The fields of human presence detection (HPD) and people counting (PC) from WiFi-Channel State Information (CSI) represent constantly evolving areas of research within the scientific community. This study aims to expand the research frontiers by exploring the application of HPD and PC in an industrial environment characterized by its considerable dimensions, the presence of obstacles and machines, and harsh propagation channel conditions. Using a convolutional neural network (CNN) incorporating CSI measurements, we demonstrated the ability to distinguish between human presence and industrial machine activity with the same model, achieving an accuracy of over 99%. In addition, a study was conducted to count up to three people, even when machines were on, with an accuracy of over 92%. Another experiment was conducted with the objective of assessing not only the feasibility of PC in a naturally challenging environment but also the impact of the WiFi channel bandwidth in this particular application. The results of this study highlight the potential of WiFi-CSI sensing to address a significant challenge in human activity detection and estimation: its application in industrial environments.

Engines and electrical devices in operation generate electromagnetic pulses, also called impulsive noise (IN), that interfere with wireless signals. The IN shall affect the channel estimation process and is, therefore, present in the channel state information (CSI) provided by wireless receivers. In this paper, impulsive noise (IN) is used as a fingerprint of electrical devices to identify the IN sources that interfere with a WiFi signal, taking into account that each individual machine has a unique pattern of impulsive noise. In this sense, the WiFi CSI provides valuable information to recognize the IN sources through deep learning (DL) strategies. Two DL models have been proposed and tested on two experimental data sets for multiclass and multilabel analysis; in multiclass, devices can operate alone during the measurement, and in multilabel, multiple devices can work simultaneously in a more realistic scenario. The model transferability between location and days has also been evaluated by analyzing two different IN feature sets for device classification with the Few-shot-learning (FSL) model. Results show that the proposed DL models can recognize multiple devices working simultaneously through the IN and also offer an acceptable transferability performance ( ∼ 80% accuracy for a five-class problem). • System to detect impulsive noise sources using WiFi signal channel state information. • Proof that combined impulsive sources are detectable when operating simultaneously. • Showing that impulsive noise features enhance machine learning model results. • Assessing transferability in machine learning models using few-shot learning algorithms.

This study presents contributions on the detection of impulsive noise sources using OFDM-Wifi signals and Machine Learning models. The influence of impulsive noise sources on WiFi signals is used to obtain the features for supervised Machine learning models. A measurement campaign has been carried out at two indoor locations, using the Atheros CSI tool to obtain the channel state information. Feature extraction for impulsive noise detection has been performed by processing the amplitude of the channel state information of each subcarrier. These features have fed two supervised Machine Learning models, a Random Forest algorithm, and a higher level algorithm such as a Deep Neural Network. The results obtained indicate that Wifi-OFDM signals can be used for impulsive noise source recognition. The main contributions of this work focus on the extraction of suitable features for the identification of impulsive noise sources through machine learning models. The accuracy greater than 0.9 in source identification validates the proposed model, which serves as a precedent for future studies in this area.

Positioning systems are essential in most of the current wireless devices applications. For that reason, a seamless positioning service is desired even when global navigation satellite systems (GNSS) are not available. Positioning systems based on observed time difference of arrival (OTDOA) such as Long Term Evolution (LTE) or the incoming 5G, result a good positioning alternative for those GNSS-disabled situations. Errors in the time of arrival (TOA) estimation can arise from non line-of-sight (NLOS) propagation conditions or multipath channels with weak line-of-sight (LOS). This manuscript elaborates on TOA estimation based on considering several first-arriving path candidates per channel. We present a new method for multipath TOA estimation through a statistical analysis of the cross-correlation. The selection of the TOA candidates can be performed with an adaptive or fixed threshold proposal. Performance analysis has been carried out with a LTE setup in terms of the cumulative distribution function (CDF) of the horizontal position error. Simulation results show that the proposed solution improves the position accuracy with respect to a traditional positioning scheme in harsh multipath propagation conditions.

In the last years multiple Internet of Things (IoT) solutions have been developed to detect, track, count, and identify human activity from people that do not carry any device nor participate actively in the detection process. When WiFi radio receivers are employed as sensors for device-free human activity recognition, channel quality measurements are preprocessed in order to extract predictive features toward performing the desired activity recognition via machine learning (ML) models. Despite the variety of predictors in the literature, there is no universally outperforming set of features for all scenarios and applications. However, certain feature combinations could achieve a better average detection performance compared to the use of a thorough feature portfolio. Such predictors are often obtained by feature engineering and selection techniques applied before the learning process. This manuscript elaborates on the feature engineering and selection methodology for counting device-free people by solely resorting to the fluctuation and variation of WiFi signals exchanged by IoT devices. We comprehensively review the feature engineering and ML models employed in the literature from a critical perspective, identifying trends, research niches, and open challenges. Furthermore, we present and provide the community with a new open database with WiFi measurements in several indoor environments (i.e., rooms, corridors, and stairs) where up to five people can be detected. This dataset is used to exhaustively assess the performance of different ML models with and without feature selection, from which insightful conclusions are drawn regarding the predictive potential of different predictors across scenarios of diverse characteristics.

