The range-based localization approaches using received signal strength (RSS) are inaccurate due to randomness in received signal power arises from shadowing and multipath effects. By modeling multipath effects using a generalized fading environment, such as$\alpha {-}\mu $distribution improves localization accuracy while maintaining low model complexity. We derive estimator’s localization error bound for model considering shadowing and multipath effects as Gaussian and$\alpha {-}\mu $distribution, respectively. We also compare the location estimator performance to assess potential areas of algorithm improvement and parameters on which CRB depends. Furthermore, this letter also provides parameters that influence localization performance based on experiments performed on two real-world testbeds.

This article proposes a nondata-aided (NDA) signal-to-noise ratio (SNR) estimation method for molecular diffusive communication channels. In contrast to previous data-aided (DA) approaches, which require a known data sequence to be shared between transmitter and receiver nanomachines, the proposed NDA method eliminates the need for prior knowledge of data symbols for the estimation process. The absence of a training sequence significantly reduces resource costs by minimizing the number of molecules required for information transmission. The expectation–maximization (EM) algorithm was used to iteratively find the maximum-likelihood (ML) estimates. The proposed method is particularly advantageous in scenarios in which accessing known data sequences is not possible, thereby ensuring robust communication in the Internet of Bio-Nano Things. By employing the ML estimates obtained through the EM algorithm, the probabilities associated with the transmission of data symbols were derived, and the bit error probabilities were obtained to assess the system performance. The numerical results validated the effectiveness of the proposed NDA SNR estimator in accurately estimating SNR over a diverse range of conditions. The proposed NDA estimator demonstrated resilience by maintaining a low bit error probability.

The Internet of Bio-Nano Things (IoBNT) enables complex tasks like health monitoring and environmental sensing, requiring precise clock synchronization between nanomachines. In this article, we propose a clock synchronization method based on the arrival times of molecules that considers both clock offset and clock skew parameters in a timing-based diffusive channel. The timing-based diffusive channel requires a degradation factor to limit the lifespan of molecules. The existing methods assume ideal conditions for the synthesis time and arrival time of the molecules. However, we consider a practical scenario where the molecule synthesis time is nonzero and the first arrival times of the molecule are correlated for the joint estimation of the clock offset and clock skew. Individual estimation of clock offset requires prior knowledge of the clock skew. However, in practice, both parameters are unknown. Therefore, we address the synchronization issues that arise from individual estimation of the clock offset by jointly estimating both parameters. Using the transmission of molecules and their correlated arrival times, the receiver nanomachine (RN) employs the maximum likelihood (ML) estimator to estimate the synchronization parameters. Numerical results demonstrate that the proposed estimator outperforms the existing one.

Scarcity of labeled data in sensitive research areas such as healthcare limits the performance of artificially intelligent (AI) models. The effort required in labeling the acquired healthcare data for carrying out a classification task is a concern faced by healthcare researchers. In this work, we design a capsule-long short-term memory (LSTM) model, abbreviated as CapsLSTM, capable of classifying human activities with scarce labeled activity data. The proposed CapsLSTM model is used to recognize multiple human activities sensed by the accelerometer and the gyroscope sensors embedded in smartphones leveraging the spatio-temporal information. We validate the proposed framework based on two human activity recognition (HAR) databases, namely, UCI-HAR and MotionSense. The CapsLSTM model yields close test accuracies for different fractions of the training data unlike the other models such as LSTM, 1-D convolutional neural network (1D-CNN), convolutional LSTM (ConvLSTM), or CNN-LSTM. For the state-of-the-art models, the test accuracies decrease significantly with the decrease in training data. Our proposed model classifies the human activities using a minimum of 20% labeled training data from each database. There is less decrease in accuracies from that obtained using 70% of training data in comparison to the existing models. The classification performance of the proposed CapsLSTM model for a small fraction of training data proves the effectiveness and reliability of the same in a data-scarce situation.

