.g., ambient temperature, luminous intensity, atmospheric pressure, and motion sensors) recorded over a full year using an experimental IoT subsystem, this article first shows that real IoT traffic does not follow the Poisson process model conventionally assumed in the literature, but rather depends on the type of application. Consequently, we estimate the empirical statistical distribution of the interarrival between data packets for several smart home applications. The empirical distribution of the packet interarrival times is fitted with some well-established classical statistical distributions using the method of moments as well as maximum-likelihood estimation techniques, and the goodness of fit is quantified using the Kolmogorov-Smirnov (KS) test. Moreover, we also carry out a regression analysis to provide mathematical relations between the distribution parameters and the considered physical input parameters (ambient temperature, luminous intensity, and atmospheric pressure), which is particularly useful in practical scenarios. Furthermore, an exhaustive analysis of the variation of parameters over different time scales and the autocorrelation characteristics of the data packet generation are included as well. In summary, this article provides accurate traffic models suitable for real-life IoT scenarios that can be used for an adequate design and optimization of future communication networks to efficiently support IoT services.

Primary channel statistics have recently gained increasing attention due to its remarkable role in the performance improvement of Dynamic Spectrum Access (DSA)/Cognitive Radio (CR) systems. These statistics can be calculated from the outcomes of spectrum sensing, which is the well-known method used to identify the available instantaneous opportunities in the spectrum. Computing statistical information from spectrum sensing, however, may sometimes be unreliable due to the fact that spectrum sensing is imperfect in the real world and errors are likely to occur in the sensing decisions. In this context, this work provides a detailed analysis of a broad range of primary channel statistics under Imperfect Spectrum Sensing (ISS) and finds a set of closed-form expressions for the calculated statistics under ISS as a function of the original primary channel statistics, probability of error, and the employed sensing period. In addition, the obtained mathematical expressions are employed to find and propose novel estimators for the primary channel statistics, which outperform the existing estimators in the literature and can provide accurate estimations of the original statistics even under high probability of error of spectrum sensing. The correctness of the obtained analytical expressions and the accuracy of the proposed estimators are corroborated with both simulation and experimental results.

This work proposes a novel experimental and mathematical framework to determine the statistical models for the Internet of Things (IoT) data traffic. Conventionally, it is assumed that the data packet generation for IoT based applications follows a Poisson process with exponentially distributed packet inter-arrival time. Based on such generalized premise, majority of the network related theoretical and practical analysis of the IoT platforms are carried out. Based on empirical data for a smart home application recorded for over 10 weeks duration using proposed IoT subsystem, in this paper we estimate the empirical statistical distribution of the IoT data traffic generated by temperature, light intensity and motion sensors. The inter-arrival between the data packets generated from different sensing modules of the IoT smart home application subsystem is determined. The Empirical Cumulative Distribution Function (ECDF) of the estimated time duration is fitted with few of the well-established classical statistical distributions using Method of Moments (MoM) and Maximum Likelihood (ML) estimation techniques. The goodness of fit is quantified using Kolmogrov-Smirnov (KS) test. The parameters of the fitted distributions are determined as a function of the physical input parameters. The results reveal source IoT traffic does not follow a Poisson process which is conventionally assumed in the literature, but rather depends on the type of application.

In this paper, we propose a framework to optimize the packet length and modulation level to determine the optimal packet size (OPS) for topology aware cognitive radio sensor networks (CRSNs) using a variable rate modulation scheme. A generalized network topology with specific node density of the Primary Users (PUs) is accounted to estimate the OPS. Based on stochastic geometry and non-linear optimization techniques, a joint multivariate optimization problem is formulated to determine the OPS for the topology dependent CRSNs.

