Article

# A Gaussian Process Regression Approach to Predict the k-barrier Coverage Probability for Intrusion Detection in Wireless Sensor Networks

If you want to read the PDF, try requesting it from the authors.

## Abstract and Figures

Sensors in Wireless Sensor Network (WSN) sense, process, and transmit information simultaneously. They mainly find applications in agriculture monitoring, environment monitoring, smart city development and defence. These applications demand high-end performance from the WSN. However, the performance of a WSN is highly vulnerable to various types of security threats. Any intrusion may reduce the performance of the WSN and result in fatal problems. Hence, fast intrusion detection and prevention is of great use. This paper aims towards fast detection and prevention of any intrusion using a machine learning approach based on Gaussian Process Regression (GPR) model. We proposed three methods (S-GPR, C-GPR and GPR) based on feature scaling for accurate prediction of k-barrier coverage probability. We have selected the number of nodes, sensing range, Sensor to Intruder Velocity Ratio (SIVR), Mobile to Static Node Ratio (MSNR), angle of the intrusion path and required k as the potential features. These features are extracted using an analytical approach. Simulation results demonstrate that the proposed method III accurately predicts the k-barrier coverage probability and outperforms the other two methods (I and II) with a correlation coefficient (R = 0.85) and Root Mean Square Error (RMSE = 0.095). Further, the proposed methods achieve a higher accuracy as compared to other benchmark schemes.
Content may be subject to copyright.

## No full-text available

... The problem at hand can be resolved by employing machine learning approaches that are exceptionally competent for computational time [20,21]. For instance, the work presented in [22] provided a mathematical framework to evaluate the k-barrier coverage probability for a given intrusion path using mobile WSNs. The authors have proposed three machine learning models based on the GPR algorithm to predict the k-barrier coverage probability to overcome the computational and time complexity problem. ...
... These datasets can either be field derived (obtained by direct measurements) or generated synthetically (obtained through simple rules, statistical modelling, and simulations) [24]. The use of synthetic data is increasing exponentially in the domain of healthcare [25,26], WSNs [22,27], and data privacy [28]. ...
... To calculate each feature's relative importance score, we created a regression ensemble through boosting ensemble learning. We leverage LSBoost (Least Square gradient Boosting) algorithm to boost hundred regression trees, each having unity learning rate [22,30]. This algorithm assumes each decision tree as a weak learner and processes them individually by identifying their weak points. ...
Article
Full-text available
The dramatic increase in the computational facilities integrated with the explainable machine learning algorithms allows us to do fast intrusion detection and prevention at border areas using Wireless Sensor Networks (WSNs). This study proposed a novel approach to accurately predict the number of barriers required for fast intrusion detection and prevention. To do so, we extracted four features through Monte Carlo simulation: area of the Region of Interest (RoI), sensing range of the sensors, transmission range of the sensor, and the number of sensors. We evaluated feature importance and feature sensitivity to measure the relevancy and riskiness of the selected features. We applied log transformation and feature scaling on the feature set and trained the tuned Support Vector Regression (SVR) model (i.e., LT-FS-SVR model). We found that the model accurately predicts the number of barriers with a correlation coefficient (R) = 0.98, Root Mean Square Error (RMSE) = 6.47, and bias = 12.35. For a fair evaluation, we compared the performance of the proposed approach with the benchmark algorithms, namely, Gaussian Process Regression (GPR), Generalised Regression Neural Network (GRNN), Artificial Neural Network (ANN), and Random Forest (RF). We found that the proposed model outperforms all the benchmark algorithms
... Also, IoT devices are very expensive and require a huge amount of financial investment. The high time complexity and financial issues can be minimised to a negligible level using machine learning approaches to validate and predict the performance of WSNs before their actual deployment in a given region (Mishra et al., 2018;Singh et al., 2021b;Kotiyal et al., 2021). However, the accurate and timely detection and prevention of intrusion through machine learning approaches is still an ill-posed problem that has been insufficiently investigated. ...
... Here, they have calculated the total area covered by an intruder traveling at a given angle to cross the RoI, Then, this total area is utilised to obtain the closed-form expression for the k-barrier coverage probability of the WSN. In an another work presented in Singh et al. (2021b), the authors have employed three machine learning approaches, namely Gaussian Process Regression (GPR), Scaling GPR (S-GPR), and Center mean GPR (C-GPR) to predict the k-barrier coverage probability of a WSN. The proposed GPR technique quickly detects and prevents any intrusion taking place at any location in the RoI. ...
... However, it does not convey how the features are associated with the predictand, i.e., whether the predictand value increases with feature (positive impact) or decreases with features (negative impact). To evaluate this, we have performed the sensitivity analysis of the features using Partial Dependence Plot (PDP) (Singh et al., 2021b;Friedman, 2001). PDP measures the average effect of a single or more feature by marginalising the effect of all other features taken into consideration. ...
Preprint
Full-text available
Wireless Sensor Networks (WSNs) is a promising technology with enormous applications in almost every walk of life. One of the crucial applications of WSNs is intrusion detection and surveillance at the border areas and in the defense establishments. The border areas are stretched in hundreds to thousands of miles, hence, it is not possible to patrol the entire border region. As a result, an enemy may enter from any point absence of surveillance and cause the loss of lives or destroy the military establishments. WSNs can be a feasible solution for the problem of intrusion detection and surveillance at the border areas. Detection of an enemy at the border areas and nearby critical areas such as military cantonments is a time-sensitive task as a delay of few seconds may have disastrous consequences. Therefore, it becomes imperative to design systems that are able to identify and detect the enemy as soon as it comes in the range of the deployed system. In this paper, we have proposed a deep learning architecture based on a fully connected feed-forward Artificial Neural Network (ANN) for the accurate prediction of the number of k-barriers for fast intrusion detection and prevention. We have trained and evaluated the feed-forward ANN model using four potential features, namely area of the circular region, sensing range of sensors, the transmission range of sensors, and the number of sensor for Gaussian and uniform sensor distribution. These features are extracted through Monte Carlo simulation. In doing so, we found that the model accurately predicts the number of k-barriers for both Gaussian and uniform sensor distribution with correlation coefficient (R = 0.78) and Root Mean Square Error (RMSE = 41.15) for the former and R = 0.79 and RMSE = 48.36 for the latter. Further, the proposed approach outperforms the other benchmark algorithms in terms of accuracy and computational time complexity.
... Also, IoT devices are very expensive and require a huge financial investment. The high time complexity and financial issues can be minimised to a negligible level using machine learning approaches to validate and predict the performance of WSNs before their actual deployment in a given region (Kotiyal, Singh, Sharma, Nagar, & Lee, 2021;Mishra, Varadharajan, Tupakula, & Pilli, 2018;Singh, Nagar, Sharma and Kotiyal, 2021). However, the accurate and timely detection and prevention of intrusion through machine learning approaches is still an ill-posed problem that has been insufficiently investigated. ...
... Here, they have calculated the total area covered by an intruder travelling at a given angle to cross the RoI, Then, this total area is utilised to obtain the closedform expression for the -barrier coverage probability of the WSN. In another work presented in Singh, Nagar et al. (2021), the authors have employed three machine learning approaches, namely Gaussian Process Regression (GPR), Scaling GPR (S-GPR), and Center mean GPR (C-GPR) to predict the -barrier coverage probability of a WSN. The proposed GPR technique quickly detects and prevents any intrusion from taking place at any location in the RoI. ...
... However, it does not convey how the features are associated with the predictand, i.e., whether the predictand value increases with feature (positive impact) or decreases with features (negative impact). To evaluate this, we have performed the sensitivity analysis of the features using the Partial Dependence Plot (PDP) (Friedman, 2001;Singh, Nagar et al., 2021). PDP measures the average effect of a single or more features by marginalising the effect of all other features taken into consideration. ...