Device heterogeneity can cause a detrimental impact on the classification of healthcare data. In this work, we propose the Maximum Difference-based Heterogeneity Mitigation (MDHM) method to address device heterogeneity. Mitigating heterogeneity increases the reliability of using multiple devices from different manufacturers for measuring a particular physiological signal. Further, we propose an attention-based bilevel GRU (Gated Recurrent Unit) model, abbreviated as AT2GRU, to classify multi-modal healthcare time-series data for human emotion recognition. The physiological signals of Electroencephalogram (EEG) and Electrocardiogram (ECG) for twenty-three persons are leveraged from the DREAMER dataset for emotion recognition. Also, from the DEAP dataset, the biosignals namely EEG, Galvanic Skin Response (GSR), Respiration Amplitude (RA), Skin Temperature (ST), Blood Volume (BV), Electromyogram (EMG) and Electrooculogram (EOG) of thirty-two persons are used for emotion recognition. The EEG and the other biosignals are denoised by the wavelet filters for enhancing the model's classification accuracy. A multi-class classification is carried out considering valence, arousal, and dominance for each person in the datasets. The classification accuracy is validated against the self-assessment obtained from the respective person after watching a movie/video. The proposed AT2GRU model surpasses the other sequential models namely Long Short Term Memory (LSTM) and GRU in performance.

Gender plays a crucial role in improving the performance quality of personalized systems. Privacy and anonymity allow users to hide their details. Based on the intuition that post contents of male and female users differ, we can predict the gender of the social media account holder via their corresponding posts. These posts can be multimodal (text + image) in nature. We investigate various emotion-assisted multimodal gender prediction models in this article. The developed models use gated recurrent units (GRUs) and ResNets for extracting features from the tweets and the images, respectively. The distribution of emotion categories is related to the gender of the target person. In response to these findings, this article describes the first attempt to use multimodal (image and text posted) information for gender prediction in a multitask setting with emotion recognition as an auxiliary task. The enriched PAN-2018 dataset with gender and emotion labels is used to train gender and emotion networks. Several models were developed to improve the gender prediction task by generating better emotion-aware features. The results show that the proposed multimodal emotion detection model outperforms single-modal (text and image)-based models and state-of-the-art systems on the benchmark PAN-2018 dataset.

Wireless Fidelity (Wi-Fi)-received signal strength (RSS) fingerprints are extensively used for localization in an indoor Internet of Things (IoT) environment. However, RSS fingerprinting-based localization methods face two major challenges. First, the labor-intensive and time-consuming task of constructing a fingerprinting database with a large number of RSS samples for offline training. Second, the RSS values vary due to device heterogeneity and temporal variance caused by the change in access points (APs), device orientation, and environmental factors. These dynamic RSS variations in the online phase decrease localization accuracy. This article proposes a Siamese embedding-based localization method for a dynamic IoT environment. The proposed method provides high localization accuracy with limited RSS samples by utilizing our proposed distance-based sampling method. Also, the Siamese embeddings capture the spatial relations between locations and remain persistent even with the RSS variations in the online phase. This is possible with the help of a lightweight, offline fine-tuning method that requires minimal RSS samples. The proposed method is validated on three real IoT testbeds with RSS variations. It outperforms the existing state-of-the-art methods on two testbeds along with the comparable results on the third.

Device heterogeneity is a social concern, especially in healthcare domain. In this work, we mitigate the problem of device heterogeneity and further classify the healthcare electrocardiogram (ECG) data with improved performance using a proposed variant of long short-term memory (LSTM). ECG data sensed from different devices are used in this work for experimentation. Device heterogeneity is addressed using the proposed multiplicative convergence-based heterogeneity mitigation (MCHM) method. The proposed warehouse LSTM, in addition to support vector machine (SVM), is leveraged in this work for healthcare data classification. The warehouse LSTM keeps a track of the data that are considered as insignificant in the initial epoch. We mitigate heterogeneity in medical devices and reduce the root-mean-squared error to the order of 10−6–10−5. Using the warehouse, the LSTM-SVM attains the classification accuracies of 98.34% and 96.27% during training on the MIT-BIH and PTB datasets, respectively. The proposed MCHM method increases the reliability on the usage of devices from multiple manufacturers. The novel warehouse LSTM-SVM model also outperforms the existing methods for classification of data.

The gender of a user plays a very important role in the development of the personalized online services. However, due to privacy, and anonymity, gender information is usually not available for many users. Since male and female users have differences in their message contents, the messages posted by users can be analyzed for finding their genders. Users with different genders express their emotions differently too. In the current paper, we introduce an emotion-aided gender prediction system. The intuition behind our approach is to predict the gender of a user based on emotional clues. The proposed approach consists of two neural network branches, one for gender and the other for emotion. Both networks share their bottom feature extraction module and are optimized within a multi-task learning framework. Gender and emotion networks are trained over our own annotated dataset having gender and emotion labels. Experimental results show the effectiveness of using emotion for predicting the gender of a user.

