Search

Dimensionality reduction is an important task for Wi-Fi-based indoor localization (IL). Most such techniques do not take into account realistic data collection issues such as the presence of outliers or inconsistent fingerprint instances. These fingerprints either represent a class boundary or an outlier. Instance hardness is a measure that better characterizes such instances. Accordingly, in this work, our contribution is to propose a convolutional autoencoder-based dimensionality reduction approach that works on the basis of feature transformation and instance hardness. The encoding process of the data input involves a two-channel representation of a fingerprint dataset that holds the normalized RSS and an instance hardness measure, that is, a k-disagreeing score. The inclusion of the k-disagreeing score into the training pipeline is made with the objective of injecting instance importance for training using 1-D CNN architectures for classification. The experimentations were performed on three benchmark datasets and a collected dataset. The proposed pipeline is found to yield an accuracy of more than 97% with error deviation ranging from 2.2– $2.37{m}$ which is quite acceptable for any localization system.

Indoor localization systems are extensively used to develop positioning in various public buildings, and warehouses, for localization and navigation of users, robots and/or tracking assets. Researchers have developed and worked on variegated technologies such as, Bluetooth Low Energy, motion planning, Received Signal Strength based fingerprinting and mapping for achieving localization. Inertial Measurement Units (IMUs) are widely used in navigation that utilizes accelerometer, magnetometer, and gyroscope to sense acceleration, magnetic field, and angular rate respectively for navigation. IMUs are not only available as wearable sensors but also present in smartphones that are widely carried by users nowadays. Thus, ubiquitous localization systems can be designed with smartphone based IMU sensors. Existing survey articles on indoor localization has mostly focused on the different technologies available, and the different approaches utilized. However, existing works on IMU sensing based user localization methods need special attention as they can be extended toward a ubiquitous localization system that requires minimal fingerprinting effort from the public buildings. Accordingly, the article focuses on providing in-depth knowledge of the working procedure and discusses the challenges smartphone IMU faces. The article also surveys the fusion-based techniques used in indoor positioning and presents a comparative study of the various approaches.

The ongoing COVID-19 pandemic has caused worldwide socioeconomic unrest, forcing governments to introduce extreme measures to reduce its spread. Being able to accurately forecast the effect of unlocking in India would allow governments to alter their policies accordingly and plan ahead. The study investigated prediction forecasts using the ARIMA model on the COVID-19 data on the lockdown period and the unlock period. In this work, we have considered not only the number of positive COVID cases but also considered the number of tests carried out. The time series data sample was collected till June 2020, and the prediction and analysis are done for August 2020. The model developed and the forecasted results align very closely with the actual number of cases, and some important inferences have been drawn through the experimentation.

The sensed data have limited or no use without knowing the location. So, estimating the location of sensors is an issue in Wireless Sensor Networks (WSNs). However, WSNs are vulnerable to malicious attacks like sybil attack affecting the accuracy of localization by forging one or many anchor identities. In this paper, an information-theoretic approach is used to detect if anchors are compromised by sybil attack. This technique can be adopted by any localization algorithm to yield secure localization against sybil attack in WSNs. Simulation results claim that it improves accuracy of localization even with sybil attack.

Remote health monitoring framework using wireless body area network with ubiquitous support is gaining popularity. However, faulty sensor data may prove to be critical. Hence, faulty sensor detection is necessary in sensor-based health monitoring. In this paper, an artificial neural network (ANN)-based framework for learning about health condition of patients as well as fault detection in the sensors is proposed. This experiment is done based on human cardiac condition monitoring setup. Related physiological parameters have been collected using wearable sensors from different people. These data are then analyzed using ANN for health condition identification and faulty node detection. Libelium MySignals HW (eHealth Medical Development Shield for Arduino) v2 sensors such as ECG sensor, pulse oximeter sensor, and body temperature sensor have been used for data collection and ARDINO UNO R3 as microcontroller device. ANN method detects faulty sensor data with classification accuracy of 98%. Experimental results and analyses are given to prove the claim.