Article
Full-text available
Wireless Sensor Networks (WSNs) is a promising technology with enormous applications in almost every walk of life. One of the crucial applications of WSNs is intrusion detection and surveillance at the border areas and in the defense establishments. The border areas are stretched in hundreds to thousands of miles, hence, it is not possible to patrol the entire border region. As a result, an enemy may enter from any point absence of surveillance and cause the loss of lives or destroy the military establishments. WSNs can be a feasible solution for the problem of intrusion detection and surveillance at the border areas. Detection of an enemy at the border areas and nearby critical areas such as military cantonments is a time-sensitive task as a delay of few seconds may have disastrous consequences. Therefore, it becomes imperative to design systems that are able to identify and detect the enemy as soon as it comes in the range of the deployed system. In this paper, we have proposed a deep learning architecture based on a fully connected feed-forward Artificial Neural Network (ANN) for the accurate prediction of the number of k-barriers for fast intrusion detection and prevention. We have trained and evaluated the feed-forward ANN model using four potential features, namely area of the circular region, sensing range of sensors, the transmission range of sensors, and the number of sensor for Gaussian and uniform sensor distribution. These features are extracted through Monte Carlo simulation. In doing so, we found that the model accurately predicts the number of k-barriers for both Gaussian and uniform sensor distribution with correlation coefficient (R = 0.78) and Root Mean Square Error (RMSE = 41.15) for the former and R = 0.79 and RMSE = 48.36 for the latter. Further, the proposed approach outperforms the other benchmark algorithms in terms of accuracy and computational time complexity.
... Recently, the authors in [19] proposed a regression-based machine learning algorithm for the accurate prediction of -barrier coverage probability. They showed how machine learning algorithms could be used for fast intrusion detection and prevention. ...
... PDP measures the average marginal effect of each feature over the response variable. In contrast, ICE shows the dis-aggregated results of PDP [19]. ...
... In earlier attempts, Singh et al. 2021 [19] proposed several novel approaches for predicting -barrier coverage probability using regression-based machine learning algorithms (support vector regression and gaussian process regression). They reported an R = 0.85 and an RMSE of 0.095 for the best machine learning model. ...
Article
Full-text available
Wireless Sensor Networks (WSNs) is one of the most widely employed technology because it has numerous applications in almost every walk of life. The analytical results available for large-scale WSNs cannot be utilised to estimate the performance of WSNs deployed in a finite region due to Boundary Effects (BEs). In addition, wireless channel characteristics are affected by diverse environmental phenomena such as the presence of impediments, interference, reflection, and refraction, etc. Therefore, we render an analytical model by considering BEs in the shadowed environments to estimate the κ-coverage metric of a WSN installed in a circular region (CR). Validation of the analytical models is a time-consuming and tedious task and requires hours. To overcome this problem, in this study, we proposed a framework based on feed-forward Artificial Neural Network (ANN) to map the κ-coverage probability using nodes, sensing range, the standard deviation of shadowing denoted by sigma, and required κ as features. These features were extracted through Monte Carlo simulations. We estimated the feature importance and performed the feature sensitivity analysis before training the ANN model. We trained two feed-forward ANN models for with and without BEs. We found sensing range is the most important feature in predicting the κ-coverage probability. Further, the proposed feed-forward framework performs equally well for both cases, with correlation coefficient (R) = 0.98 and Root Mean Square Error (RMSE) = 0.07. Furthermore, it also outperforms the results obtained through the Adaptive Neuro-Fuzzy Inference System (ANFIS).
... Technological accomplishments in micro-electromechanical systems have led to the fabrication of wireless nodes immensely being deployed to form WMNs-also recognized as wireless sensor networks (WSNs) and wireless ad hoc networks (WANETs) [1][2][3]. WMNs comprise an abundance of tiny, low-power, low-cost, but failure-prone nodes. These nodes operate in a distributed fashion without the necessity of any permanent foundation, i.e., nodes convey information with each other via either a single hop or a multihop path through a wireless channel [4]. ...
... WMNs can be set up ''on the fly'' smoothly in areas of emergency operations, in an unreachable territory, and in other hazardous surroundings. This fact has resulted in many applications of these networks in several decisive/non-decisive situations such as enemy chasing and espionage, telecommunication systems, automation, health, security, and mines detection in oceans [1][2][3][4][5][6][7][8][9], etc. A WMN is connected when there is at least one wireless route between each pair of nodes. ...
... Mainly, there are two challenges in calculating the NIP P Iso r max ð Þ in (9) and the MNDD f D j; r max ð Þin (10), viz., 1 The first challenge is to compute the overlapping area A P; r max ð Þ\R j j = L Â x ð Þ by integrating the SEs and the BEs to achieve the cumulative distribution function (cdf) (8). 2 The second one is to compute the average value of cdf in (8) over the whole RSR R in shadowing environment to estimate the NIP P Iso r max ð Þ(9) and the MNDD f D j; r max ð Þ (10). ...
Article
Full-text available
Connectivity is one of the critical performance parameters of wireless multihop networks (WMNs) explored in detail for the last few years. The assumption, a constant communication range (CR) of sensor nodes in all possible paths taken in earlier studies may not be accurate due to many fluctuations in the received signal strength (RSS) caused by the random nature of wireless channels and the presence of obstacles in the communication environment. Studies have also ignored the border effects (BEs), rendering an overestimated and erroneous result on performance parameters. In this study, we formulate an analytical framework considering BEs to investigate and analyze the influence of shadowing environments on minimum node degree distribution (MNDD), node isolation probability (NIP), and $$\kappa$$-connectivity of a WMN deployed in a rectangular-shaped region (RSR). The proposed framework provides closed-form expressions for the MNDD and NIP. Simulation results validate the outcomes obtained through the proposed framework. We observed that the NIP increases and the $$\kappa$$-connectivity degrades severely with the rise in the shadowing effects' standard deviation.
... In comparison to real data, acquiring synthetic data is efficient and cost-effective. Due to this, the use of synthetic datasets to train machine learning models is increased in the past lustrum 21,[26][27][28][29] . ...
... We calculated the relevancy of the selected predictors in estimating the k-barriers by estimating each predictor's relative predictor importance score. To do so, we have used the regression tree ensemble technique 21,34 . It is an inbuilt class with a tree-based classifier that assigns a relative score for every predictor or attribute of the data. ...
... Predictor sensitivity. We have performed the sensitivity analysis of the predictors using Partial Dependence Plot (PDP) 21,35 . PDP depicts whether a model's predicted response (outcome) changes as a single explanatory variable varies. ...
Article
Full-text available
Momentous increase in the popularity of explainable machine learning models coupled with the dramatic increase in the use of synthetic data facilitates us to develop a cost-efficient machine learning model for fast intrusion detection and prevention at frontier areas using Wireless Sensor Networks (WSNs). The performance of any explainable machine learning model is driven by its hyperparameters. Several approaches have been developed and implemented successfully for optimising or tuning these hyperparameters for skillful predictions. However, the major drawback of these techniques, including the manual selection of the optimal hyperparameters, is that they depend highly on the problem and demand application-specific expertise. In this paper, we introduced Automated Machine Learning (AutoML) model to automatically select the machine learning model (among support vector regression, Gaussian process regression, binary decision tree, bagging ensemble learning, boosting ensemble learning, kernel regression, and linear regression model) and to automate the hyperparameters optimisation for accurate prediction of numbers of k-barriers for fast intrusion detection and prevention using Bayesian optimisation. To do so, we extracted four synthetic predictors, namely, area of the region, sensing range of the sensor, transmission range of the sensor, and the number of sensors using Monte Carlo simulation. We used 80% of the datasets to train the models and the remaining 20% for testing the performance of the trained model. We found that the Gaussian process regression performs prodigiously and outperforms all the other considered explainable machine learning models with correlation coefficient (R = 1), root mean square error (RMSE = 0.007), and bias = − 0.006. Further, we also tested the AutoML performance on a publicly available intrusion dataset, and we observed a similar performance. This study will help the researchers accurately predict the required number of k-barriers for fast intrusion detection and prevention.
... Typically, each sensor has a sensing unit, power unit, processing unit and an antenna to receive or transmit the sensed information. Sensors are inherently resource-constrained as they have a limited battery, limited storage capacity, and limited processing power due to their compact size [11,12]. WSNs possess considerable potential for monitoring large or complicated-to-reach areas such as forest fire monitoring [13,14]. ...