Location estimation in an indoor Internet-of-Things (IoT) environment is a challenging task due to multipath signals and obstacles that cause shadowing and fading effects, and change the received signal power considerably. Most of the existing path-loss-based localization methods assume only a lognormal shadowing model and ignore small scale fading effects. This article considers a generic combined lognormal shadowing and Rayleigh fading model for efficient localization of smart devices in an indoor IoT environment. In particular, the maximum likelihood estimate of the location and path-loss exponent (PLE), and Cramer-Rao lower bound (CRLB) are derived. The localization parameters are estimated using a novel adaptive mini-batch gradient ascent method that maximizes the log-likelihood function with an appropriate batch size based on the convergence factor. Hence, the proposed method addresses the challenge of an arbitrary selection of a fixed batch size for a gradient ascent method by utilizing this convergence factor. Performance evaluation by a simulation study and real experiments from an indoor IoT testbed provide a more accurate joint estimation of model parameters and smart device localization.

Location-based services assume significant importance in the Internet of Things (IoT) based systems. In the scenarios where the satellite signals are not available or weak, the Global Positioning System (GPS) accuracy degrades sharply. Therefore, opportunistic signals can be utilized for smart device localization. In this paper, we propose a smart device localization method using residual neural networks. The proposed network is generic and performs smart device localization using opportunistic signals such as Wireless Fidelity (Wi-Fi), geomagnetic, temperature, pressure, humidity, and light signals in the IoT network. Additionally, the proposed method addresses the two significant challenges in IoT based smart device localization, which are noise and device heterogeneity. The experiments are performed on three real datasets of different opportunistic signals. Results show that the proposed method is robust to noise, and a significant improvement in the localization accuracy is obtained as compared to the state-of-the-art localization methods.

In this article, we recognize multiple human activities in an Internet-of-Things (IoT) network using differential channel state information (CSI) of the available wireless fidelity (Wi-Fi) signals. Different human activities in the Wi-Fi environment lead to multipath fading, resulting in a change of CSI for each activity. This CSI is sensed by smart IoT devices, such as smartphones, tablets, and laptops for activity recognition. The use of differential CSI mitigates the offset and background noise. Another advantage of the proposed method is that it eliminates the requirement of traditional wearable activity recognition sensors, such as gyroscope, pedometers, and accelerometers. A long short-term memory (LSTM) model is used for automatic feature extraction and classification of human activities from the differential CSI. Training the LSTM model with the phase of differential denoised CSI significantly improves the classification accuracy. The results show a good tradeoff between model complexity and classification accuracy, thereby ensuring better performance as compared to the previous state-of-the-art methods.

This paper analyses the performance of target detection and localization methods in heterogeneous sensor networks using compartmental model, which is an attenuation model expressing the variation of received signal strength (RSS) with propagation distance. First, we compute the threshold for the proposed target detection scheme, based on the decision fusion of different sensors and without requiring a priori probability. We also derive the bound on the threshold and subsequently the lower and upper bounds on the detection and false-alarm probabilities. Next, the location of the detected target is estimated using iterative mini-batch Singular Value Decomposition (SVD) methods in the presence of sensor location uncertainty. We highlight that the method for localization has low computational complexity which is suitable for Internet of Things (IoT) networks. The effectiveness of the compartmental model is demonstrated using both simulation study and real experiments. The model parameters are estimated using WiFi signal strength received on the mobile phones from the access points in an indoor environment.

In this paper, the problem of optimal placement of an Unmanned Aerial Vehicle (UAV) is addressed for seamless indoor IoT connectivity of the worst-case (farthest) multiple users. The optimization of the transmitted power from UAV is carried out so that all the users can receive the transmitted data from the UAV with a minimum path loss. The proposed method can be used for signals such as Wi-Fi, Bluetooth or cellular, amongst others. We utilize the ITU-R recommended path loss model and place the UAV, ensuring seamless connectivity of the worst-case users. The relationship between the locations of the user and UAV is also established along with the derivation of the optimality condition for maximum power transmitted. The theoretical results are in good agreement with the simulation results. The proposed method is validated for multiple users instead of a single user in previous methods using numerical experiments.

One read-write 8T SRAM cell with decoupled read port suffer data-dependent read bit line leakage, limiting VDDMIN of SRAMs in deep submicron technologies. This paper proposes a 10T SRAM cell in a typical sub-10nm FinFET technology with data-independent RBL leakage and enhanced RBL sensing margin: both voltage and time window to improve VDDMIN. 10T SRAM cell data-independent RBL leakage is improved by 15% compared to conventional 8T SRAM cell. At VDD=0.3V, our proposed 10T SRAM cell show significant RBL sensing voltage improvement of 155mV (~8.2x) compared to 19mV in conventional 8T SRAM. We also proposed a weak NMOS keeper to improve VDDMIN and Read `0' delay for large signal RBL sensing vs. conventional weak PMOS keeper. Our proposed weak NMOS keeper with NFIN=2 and NFIN=3 improves VDDMIN by 50mV and 90mV respectively against conventional PMOS keeper. Read `0' delay also improved for our proposed weak NMOS keeper with NFIN=2 and NFIN=3 against conventional weak PMOS keeper (0.15× at VDD=0.5V).