The emerged Dynamic Spectrum Access (DSA) concept based on Cognitive Radio (CR) is a promising solution to overcome the problems related to frequency spectrum scarcity. In DSA/CR systems, the inactivity patterns of the licensed frequency channels are exploited in an opportunistic and non-interfering manner by unlicensed users. Therefore, the knowledge of the occupancy rate (i.e., duty cycle) of these licensed channels is crucial for boosting the performance of the DSA/CR system. For example, it can help to select the lowest occupied channel which can offer higher opportunistic spectrum to the unlicensed users. Channel Duty Cycle (DC) is a statistical parameter about the activity of the licensed channel in time- domain, which is initially unknown to the DSA/CR system but can be estimated from the outcomes of spectrum sensing. However, spectrum sensing is imperfect in practice due to sensing errors, which in turn will provide incorrect estimation of the channel DC. In this context, this work successfully finds a novel method to accurately estimate the channel DC even under Imperfect Spectrum Sensing (ISS) without requiring any prior knowledge about the licensed channel activity. This is achieved after accurately analysing the impact of ISS on the estimation of the statistical moment (mean) of the channel activity periods, for which a closed form expression is obtained as a function of the true mean, probability of errors and sensing period. The achieved mathematical expression helps to find a novel method to accurately estimate the true mean of the channel activity periods and subsequently the channel DC based on the outcomes of the ISS.

This work proposes a novel mathematical approach to accurately model data traffic for the Internet of Things (IoT). Most of the conventional results on statistical data traffic models for IoT are based on the underlying assumption that the data generation follows standard Poisson or Exponential distribution which lacks experimental validation. However, in some of the use case applications a single statistical distribution is not adequate to provide the best fit for the inter-arrival time of the data packets generation. Based on the real data collected for over 10 weeks using our customized experimental IoT prototype for smart home application, in this paper we have established this very fact, citing barometric air pressure as an example. The statistical distribution of the inter-arrival time between the data packets for a specified barometric pressure fluctuation threshold is initially determined by approximating the best-fit with a set of standard classical distributions. The goodness-of-fit with the empirical data is numerically quantified using Kolmogorov-Smirnov (KS) Test. Furthermore, it is observed that any single standard distribution is unable to provide a good fit which is at least less than 10%. Therefore, a novel weighted distribution scheme is proposed that could provide an acceptable fit. The weighing factor including the location, scaling and weighing parameters of the best fitting distribution are estimated and analyzed. The distribution parameters are finally expressed as a function of the differential pressure value that can be used for different theoretical analysis and network optimization.

The application of machine learning models to spectrum sensing in cognitive radio is not uncommon in literature, but most of these models fail to consider temporal dependencies in the signal. In this paper, the temporal correlation among the spectrum data is exploited using a Long Short-Term Memory (LSTM) network. More specifically, the previous sensing event is fed along with the present sensing event to the LSTM model. The proposed sensing scheme is validated based on empirical data of various radio technologies. The proposed LSTM model is compared with other machine learning algorithms in terms of classification accuracy. Furthermore, the proposed scheme is also compared with other spectrum sensing techniques. Results indicate that the proposed scheme improves the detection performance and classification accuracy at low signal-to-noise ratio regimes. Moreover, it is observed that the achieved improvement is obtained at the expense of longer training time and nominal increase in execution time.

An accurate estimation of the noise floor is of paramount importance for an optimum performance of spectrum sensing in Cognitive Radio (CR). The most common approach followed by existing noise floor estimation methods is to attempt to isolate a set of noise-only samples based on a given energy/power threshold. However, this approach is unreliable and in general unable to provide accurate estimations of the noise floor, in particular under low SNR conditions where the power of the Primary User (PU) signal is comparable to the noise floor of the CR device. In this context, this work considers a different approach where the power observed by the CR device is modelled as a Gaussian mixture. Based on a mathematical analysis of the relation among the parameters of the obtained Gaussian mixture, a modified version of the well-known Expectation Maximisation (EM) algorithm is proposed to fit the Gaussian mixture to the observed power values and provide an estimation of the noise floor, something that the general EM algorithm fails to achieve in this scenario. The obtained results demonstrate that the proposed method provides a highly accurate estimation of the noise floor in the presence of PU signals over the whole range of SNR values.