... , n) 2 Define objective function, Obj(x); (x = x 1 , x 2 , . . . , x d ) (from Equation (5) if (x j < X min ) then 11 x j ← X min 12 if (x j > X max ) then 13 x j ← X max 14 Calculate the fitness of new cuckoos, F j = Obj(x j ) 15 Choose a random nest (x k ) from existing n nests randomly 16 if (F j > F k ) then 17 x j ← x k F j ← F k 18 Find the current best solution out of n nests 19 for all x i do 20 Compute the mutation probability, P a 21 Generate a random number, P ∈ [0,1] 22 if (P < P a ) then 23 Generate new solution (cuckoo) (x rand ) randomly within the solution range, (X min to X max ) if (F rand > F i ) then 24 Replace the existing solution with the new random solution (cuckoo) 25 Find the global best solution of the iteration 26 Calculate differences of last three consecutive pairs of best solutions: ...
Article
Full-text available
Node localisation plays a critical role in setting up Wireless Sensor Networks (WSNs). A sensor in WSNs senses, processes and transmits the sensed information simultaneously. Along with the sensed information, it is crucial to have the positional information associated with the information source. A promising method to localise these randomly deployed sensors is to use bio-inspired meta-heuristic algorithms. In this way, a node localisation problem is converted to an optimisation problem. Afterwards, the optimisation problem is solved for an optimal solution by minimising the errors. Various bio-inspired algorithms, including the conventional Cuckoo Search (CS) and modified CS algorithm, have already been explored. However, these algorithms demand a predetermined number of iterations to reach the optimal solution, even when not required. In this way, they unnecessarily exploit the limited resources of the sensors resulting in a slow search process. This paper proposes an Enhanced Cuckoo Search (ECS) algorithm to minimise the Average Localisation Error (ALE) and the time taken to localise an unknown node. In this algorithm, we have implemented an Early Stopping (ES) mechanism, which improves the search process significantly by exiting the search loop whenever the optimal solution is reached. Further, we have evaluated the ECS algorithm and compared it with the modified CS algorithm. While doing so, note that the proposed algorithm localised all the localisable nodes in the network with an ALE of 0.5-0.8 m. In addition, the proposed algorithm also shows an 80% decrease in the average time taken to localise all the localisable nodes. Consequently, the performance of the proposed ECS algorithm makes it desirable to implement in practical scenarios for node localisation.
... The work presented in literatures 22,23) studied the impact of wireless channel environment on the coverage metric of WSNs and found that interference and multipath fading have severely deteriorating influence on the cov-erage metric of the deployed network. Nowadays, artificial intelligence and machine learning algorithms are being applied in the domain of WSNs 24,25,26) . In literature 25 , the authors have proposed three machine learning algorithms to predict the average localization error in the position of SNs at the time of deployment. ...
... Nowadays, artificial intelligence and machine learning algorithms are being applied in the domain of WSNs 24,25,26) . In literature 25 , the authors have proposed three machine learning algorithms to predict the average localization error in the position of SNs at the time of deployment. The authors found that their proposed algorithms are very efficient with Root Mean Square Error (RMSE = 0.147m). ...
Article
Coverage is one of the most critical performance metric of wireless sensor networks (WSNs) because it shows how well a region of interest (RoI) is being monitored by the deployed network. In general, the RoI is either circular or rectangular in shape which have boundary regions. Sensor nodes (SNs) deployed in these regions suffer boundary effects (BEs), i.e., the useful coverage area of an SN deployed near the boundary regions is less as compared to the SNs deployed in the middle of the RoI. It is imperative to consider these BEs while evaluating the performance of WSNs because analytical results derived for large networks are not valid for finite networks. In addition, SNs of a deployed WSN are prone to failure due to a large number of factors such as battery drainage, high temperature, and other environmental conditions. Earlier researchers have ignored the impact of BEs in the presence of sensor failure while evaluating the coverage performance of WSNs. In this work, we derive an analytical model by considering BEs and sensor failure to achieve a closed form expression for the k-coverage performance of a WSN deployed in a rectangular RoI. Further, we analyze the influence of various network parameters such as number of SNs, sensing range, and sensor failure rate on the k-coverage performance of the network. The results obtained using the proposed model show a good match with simulations outcomes with Root Mean Square Error (RMSE) no more than 0.03, thus, validating our model. For 𝑟𝑟𝑠𝑠 = 80 m and N = 100, 1-coverage probabilities are found to be 0.9874, 0.9313 and 0.8069 for k = 1, 2 and 3 respectively showing that the k-coverage probability deriorates with the increase in the value of k.
... This concept is known as The Internet of Things (IoT) by Kevin Ashton [1]. IoT is the physical object network that allows specific objects to gather and share data with computers, instruments, cars, buildings, and other items embedded with processors, circuitry, applications, sensors, and network connectivity [2][3][4]. ...
... The following Eqs. (2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) describe the grey wolf surrounding the prey [37]. ...
Article
Full-text available
The intrusion detection system is a method for detection against attacks, making it one of the essential defense layers. Researchers are trying to find new algorithms to inspect all inbound and outbound activities and identify suspicious patterns that may show an attempted system attack. The proposed technique for detecting intrusions uses the Grey Wolf Optimization (GWO) to solve feature selection problems and hybridizing it with Particle Swarm Optimization (PSO) to utilize the best value to update the information of each grey wolf position. This technique preserves the individual's best position information by the PSO algorithm, which prevents the GWO algorithm from falling into a local optimum. The NSL KDD dataset is used to verify the performance of the proposed technique. The classification is done using the k-means and SVM algorithms to measure the performance in terms of accuracy, detection rate, false alarm rate, number of features, and execution time. The results have shown that the proposed technique attained the necessary improvement of the GWO algorithm when using K-means or SVM algorithms.
... This can lead to accelerated convergence of the optimization process and helps to approach optimization problems in a more global manner. The main disadvantages of Gaussian processes, on the other hand, their high computational complexity of O(n 3 ) and their difficulties in higher dimensions, do not arise in the problem at hand (Yetilmezsoy et al. 2021;Akbari et al. 2019;Singh et al. 2021). ...
Article
Full-text available
The presented paper describes a shape optimization workflow using Bayesian strategies. It is applied to a novel automotive axle system consisting of leaf springs made from glass fiber reinforced plastics (GFRP). Besides the primary objectives of cost and mass reduction, the assembly has to meet multiple technical constraints with respect to various loading conditions. The related large-scale finite element model is fully parameterized by splines, hence the general shape of the guide curve as well as the spring’s height, width and material properties can be altered by the corresponding workflow. For this purpose, a novel method is developed to automatically generate high-quality meshes depending on the geometry of the respective springs. The size and complexity of the model demands the implementation of efficient optimization techniques with a preferably small number of required response function evaluations. Therefore, an existing optimization framework is extended by state-of-the-art Bayesian methods, including different kernel combinations and multiple acquisition function approaches, which are then tested, evaluated and compared. To properly address the use of GFRP as spring material in the objective function, an appropriate cost model is derived. Emerging challenges, such as conflicting targets regarding direct material costs and potential lightweight measures, are considered and investigated. The intermediate steps of the developed optimization procedure are tested on various sample functions and simplified models. The entire workflow is finally applied to the complete model and evaluated. Concluding, ideas and possibilities in improving the optimization process, such as the use of models with varying complexity, are discussed.
... Moreover, they do not require any pre-installed infrastructure for their operations [4]. As a result, they have a large number of civilians as well as defence applications like precision agriculture through soil moisture, landslide prediction, vehicle traffic monitoring, internet of things (IoT), healthcare, telecommunication, enemy tacking, battlefield surveillance, reconnaissance [5][6][7][8][9][10][11][12][13][14][15][16][17][18], and so on. ...