Being one of the fastest growing industries in recent times, healthcare has witnessed a complete transition in the last decade. Rapid technological and communication advancements in machine learning have contributed to predict fatal diseases with datasets available from various sources. Single-layer ensemble learning has been applied in literature extensively, to perform predictive analysis mostly in the form of bagging, boosting, and stacking. In this work, a meta learning model is proposed that combines two single-layered ensemble models to render the final prediction. The two layers of the ensemble model explore different sets of features subject to different base learners. Weighted majority voting is performed to make a decision at the two candidate ensemble models based on different feature sets that are combined at a later stage. Interestingly, one feature may be important to classify a particular instance to a class, while another feature indicates distinguishing pattern for a different class. So, the effect of important features would be captured in the prediction result of a single-layered ensemble. However, combining two ensemble models subject to selected feature and all feature sets adds robustness to the design while reducing the chances of overfitting. The role of dimensionality on the choice of ensemble layers is also considered in the work. The first layer deals with all features of the datasets and the second layer works on datasets after feature scaling and feature selection is applied on them. The proposed method is implemented for multiple benchmark datasets of the healthcare domain. Experimentally, it has been observed that double ensemble methods provide a better predictive analysis in comparison to single-layered ensemble models if the dataset dimensionality is moderate. The problem of overfitting is also significantly reduced in double-layered ensemble approach. The combination of more than one classifier also eliminates the dependency on a single classifier to perform predictions.

Mobile adhoc network (MANET) provides a promising platform for pervasive computing applications. Mobile agents are found to be effective for executing such applications. The motivations behind this are advancement in technology, wireless networking, sensor network etc. However, since MANET is inherently more vulnerable to security threats and prone to topology changes, reliability and security issues must be addressed before mobile agents are commercially deployed. There are few works on securing mobile agents but even fewer focuses on MANET. This work is our attempt to design a lightweight trust based reputation scheme to protect the agents against network layer threats. The scheme is based on Dempster-Shafer belief theory. Performance of the trust based reputation scheme with respect to network and system reliability is analysed. The work is simulated and the results show that even for a fairly hostile MANET, the effective reliability of distributed application can be increased using mobile agents.

Lifetime of a battery is one of the major concerns in wireless sensor networks (WSNs). The traditional network algorithm chooses the data transmission path based on the distance between sender and receiver. For example, in directed diffusion, the nodes closer to the sink node are more active so that the continuous data flow of the network is maintained. However, in such a scenario, there is an energy hole problem because the nodes closer to the sink reduce their energy since they are more active. But if the residual energy and shortest hop counts can be taken together, then the problem can be overcome. We have analyzed different transmission cost functions with respect to approximated uniform energy dissipation directed diffusion algorithm. The final goal of the paper is to find out the suitable cost function to resist the early death of the first node of a network. We introduce dynamic priority variable, which is a tradeoff between total energy consumption and uniform energy dissipation.

Localization of sensor nodes is one of the important issues in wireless sensor networks (WSNs). The location of a node can be used as the location of the occurrence of an event. Error handling and scalability are key research issues that need to be taken care of while estimating the efficiency of any localization algorithm. In this paper, we propose an approach of error correction mechanism in addition to the minimization of error in multihop system (MEMHS) for the localization algorithm. The MEMHS algorithm deals with a scalable error correction of a multiliterate localization process using a few geographical positioning system enabled nodes. The MEMHS authors assumed that an error propagates linearly and is equal in any direction. In the present study, the authors show that an error propagates nonlinearly with respect to the hop count, and the magnitude of error (X-coordinate or Y-coordinate) depends on the direction of equator lines. This paper proposes a modified algorithm of MEMHS. Furthermore, an optimum deployment strategy is introduced so that maximum number of sensor nodes can be localized. By analyzing the proposed algorithm in comparison with the MEMHS algorithm, it is found that the proposed algorithm has a better performance in terms of error correction.

Mobile ad hoc networks (MANETs) are comprised of wireless mobile nodes and are gaining popularity due to their ease of deployment, low cost, and simple communication. However, MANETs face numerous challenges such as limited node energy, high mobility, and dynamic topologies, leading to frequent link breakage and route discovery. Although several researchers have worked on improving existing protocols, there is still ample room for improvement. This paper proposes a routing algorithm called MO-AOMDV (Multi-objective AOMDV) that selects routes based on the quality of nodes and links denoted by G(V, L), where Vrepresents the number of nodes and Lrepresents the links between them. Additionally, each node maintains a local list that updates the remaining energy of the node and its neighbors. The algorithm is developed and simulated using a network simulator and compared with traditional AODV and AOMDV algorithms based on packet drop ratio, packet delivery ratio, and network lifetime.