Most of the existing approximations for the Gaussian Q-function have been developed bearing in mind applications that require high estimation accuracy for large argument values (e.g., derivation of the bit/symbol error rates of digital communication systems, which are typically in the order of 10-6to 10-12). Such values correspond to positive arguments of the function and consequently most of the existing approximations are valid for positive arguments only. However, other relevant problems where the Gaussian Q-function can appear do not require such a level of accuracy (e.g., derivation of the detection probability of a signal detector, where accuracies of two or three decimal figures are sufficient) and, more importantly, require the evaluation of the Q-function over the whole range of values (i.e., both positive and negative arguments). In this context, this paper analyses a sigmoid approximation to the Q-function that provides adequate levels of accuracy for any real argument. As an illustrative example, this approximation is employed to obtain new closed-form expressions for the probability of detection of an energy detector under Rayleigh and Nakagami- m fading channels.

In order to cope with the increased demand of wireless services and applications, LTE over unlicensed spectrum has been proposed to extend the operation of LTE to operate also over unlicensed bands. However, this extension faces various challenges regarding the coexistence between LTE-U and different technologies that use these unlicensed spectrum bands such as Wi-Fi technology. The first scenario of LTE allowing LTE to operate over unlicensed bands is LTE-Unlicensed duty cycling (LTE-U). Specifically, LTE-U can coexist with Wi-Fi by allowing LTE-U devices to transmit only in predetermined duty cycles (DCs) or to use adaptive DCs for LTE based on the activity measurements. In this paper, we investigate the downlink performance of LTE-U and Wi-Fi under different traffic loads. The main novelty of this work is to exploit the knowledge of the existing Wi-Fi traffic activity to select a fixed DC for LTE. Moreover, two proposed methods to allocate the blank subframes within LTE frames are provided. Simulation results using NS-3 simulator for LTE-U and Wi-Fi coexistence mechanism under different traffic loads are provided. In particular, the results show that the coexistence mechanism between LTE-U and Wi-Fi in the 5 GHz band achieves better total aggregated throughputs for the coexisting technologies using the proposed approach. Moreover, the location of the blank subframes plays a key role in this coexistence in terms of the total aggreaated throughputs.

The 3rd Generation Partnership Project (3GPP) has recently defined a Licensed Assisted Access (LAA) scheme to enable Long Term Evolution (LTE) networks to use unlicensed frequency bands. However, the unlicensed bands are mainly occupied by the Wi-Fi technology. Hence, achieving fairness between LAA using LTE (LTE-LAA) and Wi-Fi in the unlicensed bands is a primary challenge. The 3GPP has recently standardised in Release 13 a Listen Before Talk (LBT) algorithm to ensure the fairness among these two technologies (LTE and Wi-Fi) over the unlicensed bands. In this paper, we focus on the downlink performance of LTE-LAA and Wi-Fi with different traffic loads. To achieve not only better fairness but also higher total aggregated throughputs for the coexisting networks, a static Contention Window (CW) selection method based on the fairness definition is proposed. The main novelty of this work is that the knowledge of Wi-Fi activity statistics is exploited effectively to select the CW of LAA. We show that the fairness between LAA and Wi-Fi networks depends on the LAA CW size adaptation criterion. Simulation results validate that the proposed method is effective in LAA/Wi-Fi coexistence scenario, can improve fairness performance and provide higher total aggregated throughputs for both coexisting networks compared with the current Category 4 LBT (Cat 4 LBT) algorithm defined in the 3GPP standard.

Primary channel activity statistics (such as the minimum idle/busy periods of the channel, the moments or the underlying distributions) can be exploited by cognitive radio (CR) systems to adapt their operation and improve their performance. Such statistics can be directly estimated from periodic observations of the instantaneous idle/busy state of the primary channel (i.e., periodic spectrum sensing). However, the periodicity of such observations (i.e., the sensing period) imposes a fundamental limit on the time resolution in which idle/busy periods can be observed, and consequently on the accuracy of any subsequent estimated statistics. In this context, this paper provides a comprehensive analysis on the estimation of the primary activity statistics based on periodic channel state observations performed with a finite sensing period. In particular, this paper provides a comprehensive set of closed-form expressions for the estimated statistics as a function of the true primary activity statistics and the employed sensing period. These expressions can find a wide range of applications in the analysis, design, and simulation of CR systems. Moreover, several methods to minimize the estimation errors and improve the accuracy are proposed and validated with both simulations and hardware experiments.