Cognitive medium access control (C-MAC) protocols employ several mechanisms that deal with the heterogeneity of a coexistence scenario with multiple wireless systems in the same band. The underutilized channels of some bands such as the TV white spaces (TVWS), are valuable and tangible resources within the congested spectrum for providing new wireless services on licensed and license-exempt basis. This paper presents a new C-MAC protocol based on a sequential frequency hopping scheme with a proactive spectrum sensing phase for opportunistic spectrum sharing. The proposal has been implemented over a software defined radio (SDR) testbed and experimentally assessed in indoor radio-propagation conditions. In addition, the performance of the proposed protocol has been compared with other C-MAC schemes, also implemented over the SDR-based framework. Performance results in terms of free-collision channel occupation rate and transmission throughput show that the proposed architecture outperforms other C-MAC protocols in scenarios with low and medium density of cognitive users per available channel. In contrast, experimental results have shown that partially observable Markov decision process (POMDP)-based solutions are more suitable when the available channels are fewer than the cognitive links.

In licensed shared access (LSA), incumbents (i.e. primary users) control the access to their spectrum bands through temporarily leasing to a third party (called LSA licensees) in a specific spatial region. Such shared access takes place based on spectrum availability information acquired from LSA repositories. In this context, this paper presents the following contributions: 1) The use of a cognitive cooperative framework based on measurement-capable devices (MCDs) in order to provide valuable information both to generate and to update radio-environment maps for LSA repositories, thus enabling more dynamic spectrum-access opportunities. 2) A new distributed and cooperative scheme among cognitive MCDs to detect and classify incumbents and LSA licensees in the LSA network. The proposal suggests a clustering strategy for cooperative spectrum sensing when neighboring MCDs observe different licensed users-whether incumbents or LSA licensees. The clustering considers the differences in underlying hidden Markov models associated with the detection of distinct licensed users. 3) Theoretical expressions for the overall probability of detection and false alarm are derived for hard cooperation strategies that may include failed MCDs in the clustering process.

The increasing scarcity in the available spectrum for wireless communication is one of the current bottlenecks impairing further deployment of services and coverage. The proper exploitation of white spaces in the radio spectrum requires fast, robust, and accurate methods for their detection. This paper proposes a new strategy to detect adaptively white spaces in the radio spectrum. Such strategy works in cognitive radio (CR) networks whose nodes perform spectrum sensing based on energy detection in a cooperative way or not. The main novelty of the proposal is the use of a cost-function that depends upon a single parameter which, by itself, contains the aggregate information about the presence or absence of primary users. The detection of white spaces based on this parameter is able to improve significantly the deflection coefficient associated with the detector, as compared to other state-of-the-art algorithms. In fact, simulation results show that the proposed algorithm outperforms by far other competing algorithms. For example, our proposal can yield a probability of miss-detection 20 times smaller than that of an optimal soft-combiner solution in a cooperative setup with a predefined probability of false alarm of 0.1.

We analyze two algorithms, viz. the affine projection algorithm for sparse system identification (APA-SSI) and the quasi APA-SSI (QAPA-SSI), regarding their stability and steady-state mean-squared error (MSE). These algorithms exploit the sparsity of the involved signals through an approximation of the l0norm. Such approach yields faster convergence and reduced steady-state MSE, as compared to algorithms that do not take the sparse nature of the signals into account. In addition, modeling sparsity via such approximation has been consistently verified to be superior to the widely used l1norm in several scenarios. In this paper, we show how to properly set the parameters of the two aforementioned algorithms in order to guarantee convergence, and we derive closed-form theoretical expressions for their steady-state MSE. A key conclusion from the proposed analysis is that the MSE of these two algorithms is a monotonically decreasing function of the sparsity degree. Simulation results are used to validate the theoretical findings.

Last-generation and future wireless communication standards, such as DVB-T2 or DVB-NGH, are including multi-antenna transmission and reception in order to increase bandwidth efficiency and receiver robustness. The main goal is to combine diversity and spatial multiplexing in order to fully exploit the multiple-input multiple-output (MIMO) channel capacity. Full-rate full-diversity (FRFD) space-time codes (STC) such as the Golden code are studied for that purpose. However, despite their larger achievable capacity, most of them present high complexity for soft detection, which hinders their combination with soft-input decoders in bit-interleaved coded modulation (BICM) schemes. This article presents a novel low-complexity soft detection algorithm for the reception of FRFD space-frequency block codes in BICM orthogonal frequency-division multiplexing (OFDM) systems. The proposed detector maintains a reduced and fixed complexity, avoiding the variable nature of the list sphere decoder (LSD) due to its dependence on the noise and channel conditions. Complexity and simulation-based performance results are provided which show that the proposed detector performs close to the optimal log-maximum a posteriori (MAP) detection in a variety of DVB-T2 broadcasting scenarios.

Recent and next-generation wireless broadcasting standards, such as DVB-T2 or DVB-NGH, are considering distributed multi-antenna transmission in order to increase bandwidth efficiency and signal quality. Full-rate full-diversity (FRFD) space-time codes (STC), such as the Golden code, have been reported to be excellent candidates, being their main drawback their detection complexity, which is enhanced when soft output is required when combined with a bit interleaved coded modulation (BICM) scheme based on low-density parity check (LDPC) codes. We present a novel low-complexity soft detection algorithm for the reception of Golden codes in LDPC based orthogonal frequency-division multiplexing (OFDM) systems. Complexity and simulation-based performance results are provided which show that the proposed detector performs close to the optimal detector in a variety of DVB-T2 broadcasting scenarios.