Article
Coverage is a crucial quality of service (QoS) parameter for wireless sensor networks (WSNs), which tells how effectively the deployed network monitors a given region. Analytical models available for the coverage analysis of finite WSNs are not scalable for large networks due to boundary effects (BEs).The effective coverage area (ECA) of a sensor node lying near the boundary regions is less than the ECA of a sensor node lying in the region's middle. Also, with the presence of obstacles in the transmission path and the radio irregularities, there are frequent changes in the wireless channel characteristics known as shadowing effects (SEs). Therefore, it becomes crucial to include BEs and SEs while investigating the coverage performance of WSNs. In this work, we analyze the-coverage performance of a WSN spread in a circular region of interest (RoI) by considering BEs and using a binary and a log-normal sensing range model.Furthermore, we also assess the effect of various network parameters viz.,the number of sensor nodes, maximum sensing range, and standard deviation of SEs on the-coverage of the WSN. Also, we compare the-coverage outcomes obtained by considering BEs with the results obtained by ignoring BEs. It is found that both BEs and SEs have a significant effect on the-coverage performance of the WSNs. The simulation outcomes substantiate analytical results and match upto a great extent.
... Then, the authors adopted the Pearson correlation coefficient (PCC) method to reduce the number of the feature temperature points to search for a more efficient and economic thermal error model for the spindle. GPR [35,36] has high robustness and accuracy and is easy to implement. It can adjust the hyperparameters by maximizing the marginal likelihood and by accurately optimizing them according to the value of the hyperparameters. ...
Article
Full-text available
Thermal error is one of the main sources of machining error of machine tools. Being a key component of the machine tool, the spindle will generate a lot of heat in the machining process and thereby result in a thermal error of itself. Real-time measurement of thermal error will interrupt the machining process. Therefore, this paper presents a machine learning model to estimate the thermal error of the spindle from its feature temperature points. The authors adopt random forests and Gaussian process regression to model the thermal error of the spindle and Pearson correlation coefficients to select the feature temperature points. The result shows that random forests collocating with Pearson correlation coefficients is an efficient and accurate method for the thermal error modeling of the spindle. Its accuracy reaches to 90.49% based on only four feature temperature points—two points at the bearings and two points at the inner housing—and the spindle speed. If the accuracy requirement is not very onerous, one can select just the temperature points of the bearings, because the installation of temperature sensors at these positions is acceptable for the spindle or machine tool manufacture, while the other positions may interfere with the cooling pipeline of the spindle.
... A wireless multihop network (WMN) is formed by spreading a hundred to thousands of miniature, lowpower, and low-cost sensor nodes (SNs) with identical sensing, computation, and communication capabilities [1]. These networks do not need any fixed infrastructure and operate in a decentralized and self-organized manner [2,3]. Therefore, WMNs have found a colossal number of applications such as battlefield surveillance, reconnaissance, enemy tracking, vehicle traffic monitoring, monitoring in underground coal mines, telecommunication, internet of things (IoT), environmental monitoring, structural health monitoring, industrial automation, health sector, to name a few, [4][5][6][7][8][9][10][11]. ...
Article
The performance of wireless multihop networks (WSNs) deployed in a finite region is affected by environmental phenomena such as shadowing and boundary effects. The shadowing effect affects the actual sensing range of sensor nodes due to varying amounts of signal path-losses, whereas the useful coverage area of a sensor node gets affected by its location within the deployed region-known as boundary effects. The existing analytical solutions for large networks are not conformable for finite networks as boundary effects distort the linearities between the network variables and cause non-linear behaviour of network dynamics. Earlier works have provided the analytical solution for the performance metrics of (WSNs) by considering either shadowing or boundary effects separately but considering these effects together is still non-existent. Therefore, it is imperative to study the impact of boundary and shadowing effects simultaneously while estimating the performance of a WMN as it indicates – how well the network monitors a given region? This work provides an analytical solution by considering boundary and shadowing effects simultaneously to compute the κ-coverage probability of a random location within a circular region. The proposed analytical solution is verified through exhaustive simulations with a root mean square error of less than or equal to 0.0189 between the analytical and simulation outcomes. This work has also analysed the influence of different network parameters on the network κ-coverage performance and concludes that the sensing range and the number of sensor nodes have positive, whereas the standard deviation of the shadowing effect has a negative impact on the coverage.
... This is considered a non-linear sensitivity analysis that disaggregates the averaging effects and evaluates the model at each instance. The average of all the ICE lines provides the PDP plot [94][95][96]. The averaging effect of PDP conceals any heterogeneous relationship present at any particular instance. ...
Article
Full-text available
We apply the Support Vector Regression (SVR) machine learning model to estimate surface roughness on a large alluvial fan of the Kosi River in the Himalayan Foreland from satellite images. To train the model, we used input features such as radar backscatter values in Vertical–Vertical (VV) and Vertical–Horizontal (VH) polarisation, incidence angle from Sentinel-1, Normalised Difference Vegetation Index (NDVI) from Sentinel-2, and surface elevation from Shuttle Radar Topographic Mission (SRTM). We generated additional features (VH/VV and VH–VV) through a linear data fusion of the existing features. For the training and validation of our model, we conducted a field campaign during 11–20 December 2019. We measured surface roughness at 78 different locations over the entire fan surface using an in-house-developed mechanical pin-profiler. We used the regression tree ensemble approach to assess the relative importance of individual input feature to predict the surface soil roughness from SVR model. We eliminated the irrelevant input features using an iterative backward elimination approach. We then performed feature sensitivity to evaluate the riskiness of the selected features. Finally, we applied the dimension reduction and scaling to minimise the data redundancy and bring them to a similar level. Based on these, we proposed five SVR methods (PCA-NS-SVR, PCA-CM-SVR, PCA-ZM-SVR, PCA-MM-SVR, and PCA-S-SVR). We trained and evaluated the performance of all variants of SVR with a 60:40 ratio using the input features and the in-situ surface roughness. We compared the performance of SVR models with six different benchmark machine learning models (i.e., Gaussian Process Regression (GPR), Generalised Regression Neural Network (GRNN), Binary Decision Tree (BDT), Bragging Ensemble Learning, Boosting Ensemble Learning, and Automated Machine Learning (AutoML)). We observed that the PCA-MM-SVR perform better with a coefficient of correlation (R = 0.74), Root Mean Square Error (RMSE = 0.16 cm), and Mean Square Error (MSE = 0.025 cm2). To ensure a fair selection of the machine learning model, we evaluated the Akaike’s Information Criterion (AIC), corrected AIC (AICc), and Bayesian Information Criterion (BIC). We observed that SVR exhibits the lowest values of AIC, corrected AIC, and BIC of all the other methods; this indicates the best goodness-of-fit. Eventually, we also compared the result of PCA-MM-SVR with the surface roughness estimated from different empirical and semi-empirical radar backscatter models. The accuracy of the PCA-MM-SVR model is better than the backscatter models. This study provides a robust approach to measure surface roughness at high spatial and temporal resolutions solely from the satellite data. Keywords: surface roughness; Sentinel-1; Sentinel-2; machine learning models; AutoML; backscatter models
... Despite the useful contributions of these authors, not much attention has been placed on the choice of kernels in GPR. The authors in [17] used three GPR models to show the efficacy of GPR in intrusion detection in wireless sensors without concern for time complexity. In particular, the three GPR models were implemented using squared exponential kernel. ...
... A wireless multihop network (WMN) is formed by spreading a hundred to thousands of miniature, low-power, and low-cost sensor nodes (SNs) with identical sensing, computation, and communication capabilities [1]. These networks do not need any fixed infrastructure and operate in a decentralized and selforganized manner [2,3]. Therefore, WMNs have found a colossal number of applications such as battlefield surveillance, reconnaissance, enemy tracking, vehicle traffic monitoring, monitoring in underground coal mines, telecommunication, internet of things (IoT), environmental monitoring, structural health monitoring, industrial automation, health sector, to name a few, [4][5][6][7][8][9][10]. ...