Recent advancements in Full Duplex (FD) radio have built a promising idea for faster channel sensing in cognitive radio. Full duplex cognitive radio (FDCR) provides an efficient way to utilize the idle channel without interrupting the ongoing transmission. Currently, non-parametric sensing techniques like energy detection and its modified techniques are implemented for FDCR. In this paper, we introduce a Goodness of Fit based distribution-free sensing in FDCR. With Monte Carlo simulations and analytical approximation, we show that the proposed technique outperforms energy detection and other goodness-of-fit based sensing algorithms for FDCR.

An accurate estimation of the primary statistics is essential for Cognitive Radio (CR) systems. This knowledge can be exploited to enhance CR performance and reduce the interference with the primary users. In this work, we propose a method based on the Method of Moments (MoM) to improve the distribution estimation. A Modified Method of Moments (MMoM) with a correction factor is proposed to improve the estimation of moments and thus the resulting primary distribution. The simulation and experimental results show that the MMoM approach is notably more accurate. Finally, we study the importance of having a sufficiently large sample space and the effect of sample size on the moments and the primary distribution estimation.

Dynamic Spectrum Access (DSA) / Cognitive Radio (CR) systems utilize spectrum sensing to monitor spectrum status and decide transmission time in an opportunistic manner. This results in an increase in wireless spectrum efficiency. Spectrum sensing can also be used to monitor the statistics of primary users to gain information on occupation patterns and estimate the statistics of the primary traffic activity, a useful knowledge that can be exploited in many ways. In this research, three novel algorithms are proposed to enhance the estimation of primary user activity statistics under imperfect spectrum sensing given the knowledge of minimum transmission time. Simulation results show that the proposed methods enable accurate estimation for the primary user statistics. Moreover, the proposed methods are compared to previously proposed methods and it is shown they provide a significantly better estimation accuracy.

Duty cycle (DC) has been used to express the deterministic and stochastic aspects of spectrum usage. Specifically, a deterministic model for the mean of the duty cycle (MDC) has been proposed in a previous work. On the other hand, the observed DC (O-DC) during short time duration has randomness and a stochastic model is used to express the randomness. In this paper, we extend the conventional approach to a combined deterministic-stochastic (DS) model which represents both the deterministic and stochastic aspects. For the distribution of the O-DC, the beta distribution has been used as stochastic model, but we employ a mixture of beta distributions. The mixture-beta distribution can achieve higher accuracy but requires more capacity for data storage in spectrum usage measurements since it has a higher number of parameters than the beta distribution. For this issue, we employ regression analysis in DS-model to reduce the number of parameters while retaining the accuracy. We show the validity of DS-model based on exhaustive spectrum measurements in IEEE 802.11-based wireless local area networks.

For achieving an efficient spectrum sharing in a context of dynamic spectrum access, understanding the spectrum usage by licensed users [primary users (PUs)], is important for secondary users (SUs). Duty cycle (DC) has been used to express the deterministic and stochastic aspects of spectrum usage. Specifically, a deterministic model for the mean of the duty cycle (M-DC) has been proposed in a previous work. The deterministic aspect of M-DC is affected by social behavior, and common habits of users, which can be confirmed in cellular systems. On the other hand, the observed DC (O-DC) during short time duration has randomness and a stochastic model is more suitable, e.g. distribution of O-DC. In this paper, we extend the conventional approach, in which only either the deterministic or stochastic aspect is considered, to a combined deterministic-stochastic (DS) model, which represents both the deterministic and stochastic aspects at once. For the distribution of the O-DC, the beta distribution has been used as stochastic model, but we employ a mixture of beta distributions. The mixture-beta distribution can achieve higher accuracy but requires more capacity for data storage in spectrum usage measurements since it has a higher number of parameters than the beta distribution. For this issue, we employ regression analysis in DS-model since this approach can reduce the number of parameters while retaining the accuracy. We show the validity of DS-model based on exhaustive spectrum measurements in IEEE 802.11-based wireless local area networks and long-term evolution uplink.

Cognitive Radio (CR) systems utilize spectrum sensing to decide transmission time in an opportunistic manner. Spectrum sensing can also be used not only to determine the instantaneous on/off state of the channel but also to monitor the statistics of primary user to gain information on occupancy pattern. This knowledge can be exploited in many ways to improve CR systems. In this paper, we propose an analytical model to link the sensing period with the observed spectrum occupancy. Moreover, the effect of spectrum sensing periods on the estimated primary activity pattern is analysed. Simulation results show that the proposed model captures with reasonable accuracy the spectrum occupancy observed at the CR.