A novel method for device localization under mixed line-of-sight/non-line-of-sight (LOS/NLOS) conditions based on second order cone programming (SOCP) is presented in this paper. The devices can communicate cooperatively among themselves in a large internet of things (IoT) network. SOCP methods have, hitherto, not been utilized in the node localization under mixed LOS/NLOS conditions. Unlike semidefinite programming (SDP) formulation, SOCP is computationally efficient for resource constrained IoT network. The proposed method can work seamlessly in mixed LOS/NLOS conditions. The robustness of the method is due to the fair utilization of all measurements obtained under LOS and NLOS conditions. The computational complexity of this method is quadratic in the number of nearest neighbours of the unknown node. Cramér-Rao bound and localization error are analyzed to illustrate the effectiveness of the proposed method. The experimental results of the proposed method indicate a reasonable improvement when compared to recent state of the art methods.

In this paper, we propose a system for monitoring, tracking, and automating the trains. In contrast to the existing methods, we employ a global position system (GPS) and Global System for Mobile communication (GSM) by which each train is individually monitored and necessary messages are passed on proactively. The proposed system has advantages in terms of communication range and accuracy with respect to Zigbee, Wi-Fi, RFID based rail tracking method. The work has potential applications in bad weather and emergency situations like collision.

In this paper, a method for estimation of speed for the mobile phone user using WiFi signal-to-noise ratio (SNR) is proposed. The proposed method does not utilize additional hardware like accelerometer or gyroscope to estimate the speed unlike existing methods. Time-domain features like mean, maximum, and auto-correlation are derived using WiFi signal-to-noise ratio. The experiments are carried out extensively in two different environments. The feature that is taken into account for estimating the speed is maximum SNR. This proposed method has low complexity and a reasonable accuracy in a WiFi or cellular environment. The proposed method is low-cost and is easily adaptable. This method is unaffected in a multi-receivers environment.

In Internet of Things (IoT), numerous and diverse types of sensors generate a plethora of data that needs to be stored and processed with minimum loss of information. This demands efficient compression mechanisms where loss of information is minimized. Hence data generated by diverse sensors with different signal features require optimum balance between compression gain and information loss. This paper presents a unique analysis of contemporary lossy compression algorithms applied on real field sensor data with different sensor dynamics. The aim of the work is to classify the compression algorithms based on the signal characteristics of sensor data and to map them to different sensor data types to ensure efficient compression. The present work is the stepping stone for a future recommender system to choose the preferred compression techniques for the given type of sensor data.

In this paper, a hybrid maximum depth - k Nearest Neighbour (hybrid MD-kNN) method for real time sensor node tracking and localization is proposed. The method combines two individual location hypothesis functions obtained from generalized maximum depth and generalized kNN methods. The individual location hypothesis functions are themselves obtained from multiple sensors measuring visible light, humidity, temperature, acoustics, and link quality. The hybridMD-kNN method therefore combines the lower computational power of maximum depth and outlier rejection ability of kNN method to realize a robust real time tracking method. Additionally, this method does not require the assumption of an underlying distribution under non-line-of-sight (NLOS) conditions. Additional novelty of this method is the utilization of multivariate data obtained from multiple sensors which has hitherto not been used. The affine invariance property of the hybrid MD-kNN method is proved and its robustness is illustrated in the context of node localization. Experimental results on the Intel Berkeley research data set indicates reasonable improvements over conventional methods available in literature.

In this paper, a single mobile beacon based method to localize nodes using principle of maximum power reception is proposed. Optimal positioning of the mobile beacon for minimum energy consumption is also discussed. In contrast to existing methods, the node localization is done with prior location of only three nodes. There is no need of synchronization, as there is only one mobile anchor and each node communicates only with the anchor node. Also, this method is not constrained by a fixed sensor geometry. The localization is done in a distributed fashion, at each sensor node. Experiments on node-source localization are conducted by deploying sensors in an ad-hoc manner in both outdoor and indoor environments. Localization results obtained herein indicate a reasonable performance improvement when compared to conventional methods.