Preprint
Full-text available
The performance of wireless multihop networks deployed in a finite region is affected by environmental phenomena such as shadowing and boundary effects. The shadowing effect affects the actual sensing range of sensor nodes due to varying amounts of signal path-losses whereas the useful coverage area of a sensor node gets affected by its location within the deployed region-known as boundary effects. The existing analytical solutions for large networks are not conformable for finite networks as boundary effects distorts the linearities between the network variables and cause non-linear behaviour of network dynamics. Earlier works have provided the analytical solution for the performance metrics of wireless multihop networks by considering either shadowing or boundary effects separately but considering these effects together is still non-existent. Therefore, it is imperative to study the impact of boundary and shadowing effects simultaneously while estimating the performance of a wireless multihop network. Network coverage is one of the critical quality of service metric of a wireless multihop network indicating-how well the network monitors a given region of deployment? This work provides an analytical solution by considering boundary and shadowing effects simultaneously to compute the-coverage probability of a random location within a circular region. The proposed analytical solution is verified through exhaustive simulations with a root mean square error of less than or equal to 0.0189 between the analytical and simulation outcomes. This work has also analysed the influence of different network parameters on network-coverage performance and concludes that the sensing range and the number of sensor nodes have positive whereas the standard deviaton of shadowing effect has a negative impact on the coverage.
... Most of these threats are propagated through the Internet and other network typologies. However, many security mechanisms exist [1]- [4] to counter these threats, but the unpredictable actions of these threats become a nightmare for network engineers and security experts. In many cases, these external threats stand far ahead of the existing security mechanisms, such as firewalls [5]. ...
Article
Researchers are motivated to build effective Intrusion Detection Systems because of the implications of malicious actions in computing, communication, and cyber–physical systems (IDSs). In order to develop signature-based intrusion detection techniques that are suitable for use in cyber–physical environments, state-of-the-art supervised learning algorithms are devised. The main contribution of this research is the introduction of a signature-based intrusion detection model that is based on a hybrid Decision Table and Naive Bayes technique. In addition, the contribution of the suggested method is evaluated by comparing it to the existing literature in the field. In the preprocessing stage, Multi-Objective Evolutionary Feature Selection (MOEFS) feature selection has been used to select only five attack features from the recent CICIDS017 dataset. Keeping in view the class imbalance nature of CICIDS2017 dataset, adequate attack samples has been selected with more weightage to the attack classes having a smaller number of instances in the dataset. A hybrid of Decision Table and Naive Bayes models were combined to train and detect intrusions. Detection of botnets, port scans, Denial of Service (DoS)/Distributed Denial of Service (DDoS) attacks, such as Golden-Eye, Hulk, Slow httptest, slowloris, Heartbleed, Brute Force attacks, such as Patator (FTP), Patator (SSH), and Web attacks such as Infiltration, Web Brute Force, SQL Injection, and XSS, are all successfully detected by the proposed hybrid detection model. The proposed approach shows and accuracy 96.8% using five features of CICIDS2017, which is higher than the accuracy of methods discussed in the literatures.
... Bio-inspired algorithms [4,5] normally are mathematical models of animals and insects' social behaviour in a manner that leads problems to an optimal solution during iterations by specific number of populations in each generation. These algorithms could be employed in multiple optimization tasks, such as regression [6], clustering [7,9], feature selection [8], Minimum Spanning Tree (MST), Hub Location Allocation (HLA) and more. One of these bio-inspired algorithms which has highly efficiency is called Bees Algorithm (BA) [10,11,18]. ...
Conference Paper
Feature selection could be defined as an optimization problem and solved by bio-inspired algorithms. Bees Algorithm (BA) shows decent performance in feature selection optimization task. On the other hand, Local Phase Quantization (LPQ) is a frequency domain feature which has excellent performance on Depth images. Here, after extracting LPQ features out of RGB (colour) and Depth images from the Iranian Kinect Face Database (IKFDB), the Bees feature selection algorithm applies to select desired number of features for final classification tasks. IKFDB is recorded with Kinect sensor V.2 and contains colour and depth images for facial and facial micro-expressions recognition purposes. Here five facial expressions of Anger, Joy, Surprise, Disgust and Fear are used for final validation. The proposed Bees LPQ method is compared with Particle Swarm Optimization (PSO) LPQ, PCA LPQ, Lasso LPQ, and just LPQ features for classification tasks with Support Vector Machines (SVM), K-Nearest Neighbourhood (KNN), Shallow Neural Network and Ensemble Subspace KNN. Returned results, show a decent performance of the proposed algorithm (99 % accuracy) in comparison with others.
... Therefore, they can be deployed in almost no time in emergency situations and no man lands easily. All these features of WSN have [13752] made this technology popular and is being employed for an enormous number of military as well as civilian applications like smart border surveillance, intrusion detection, precision agriculture, disaster management, health monitoring, internet of things, forest fire detections, and so on [5][6][7][8][9][10][11]. ...
Article
Binary sensing range model does not imitate the true sensing characteristics of sensor nodes (SNs) as it ignores the environmental factors affecting the sensing abilities of SNs, thus, over estimates the coverage performance. To consider the influence of environmental factors in the sensing characteristics, Elfes sensing model was proposed which is a probabilistic sensing range model. Further, sensor failure is an important issue and may cause disrupted and declined coverage and connectivity. This paper investigates the coverage performance of WSN spread in a circular region (CR) employing both Elfes and binary sensing range models. The coverage metric obtained using the derived expression employing binary sensing range model provides higher coverage as compared to coverage achieved using Elfes model. Besides, we analysed the impact of different network parameters on network coverage and found that the sensing range and the SNs count have positive impact whereas sensor failure probability and required value of k have negative impact on the networks k-coverage metric.
... The use of machine learning to improve these identifications has been extensively discussed in the literature, thus making it a subject of great relevance and scope in science [17]. Authors have already proposed solutions to the problem using Gaussian processes [18], coverage probability in a mobile sensor network [19], low-power sensors inside the region of interest [20] and on-demand distance vector routing [21]. However, until the present moment, the interpretability of these sensors during identification has not been specified. ...
Article
Full-text available
Evolving fuzzy neural networks have the adaptive capacity to solve complex problems by interpreting them. This is due to the fact that this type of approach provides valuable insights that facilitate understanding the behavior of the problem being analyzed, because they can extract knowledge from a set of investigated data. Thus, this work proposes applying an evolving fuzzy neural network capable of solving data stream regression problems with considerable interpretability. The dataset is based on a necessary prediction of k barriers with wireless sensors to identify unauthorized persons entering a protected territory. Our method was empirically compared with state-of-the-art evolving methods, showing significantly lower RMSE values for separate test data sets and also lower accumulated mean absolute errors (MAEs) when evaluating the methods in a stream-based interleaved-predict-and-then-update procedure. In addition, the model could offer relevant information in terms of interpretable fuzzy rules, allowing an explainable evaluation of the regression problems contained in the data streams.
... It is countered by choosing a trust-based system [23,24]. Any IDS detects a suspicious activity with the help of any technique such as machine learning [25] and deep learning, which trains the system against various scenarios. Then with the help of the trained scenarios, any intrusion can be detected [26]. ...
Article
Full-text available
Wireless communication networks have much data to sense, process, and transmit. It tends to develop a security mechanism to care for these needs for such modern-day systems. An intrusion detection system (IDS) is a solution that has recently gained the researcher's attention with the application of deep learning techniques in IDS. In this paper, we propose an IDS model that uses a deep learning algorithm, conditional generative adversarial network (CGAN), enabling unsupervised learning in the model and adding an eXtreme gradient boosting (XGBoost) classifier for faster comparison and visualization of results. The proposed method can reduce the need to deploy extra sensors to generate fake data to fool the intruder 1.2-2.6%, as the proposed system generates this fake data. The parameters were selected to give optimal results to our model without significant alterations and complications. The model learns from its dataset samples with the multiple-layer network for a refined training process. We aimed that the proposed model could improve the accuracy and thus, decrease the false detection rate and obtain good precision in the cases of both the datasets, NSL-KDD and the CICIDS2017, which can be used as a detector for cyber intrusions. The false alarm rate of the proposed model decreases by about 1.827%.
... Moreover, external causes and dynamic changes in the wireless sensor networks affect cluster head selection, routing, delay, localization, quality of service, fault detection, coverage, reliability, and security [46][47][48][49][50][51]. Hence, the network may require a redesign, but the classical approaches for wireless sensor networks are programmed explicitly. ...