Spectrum sensing is a key aspect of Cognitive Radio (CR). The main requirement in CR systems is the ability to sense the primary signal accurately and rapidly. In this paper, a novel hybrid spectrum sensing scheme in CR is proposed which considers the hypothesis problem as a binary classification problem. The proposed scheme is a combination of classical energy detection, Likelihood Ratio Test statistic (LRS-G2) and Artificial Neural Network (ANN). The scheme utilises energy from energy detection and Zhang test statistic from LRS-G2as features to train the ANN while ANN provides the adaptive learning and stable performance to the scheme. The performance of proposed sensing scheme is evaluated on several real-world primary signals of various radio technologies and it has been found out that for all those radio technologies the proposed scheme outperforms the classical energy detection and the improved energy detection.

Dynamic spectrum access/cognitive radio (DSA/CR) systems can benefit from an accurate knowledge of the primary channel ON/OFF activity statistics. This information can be readily obtained from spectrum sensing observations. However, given the practical limitations of spectrum sensing, an accurate and realistic evaluation of estimation methods is required before they can be implemented in real systems. This paper presents the design and implementation of a low-cost Prototype for the Estimation of Channel Activity Statistics (PECAS) in DSA/CR systems. An in-depth description of the hardware design (based on common and inexpensive components) and software implementation (based on free open source code) is provided along with an illustrative applicability example. The developed platform provides researchers and engineers with a low-cost fully functional tool for proof-of- concept, validation and optimisation of algorithms and designs.

Modern wireless communication systems adapt the employed Modulation and Coding Scheme (MCS) to the instantaneous channel quality in order to maximise the amount of bits correctly transmitted per time unit (i.e., the throughput). An important aspect in the study of this technique, referred to as Adaptive Modulation and Coding (AMC), is the characterisation of the link throughput performance under realistic fading channels. In this context, this work derives analytically tractable closed-form expressions for the link throughput performance of AMC under Rayleigh, Nakagami-m, Nakagami-q (Hoyt), Nakagami-n (Rice), η-μ and κ-μ fading channels. The correctness and accuracy of the obtained expressions is corroborated numerically and with results from an LTE link level AMC simulator.

Dynamic Spectrum Access (DSA)/Cognitive Radio (CR) systems can benefit from the knowledge of the activity statistics of primary channels. A particularly relevant statistic is the Channel Occupancy Rate (COR) of a primary channel, which represents the probability that a channel is occupied by a primary user. The COR can be estimated based on a set of binary (idle/busy) spectrum sensing decisions. However, an important practical question is how many sensing observations are necessary (i.e., the sensing sample size) in order to estimate the COR of a primary channel to a certain desired level of accuracy. This work analyses the problem of estimating the COR of a primary channel based on spectrum sensing decisions and derives a tight closed-form expression for the required sensing sample size. Moreover, an iterative algorithm is proposed to accurately determine the sensing sample size required to estimate the primary COR to a desired level of accuracy. The obtained results demonstrate the ability of the proposed algorithm to make an arbitrarily accurate estimation of the real COR of unknown primary channels using the essential minimum number of sensing samples.

Dynamic Spectrum Access (DSA)/Cognitive Radio (CR) systems can detect transmission opportunities by means of periodic spectrum sensing. The design and configuration of spectrum sensing is commonly aimed at optimising the instantaneous detection of such opportunities. Besides the detection of transmission opportunities, spectrum sensing can also be exploited to provide DSA/CR systems with more sophisticated and elaborated information, including for instance statistical information on the occupancy pattern of primary channels. However, the configuration of spectrum sensing in order to minimise the estimation error of channel activity statistics has received much less attention. In this context, this work explores the configuration of an energy detector in order to enable an accurate estimation of the real occupancy rate of a primary channel, thus providing DSA/CR systems with accurate statistical information of primary channels that can be used effectively in spectrum and radio resource management decisions.