Huge amount of data is produced and processed in recent data-centric applications. Secure management as well as maintaining privacy of the data is a challenging scenario as data itself store the sensitive data along with other application data. Protecting sensitive data is very challenging as it could not be quantified directly. Here, we formulate a metric sensitivity-score to calculate the sensitivity value of the data attributes in a dataset. Sensitive attributes are segregated carefully to avoid possible data linkage attacks by the legitimate users of the application data. Micro-data format is good for maintaining privacy for sensitive data. However, the utility of the data will decrease exponentially. So here in this paper, the authors try to model the data in such a way that a balance between privacy and utility is maintained. The entire data set is segregated in micro-data format with attributes based on the sensitivity value. A Decision Tree-based classifier is used to label the attributes of a sample healthcare dataset as Sensitive or not. Experiments are also conducted to compare the utility and the privacy factor of the proposed method with other existing data partitioning algorithm.

Formation of scatternet using Bluetooth devices increases device tractability thereby inviting new networking applications to be designed on it. In this paper we propose Bluetooth Scatternet Formation and Routing Protocol (BSFRP). It is a distributed protocol that handles node mobility and enables multi-hop communication. BSFRP defines rules for topology discovery, scatternet formation and routing. The scatternet phase of the protocol works on the principle of leader election. For routing, AODV is modified to address the constraints of scatternets. It improves the AODV route discovery phase by considering hop count, residual node's power, and route lifetime for best route selection. Simulation results show that the scatternet formed by BSFRP has the following properties: the number of piconets formed is close to the universal lower bound, each device on an average does not assume more than 1.15 roles, and the scatternet does not contain any master-slave bridge.

COVID-19 is creating havoc on the lives of human beings all around the world. It continues to affect the normal lives of people. As number of cases are high, a cost effective and fast system is required to detect COVID-19 at appropriate time to provide the necessary healthcare. Chest X-rays have emerged as an easiest way to detect COVID-19 in no time as RT-PCR takes time to detect the infection. In this paper we propose a concatenation-based CNN model that will detect COVID-19 from chest X-rays. We have developed a multiclass classification problem which can detect and classify a chest X-ray image as either COVID + ve, or viral pneumonia, or normal. We have used chest X-rays collected from different open sources. To maintain class balancing, we took 500 images of COVID, 500 normal images, and 500 pneumonia images. We divided our dataset in training, validation, and test set in 70:10:20 ratio respectively. We used four CNNs as feature extractors from the images and concatenated their feature maps to get better efficiency of the network. After training our model for 5 folds, we have obtained around 96.31% accuracy, 95.8% precision, 92.99% recall, and 98.02% AUC. We have compared our work with state-of-the-art pretrained transfer learning algorithms and other state-of-the-art CNN models referred in different research papers. The proposed model (Concat_CNN) exhibits better accuracy than the state-of-the-art models. We hope our proposed model will help to classify chest X-rays effectively and help medical professionals with their treatment.

In this work, we have proposed a mobility-based secure checkpointing and log-based rollback recovery technique to provide fault tolerance to mobile hosts (MHs) in infrastructured wireless/mobile computing system, like, wireless cellular network. This work has been modified accordingly to be applicable in infrastructure less network, i.e., mobile ad hoc network (MANET) in which MHs save checkpoint adaptively based on their energy and mobility. Transient failures in cluster-based mobile ad hoc network (MANET) may be caused by insufficient energy of mobile hosts, among other reasons. To prevent data loss due to transient failure caused by lack of sufficient energy, mobile hosts save checkpoint if remaining energy becomes equal to a predefined threshold. A trust-based recovery technique is used for successful recovery of failed mobile hosts. To make checkpoints secure, only trusted mobile hosts participate in the proposed checkpoint-recovery process.

Nowadays there is an increasing trend of applying wireless technologies in industrial automation. However, the industrial control environment is harsher and noisier posing more stringent requirements on real-time communication. The variation in wireless signal strength with time and location, and power limitation due to battery usage make the problem even worse. Accordingly, availability and reliability of such system should be seriously analyzed before deployment. Modeling the reliability of these distributed real-time systems with nodes having embedded software has become an integral part of the design process. The time constraints that real-time systems must meet directly affect the reliability measure itself irrespective of the nature of the measure. A metric for calculating system reliability is proposed in this paper and a Monte Carlo simulation based algorithm for calculating the same is presented here. The results show that system reliability stabilizes with time even with increasing network size.