Article
Wireless Sensor Networks (WSNs) have attracted various academic researchers, engineers, science, and technology communities. This attraction is due to their broad research areas such as energy efficiency, data communication, coverage, connectivity, load balancing, security, reliability, scalability, and network lifetime. Researchers are looking towards cost-effective approaches to improve the existing solutions that reveal novel schemes, methods, concepts, protocols, and algorithms in the desired domain. Generally, review studies provide complete, easy access or solution to these concepts. Considering this as a driving force and the impact of clustering on the deterioration of energy consumption in wireless sensor networks, this review focus on clustering methods based on different aspects. This study’s significant contribution is to provide a brief review in the field of clustering in wireless sensor networks based on three different categories, such as classical, optimization, and machine learning techniques. For each of these categories, various performance metrics and parameters are provided, and a comparative assessment of the corresponding aspects like cluster head selection, routing protocols, reliability, security, and unequal clustering are discussed. Various advantages, limitations, applications of each method, research gaps, challenges, and research directions are considered in this study, motivating the researchers to carry out further research by providing relevant information in cluster-based wireless sensor networks.
... Especially, presented and adopted Greedy Randomized Adaptive Search Procedure (GRASP) metaheuristics model for improving the classification accuracy in CPS perception layers. Singh et al. [16] focused on faster prevention and detection of intrusion with an ML method based Gaussian Process Regression (GPR) technique. Also developed three models based feature scaling for precise estimation of k-barrier coverage possibility. ...
Article
Full-text available
Recently, cyber physical system (CPS) has gained significant attention which mainly depends upon an effective collaboration with computation and physical components. The greatly interrelated and united characteristics of CPS resulting in the development of cyber physical energy systems (CPES). At the same time, the rising ubiquity of wireless sensor networks (WSN) in several application areas makes it a vital part of the design of CPES. Since security and energy efficiency are the major challenging issues in CPES, this study offers an energy aware secure cyber physical systems with clustered wireless sensor networks using metaheuristic algorithms (EASCPSMA). The presented EASCPS-MA technique intends to attain lower energy utilization via clustering and security using intrusion detection. The EASCPSMA technique encompasses two main stages namely improved fruit fly optimization algorithm (IFFOA) based clustering and optimal deep stacked autoencoder (OSAE) based intrusion detection. Besides, the optimal selection of stacked autoencoder (SAE) parameters takes place using root mean square propagation (RMSProp) model. The extensive performance validation of the EASCPS-MA technique takes place and the results are inspected under varying aspects. The simulation results reported the improved effectiveness of the EASCPS-MA technique over other recent approaches interms of several measures.
Article
The rational distribution of public bicycle rental fleets is crucial for improving the efficiency of public bicycle programs. The accurate prediction of the demand for public bicycles is critical to improve bicycle utilization. To overcome the shortcomings of traditional algorithms such as low prediction accuracy and poor stability, using the 2011–2012 hourly bicycle rental data provided by the Washington City Bicycle Rental System, this study aims to develop an optimized and innovative public bicycle demand forecasting model based on grid search and eXtreme Gradient Boosting (XGBoost) algorithm. First, the feature ranking method based on machine learning models is used to analyze feature importance on the original data. In addition, a public bicycle demand forecast model is established based on important factors affecting bicycle utilization. Finally, to predict bicycle demand accurately, this study optimizes the model parameters through a grid search (GS) algorithm and builds a new prediction model based on the optimal parameters. The results show that the optimized XGBoost model based on the grid search algorithm can predict the bicycle demand more accurately than other models. The optimized model has an R-Squared of 0.947, and a root mean squared logarithmic error of 0.495. The results can be used for the effective management and reasonable dispatch of public bicycles.
Article
For an efficient detection of intrusion in the network, the system requires sensitive information. This information contains various features depending upon the network. The selection of useful features and relevant information from the data set has always been difficult for the user. These features bound the performance of the system and also restrict them in making accurate decisions. This irrelevant feature plays a major role in making confusion between the normal data pattern against the attack data pattern. This article proposes a novel Maximum correlation-based mutual information technique for efficient feature selection (MCMIFS) in a data network. The proposed method is utilized with Kernel Extreme Learning Machine (KELM) based multiclass classifier for effective intrusion detection. The performance of the proposed scheme is evaluated using standard intrusion detection datasets with feature ranging from 42 to 155. The hybrid of MCMIFS with KELM reveals the effectiveness of the proposed scheme by improving the accuracy of detection, i.e., 99.97%, decreasing the FPR, i.e., 0.19, by decreasing the computational complexity. The proposed Intrusion Detection Scheme (IDS) results are analyzed and compared with the other existing techniques.
Article
In wireless sensor networks (WSNs), energy efficiency is a significant design challenge that can be resolved by clustering and routing approaches. They are considered as Non-deterministic Polynomial (NP)-hard optimization problems, and the optimal or near-optimal solutions can be determined by using Swarm-Intelligence (SI) based algorithms. With this inspiration, this study focuses on two approaches, viz. the Hybrid Butterfly and Ant Colony optimization algorithm along with Static sink node (HBACS) and HBAC along with Mobile sink node (HBACM), which is a hybridization of Butterfly Optimization (BOA) and Ant Colony Optimization (ACO) algorithm. BOA determines the optimal cluster head, and ACO performs energy-efficient routing, thereby minimizing the energy consumption and maximizing the network’s lifetime. Furthermore, in this study, mobility of the sink node is used to eliminate the multi-hop communication between cluster heads and sink nodes, hence addressing the hot-spot issue and further extending the network lifetime. The proposed HBACS and HBACM approaches are implemented in the NS2 simulator. The simulation findings reveal that the HBACS shows percentage improvement regarding residual energy by 24.23%, 41.98%, and 66.67%; an improved number of alive nodes by 28.19%, 37.81%, and 53.12%; and improved throughput by 8.11%, 14.29%, and 17.65% over CRWO, ERP, and IHSBEER algorithms respectively. Moreover, the HBACM approach performs better in LDN by 18.76%, 63.66%, and 66.28%; HDN by 8.35%, 56.26%, and 58.15%, and FDN by 7.77%, 52.76%, and 74.29% over HGWSFO, SFO, and GWO based approaches, respectively.
Article
Full-text available
The evolution of the internet of things as a promising and modern technology has facilitated daily life. Its emergence was accompanied by challenges represented by its frequent exposure to attacks and its being a target for intruders who exploit the gaps in this technology in terms of the nature of its heterogeneous data and its large quantity. This made the study of cyber security an urgent necessity to monitor infrastructures It has network flaw detection and intrusion detection that helps protect the network by detecting attacks early and preventing them. As a result of advances in machine learning techniques, especially deep learning and its ability to self-learning and feature extraction with high accuracy, the research exploits deep learning to analyze the real data set of CSE-CIC-IDS2018 network traffic, which includes normal behavior and attacks, and evaluate our deep model long short-term memory (LSTM), That achieves accuracy of detection up to 99%.
Article
Nowadays, several kinds of attacks exist in cyberspace, and hence comprehensive research has been implemented to overcome these drawbacks. One such method to provide security in WSN (Wireless sensor network) is Intrusion Detection System. However, the determination of unknown attacks remains a major challenge in the intrusion detection system. Hence, the usage of deep learning methodologies remains to be an active area in cyber security. However, prevailing a deep learning algorithm possesses limitations such as comparatively low accuracy and heavily depends on the manual selection of the features. These problems have been analysed, and corresponding to the problems, the present study proposed an enhanced empirical-based component analysis to select relevant features. This proposed feature selection model integrates the advantages of both empirical mode decomposition and principal component analysis to retain most of the relevant features. The classification of the attack node with the selected features was performed with LSTM (Long Short Term Memory). The proposed framework validated datasets, namely NSL – KDD, CICIDS 2017, UNSW NB 2015, and KDD99 datasets, compared with the state of art methods. The comparative analysis with the prevailing methods proved the effectiveness of the presented system in terms of performance metrics such as accuracy, F1 score, Recall, FPR, FAR, etc.