Channel occupancy rate (COR) is the fraction of the time that a channel is occupied, i.e., contains signal(s) in addition to noise. Estimation of COR is important, e.g., in cognitive radio systems, which can use this information for intelligently adapting their spectrum use to the operating environment. For COR estimation, both the ability to operate with weak signals (sensitivity) and closeness of the estimate to the true COR value (accuracy) are important. In this paper, an improved COR estimation (iCOR) method is proposed enabling the use of high false alarm probabilities to improve sensitivity without the overestimation usually associated with high false alarm probabilities. The iCOR method is compared with the conventional method in terms of worst-case root-mean-square error (RMSE), which refers to the RMSE for the COR level yielding the maximum RMSE. To fairly compare different COR estimation methods, it is required that the RMSE for strong signals equals a target value and the considered methods are compared by their RMSE for weaker signals. Comprehensive theoretical analysis is performed and both exact results and approximations are derived. Experimental results verify the theoretical analysis and show significant sensitivity gains from the iCOR method (around 5 dB).

Owing to the opportunistic nature of the Dynamic Spectrum Access (DSA) paradigm based on the Cognitive Radio (CR) concept, the behaviour and performance of DSA/CR systems depend on the spectrum occupancy pattern of the primary system. DSA/CR systems can monitor periodically the occupancy state of licensed channels in order to gain statistical information on their occupancy patterns, and exploit this information in decision-making processes. Based on the outcomes of periodic spectrum sensing decisions, DSA/CR systems can make an estimation of the duration of the channel idle/busy periods and compute relevant statistics such as the minimum period duration, the moments (e.g., mean and variance) and the underlying distribution. However, the imperfect performance of spectrum sensing methods may lead to inaccuracies on the estimated statistics. In this context, this work carries out a detailed simulation-based study on the impact of imperfect spectrum sensing performance on the estimation of the primary activity statistics. The relative impact of various parameters is analysed and quantified, and guidelines to properly configure the spectrum sensing function of DSA/CR systems are provided. Moreover, several methods to overcome sensing errors are proposed and evaluated.

The inability to perfectly know the system noise properties to infinite precision, referred to as noise uncertainty, results in noise power calibration errors that have been proven to impose fundamental limitations on the detection performance of any spectrum sensing (signal detection) method in cognitive radio networks. In this work we argue that the inability of cognitive radio users to perfectly know beforehand the primary signals that might be present in the sensed band and their properties, referred to as signal uncertainty in this work, also results in an additional detection performance degradation. The noise uncertainty consequences have widely been studied, verified experimentally and distilled into tractable mathematical models. However, the potential effects of the particular primary signal properties on the resulting detection probability of generic spectrum sensing algorithms, such as energy detection, have not been taken into account in the analysis and performance evaluation of spectrum sensing in cognitive radio networks. In this context, this work develops a mathematical model for signal uncertainty and, based on such model, analyzes the impact of signal uncertainty on the resulting detection performance of spectrum sensing, with and without noise uncertainty, and compares the practical consequences of both degrading effects.

Energy detection constitutes a preferred approach for spectrum sensing in cognitive radio owing to its simplicity and applicability (it works irrespective of the signal format to be detected) as well as its low computational and implementation costs. The main drawback, however, is its well-known detection performance limitations. Various alternative detection methods have been shown to outperform energy detection, but at the expense of increased complexity and confined field of applicability. In this context, this work proposes and evaluates an improved version of the energy detection algorithm that is able to outperform the classical energy detection scheme while preserving a similar level of algorithm complexity as well as its general applicability regardless of the particular signal format or structure to be detected. The performance improvement is evaluated analytically and corroborated with the experimental results.

Cognitive radio nodes of a dynamic spectrum access network monitor the spectrum around them in order to detect and opportunistically exploit temporarily unoccupied spectrum gaps. The overall behavior of the whole network depends on the spectrum occupancy perceived by each node at its local environment. An appropriate understanding of the local perception of cognitive radio nodes in various scenarios and locations as well as its dependence on the propagation environment can provide valuable information for the design, dimensioning and performance evaluation of dynamic spectrum access networks. In this context, this work presents a novel and practical approach to predict the spectrum occupancy that would be perceived by cognitive radio nodes in realistic scenarios. Based on such approach, we explore the dependence of the spectrum occupancy perception with the considered scenario and location as well as the impact of particular parameters of the radio environment.