Article
Full-text available
Security plays an important part in this Internet world because of the hasty improvement of Internet customers. Different Intrusion Detection Systems (IDS) have been advanced for various departments in history to describe and identify intruders utilizing data processing methods. Nonetheless, when using data processing, existing systems do not achieve adequate detection accuracy. For this reason, we suggest new IDS to offer preservation in statistics communications by completely describing intruders on wireless systems. Here, a new feature selection algorithm called enhanced conditional random field based feature selection to select the most contributed features and optimized hybrid deep neural network (OHDNN) is presented for the classification process. The hybrid deep neural network is a hybridization of convolution neural network (CNN) and long short-term memory (LSTM). To enhance the performance of the HDNN classifier, the parameters are optimized using adaptive golden eagle optimization. The performance of the presented approach is analyzed based on different metrics. For experimental analysis, the NSL-KDD and UNSW-NB15 datasets are used to compare its performance with other popular machine learning algorithms such as ANN, SVM, LSTM and CNN.
Article
In the recent days all the digital devices including devices used in Healthcare, Personal Digital Assistance etc. are connected to the network, consequently data is exposed to various types of attacks. Wide range of attacks has breached the public networks so far and still the number of attacks keeps increasing. Thus, intrusion detection systems (IDS) are developed to monitor and secure the data from such cyber-attacks. Every time the intruders use different approach to hack the networks. In order to effectively secure the networks from cyber-attacks, Machine Learning methodology is being adopted. The network data usually have large number of features and using all the features in identifying the attack will affect the performance of the IDS in terms of accuracy and processing speed. Hence there is a need for an effective IDS which adopts feature selection and classification methodology. In this paper, an enhanced Intrusion Detection scheme is suggested by employing the Greyish Wolf Optimizer (GWO) with LSTM based Recurrent Neural Networks (GWO LSTM). Further, the NSL KDD and UNSW-NB15 datasets are utilized to exhibit the overall performance of suggested scheme and also equate it along with other pre-existing strategies. The results clearly shows that our proposed method adds significant improvement in performance with accuracy of 96.1% and 97.4% for NSL-KDD and UNSW-NB15 datasets respectively.
Article
Full-text available
With the development of wireless sensor networks (WSNs), energy constraints and network security have become the main problems. This paper discusses the dynamic of the Susceptible, Infected, Low-energy, Susceptible model under pulse charging (SILS-P) in wireless rechargeable sensor networks. After the construction of the model, the local stability and global stability of the malware-free T-period solution of the model are analyzed, and the threshold R0 is obtained. Then, using the comparison theorem and Floquet theorem, we obtain the relationship between R0 and the stability. In order to make the conclusion more intuitive, we use simulation to reveal the impact of parameters on R0. In addition, the paper discusses the continuous charging model, and reveals its dynamic by simulation. Finally, the paper compares three charging strategies: pulse charging, continuous charging and non-charging and obtains the relationship between their threshold values and system parameters.
Article
Remote sensing image change detection is to analyze the change information of two images from the same area at different times. It has wide applications in urban expansion, forest detection, and natural disaster. In this paper, Feature Fusion Network is proposed to solve the problems of slow change detection speed and low accuracy. The MobileNetV3 block is adopted to efficiently extract features and a self-attention module is applied to investigate the relationship between heterogeneous feature maps (image features and concatenated features). The method is tested in data sets SZTAKI and LEVIR-CD. With 98.43 percentage correct classification, it is better than other comparative networks, and its space complexity is reduced by about 50%. The experimental results show that it has better performance and can improve the accuracy or speed of change detection.
Article
Full-text available
Node localisation is one of the significant concerns in Wireless Sensor Networks (WSNs). It is a process in which we estimate the coordinates of the unknown nodes using sensors with known coordinates called anchor nodes. Several bio-inspired algorithms have been proposed for accurate estimation of the unknown nodes. However, use of bio-inspired algorithms is a highly time-consuming process. Hence, finding optimal network parameters for node localisation during the network set-up process with the desired accuracy in a short time is still a challenging task. In this paper, we have proposed an efficient way to evaluate the optimal network parameters that result in low Average Localisation Error (ALE) using a machine learning approach based on Support Vector Regression (SVR) model. We have proposed three methods (S-SVR, Z-SVR and R-SVR) based on feature standardisation for fast and accurate prediction of ALE. We have considered the anchor ratio, transmission range, node density and iterations as features for training and prediction of ALE. These feature values are extracted from the modified Cuckoo Search (CS) simulations. In doing so, we found that all the methods perform exceptionally well with method R-SVR outperforming the other two methods with a correlation coefficient (R = 0.82) and Root Mean Square Error (RMSE = 0.147m).
Article
Full-text available
Intrusion detection is a very sensitive and major concern in wireless sensor networks. In border areas, sensors are installed to discover the presence of enemies or any trace passing in a prohibited area or any vicious moving object in the region of interest. In order to gain access to a region of interest, an intruder can take various paths, i.e., it can move along a straight line, follow a zigzag path or a curved path and move at a particular angle to cross the region without being detected and to improve its attacking ability. This paper formulates and analyzes $$\kappa$$-barrier coverage probability which acts as the intrusion detection probability for an intruder when the intruder follows different paths at different path angles with respect to the shortest path to cross the region of interest in a mobile sensor network. Furthermore, the effect of different network variables such as node density, sensing range, intrusion path angle and the ratio of sensor to intruder velocity on intrusion detection probability are also investigated. It is believed that the proposed model renders an effective tool to incorporate the intruder’s movement pattern in the design of an advanced finite wireless sensor network.
Article
Full-text available
Security is playing a major role in this Internet world due to the rapid growth of Internet users. The various intrusion detection systems were developed by many researchers in the past to identify and detect the intruders using data mining techniques. However, the existing systems are not able to achieve sufficient detection accuracy when using the data mining. For this purpose, we propose a new intrusion detection system to provide security in data communication by identifying and detecting the intruders effectively in wireless networks. Here, we propose a new feature selection algorithm called conditional random field and linear correlation coefficient-based feature selection algorithm to select the most contributed features and classify them using the existing convolutional neural network. The experiments have been conducted for evaluating the proposed intrusion detection system that achieves 98.88% as overall detection accuracy. The tenfold cross-validation has been done for evaluating the performance of the proposed model.
Article
Full-text available
Intrusion detection as one of the most important approaches to guarantee wireless sensing network security has been studied adequately in previous work. However, with the development of electronic anti-reconnaissance technology, the intruder may obtain the location information of detection nodes and perform path planning to avoid being detected. Such intruder is defined as an “empowered intruder” who will bring new challenges for traditional intrusion detection methods. Moreover, some subareas may have coverage holes due to random initial deployment of detection nodes, the desired effect of detection cannot be achieved. To address these issues, we propose a vehicle collaboration sensing network model, where mobile sensing vehicles and static sensor nodes cooperate to provide intrusion detection against empowered intruders. Our proposal (named as IDEI) consists of a target pursuit algorithm of mobile sensing vehicles and a sleep-scheduling strategy of static nodes. Mobile sensing vehicles will track the empowered intruder and fill up the coverage breaches, while static nodes follow a sleep-scheduling mechanism and will be awakened by detection nodes nearby when the intruder is detected. Simulation experiments are conducted to compare our proposal with existing methods such as KMsn and MTTA in terms of intrusion detection performance, energy consumption and moving distance of sensor nodes. The parameter sensitivity of IDEI is also studied with extensive simulations. The theoretical analysis and simulation results indicate that our proposal can achieve better efficiency and availability.