The existing approximations for the Gaussian Q-function have been developed bearing in mind applications that require high estimation accuracies (e.g., derivation of the error probability for digital modulation schemes). Unfortunately, the associated mathematical expressions are too complex to be easily employed in many other analytical studies, even if they do not require such a high accuracy. This letter proposes a simple, yet accurate mathematical approximation to the Gaussian Q-function. When compared to other existing approximations, the proposed model provides an adequate balance between accuracy and analytical tractability. Its simplicity enables its application over a wider range of analytical studies at reasonable accuracy levels. As an illustrative example, the proposed approximation is employed to obtain a new and simple closed-form expression for the probability of detection of an energy detector under Rayleigh fading channels.

Heterogeneous wireless systems are envisaged as the integration and joint cooperative management of diverse radio access networks and technologies through which network providers can satisfy the wide variety of user/service demands in a more efficient manner by exploiting their varying characteristics and properties. To achieve this objective, a key tool is common radio resource management technique designed to jointly manage the radio resources from different radio access technologies. In this context, this work proposes and optimizes new common radio resource management techniques designed to efficiently distribute traffic among the available radio access technologies while providing adequate quality of service levels under heterogeneous traffic scenarios. The obtained results demonstrate the ability of the proposed solutions to provide high user/service satisfaction levels while adequately exploiting the overall system resources.

Spectrum occupancy modeling in the context of Dynamic Spectrum Access/Cognitive Radio (DSA/CR) constitutes a rather unexplored research area that still requires much more effort. This paper addresses the problem of modeling spectrum occupancy in the spatial domain by proposing a novel theoretical approach that enables modeling the occupancy level perceived at any geographical location based on the knowledge of some simple primary signal parameters. The validity of the theoretical model is verified with extensive empirical measurement results. Some examples of its potential applicability are discussed as well.

Cognitive radio (CR) has been identified as a promising solution to the so-called spectrum scarcity problem. The basic idea of this paradigm is to allow unlicensed users to access in an opportunistic and non-interfering manner some licensed bands that are temporarily unoccupied by licensed users. CR is expected to significantly increase current spectrum usage. However, before this paradigm can turn into reality, a full understanding of the dynamic usage of spectrum is required. Current spectrum utilization has already been evaluated in some measurement campaigns. However, most of them were performed in USA and therefore evaluated the American spectrum regulation and utilization. Few studies have been carried out in other places. In this context, this paper presents spectrum occupancy measurements conducted in the frequency range from 75 MHz to 3 GHz in an outdoor environment in urban Barcelona, Spain. The measurements are analyzed and compared to the official spectrum regulations. The obtained results demonstrate the existence of a significant amount of spectrum available for the deployment of future CR networks.

This paper presents and describes the real-time testbed for all-IP Beyond 3G heterogeneous wireless networks that has been developed in the framework of the European IST AROMA project. The main objective of the AROMA testbed is to provide an advanced and realistic framework where the benefits of the algorithms developed within the AROMA project for both the radio access network and core network parts, as well as for the management of the end-to-edge quality of service, can be demonstrated. In particular, this paper focuses on the radio access part of the testbed, providing an in-depth description of its implementation and showing performance results with numerous supporting data that demonstrate the potentials and capabilities of the developed tool for the real-time evaluation of Radio Access Technology (RAT) selection policies under realistic scenarios. To this end, the behavior of two innovative RAT selection algorithms recently proposed in the literature has been assessed, analyzed, and compared in order to demonstrate the applicability of the tool.

Oriented to heterogeneity in every sense, next generation wireless networks are aimed to fulfil constantly increasing users' prospects. Relying on the all-IP paradigm, expectations suppose ubiquitous wireless access in heterogeneous networks. When efforts are made to make such an aspiration a realit

The paper presents and evaluates a link adaptation algorithm for real-time video transmission within a mobile radio environment. The aim of the algorithm is to improve performance, not only in terms of throughput, but also in terms of delay. The results show that the objectives set by the proposed algorithm are more effectively fulfilled compared to more traditional algorithms aimed at system throughput optimisation.