Article
Full-text available
Wireless sensor networks (WSNs) are growing rapidly in various fields of commerce, medicine, industrial, agriculture, research, meteorology, etc. that eases complicated tasks. The most active and recent research areas in wireless sensor networks are deployment strategies, energy efficiency and coverage. Besides energy harvesting, network lifetime of the sensors can be increased by decreasing the consumption of energy. This becomes the most challenging areas of utilizing wireless sensor network in practical applications. Deployment in WSNs directly influence the performance of the networks. The usage of sensor nodes in large quantity in the random deployment improves concerns in reliability and scalability. Coverage in wireless sensor networks measures how long the physical space is monitored by the sensors. Barrier coverage is an issue in wireless sensor networks, which is used for security application aims in intruder detection of the protected area. Several ongoing research work focuses on energy efficiency and coverage in wireless sensor networks and numerous schemes, algorithms, methods and architectures have been proposed. Still, there is no comprehensive solution applicable universally. Hence,this work provides with a state-of-the-art of the classification of wireless sensor networks based on different dimensions, such as, types of sensors, deployment strategies, sensing models, coverage and energy efficiency.
Article
Full-text available
This paper presents a generalized analytical model to estimate the commonly employed network performance metrics, viz., node isolation probability, minimum node degree distribution, and the probability of κ-connectivity in a wireless multi-hop network deployed in a finite regular rectangular region rather than a square region of limited scope. We derive and provide closed form expressions for the effective area covered by a node positioned in various sub regions of the rectangular region. Our derivation considers the boundary effects of the region that has significant impact on the effective coverage area by a node, which is important to accurately estimate the minimum node degree distribution, κ-connectivity and node isolation probability. We validate and compare our proposed analytical model by performing simulation to show that the results obtained through our proposition are more plausible and also conform to the specific cases of square region considered by other researchers. Later, we also discuss the tradeoffs between the transmission range of the nodes and the number of nodes that result in a fully κ-connected network with finite number of nodes deployed in a rectangular region of varying dimensions.
Conference Paper
Full-text available
WirelessSensor Network systems (WSNs) have received noteworthy research consideration because of their elite qualities like broad capacity to detect the physical world phenomenon and their broad scope of utilizations, extensive variety of uses due to low cost. One of the real requirements of WSN is lifetime of system because of its battery usage. To determine this issue to some degree is solved by creating various approaches on routing protocol to make energy efficient. EESR-ART2 approach is an endeavor to make organize vitality proficient. ART2 Neural Network Technique is actualized with Energy Efficient Sensor Routing (EESR) Protocol that observed to be a much viable outcome in recreation and upgrade the system lifetime regarding FND (first node dead) and HNA (half node alive) when contrasted with different protocols. Security is the major challenge faced by the existing WSN system. This paper focuses on intrusion detection in energy efficient sensor network using neuro-fuzzy approach. The proposed model discusses how anomalies detection scheme is improved using neuro-fuzzy approach.
Article
Full-text available
Providing reliable communications in a remotely monitored large-scale deployment of Wireless Sensor Networks is a challenging task. In this paper, we deal with such a deployment to monitor a landslide prone zone where the nodes sense geological attributes essential for early warning. The deployment area is a hilly mountain with different demographic characteristics. Establishing network connectivity in this deployment site for real-time streaming of sensor data involves dealing with site specific challenges such as asymmetric links, dynamic network conditions due to rough weather, inadequate solar power, network fail-over and re-connection problem etc. Our deployment makes use of only a few number of relay nodes for connecting the wireless sensor network to a field management center through an IoT gateway. The field management center makes use of multiple fault-tolerant WAN networks to relay the data to a remote central data management center for deep data analysis and for generating early warnings prior to a catastrophic event. This paper provides insights into our experiences from successful deployment of a WSN system and reports real-time measurements taken in the field for ensuring network reliability. Evaluation of our system reflects that the network is able to provide reliable communication by providing a fail-over for streaming data with less than a minute to re-establish the connection and the gateway software is capable of handling heterogeneous sensor readings at a rate up-to 1700/s within a latency of 10 s while delivering the data to the data center. The system also achieves highly accurate time synchronization in the order of microseconds throughout the network.
Article
Full-text available
With the wide application of wireless sensor networks in military and environmental monitoring, security issues have become increasingly prominent. Data exchanged over wireless sensor networks is vulnerable to malicious attacks due to the lack of physical defense equipment. Therefore, corresponding schemes of intrusion detection are urgently needed to defend against such attacks. Considering the serious class imbalance of the intrusion dataset, this paper proposes a method of using the synthetic minority oversampling technique (SMOTE) to balance the dataset and then uses the random forest algorithm to train the classifier for intrusion detection. The simulations are conducted on a benchmark intrusion dataset, and the accuracy of the random forest algorithm has reached 92.39%, which is higher than other comparison algorithms. After oversampling the minority samples, the accuracy of the random forest combined with the SMOTE has increased to 92.57%. This shows that the proposed algorithm provides an effective solution to solve the problem of class imbalance and improves the performance of intrusion detection.
Article
Full-text available
This paper presents the design of a wireless sensor network particularly designed for remote monitoring and control of industrial parameters. The article describes the network components, protocol and sensor deployment, aimed to accomplish industrial constraint and to assure reliability and low power consumption. A particular case of study is presented. The system consists of a base station, gas sensing nodes, a tree-based routing scheme for the wireless sensor nodes and a real-time monitoring application that operates from a remote computer and a mobile phone. The system assures that the industrial safety quality and the measurement and monitoring system achieves an efficient industrial monitoring operations. The robustness of the developed system and the security in the communications have been guaranteed both in hardware and software level. The system is flexible and can be adapted to different environments. The testing of the system confirms the feasibility of the proposed implementation and validates the functional requirements of the developed devices, the networking solution and the power consumption management.
Article
Full-text available
Building performance simulations (BPS) are used to test different designs and systems with the intention of reducing building costs and energy demand while ensuring a comfortable indoor climate. Unfortunately, software for BPS is computationally intensive. This makes it impractical to run thousands of simulations for sensitivity analysis and optimization. Worse yet, millions of simulations may be necessary for a thorough exploration of the high-dimensional design space formed by the many design parameters. This computational issue may be overcome by the creation of fast metamodels. In this paper, we aim to find suitable metamodeling techniques for diverse outputs from BPS. We consider five indicators of building performance and eight test problems for the comparison six popular metamodeling techniques – linear regression with ordinary least squares (OLS), random forest (RF), support vector regression (SVR), multivariate adaptive regression splines, Gaussian process regression (GPR), and neural network (NN). The methods are compared with respect to accuracy, efficiency, ease-of-use, robustness, and interpretability. To conduct a fair and in-depth comparison, a methodological approach is pursued using exhaustive grid searches for model selection assisted by sensitivity analysis. The comparison shows that GPR produces the most accurate metamodels, followed by NN and MARS. GPR is robust and easy to implement but becomes inefficient for large training sets compared to NN and MARS. A coefficient of determination, R², larger than 0.9 have been obtained for the BPS outputs using between 128 and 1024 training points. In contrast, accurate metamodels with R² values larger than 0.99 can be achieved for all eight test problems using only 32–256 training points.
Article
Full-text available
Wireless technology development has increased rapidly due to it’s convenience and cost effectiveness compared to wired applications, particularly considering the advantages offered by Wireless Sensor Network (WSN) based applications. Such applications exist in several domains including healthcare, medical, industrial and home automation. In the present study, a home-based wireless ECG monitoring system using Zigbee technology is considered. Such systems can be useful for monitoring people in their own home as well as for periodic monitoring by physicians for appropriate healthcare, allowing people to live in their home for longer. Health monitoring systems can continuously monitor many physiological signals and offer further analysis and interpretation. The characteristics and drawbacks of these systems may affect the wearer’s mobility during monitoring the vital signs. Real-time monitoring systems record, measure, and monitor the heart electrical activity while maintaining the consumer’s comfort. Zigbee devices can offer low-power, small size, and a low-cost suitable solution for monitoring the ECG signal in the home, but such systems are often designed in isolation, with no consideration of existing home control networks and smart home solutions. The present study offers a state of the art review and then introduces the main concepts and contents of the wireless ECG monitoring systems. In addition, models of the ECG signal and the power consumption formulas are highlighted. Challenges and future perspectives are also reported. The paper concludes that such mass-market health monitoring systems will only be prevalent when implemented together with home environmental monitoring and control systems.
Article
Full-text available
Wireless Sensor Networks have profound applications in military systems. Intrusion at unmanned borders and at other sensitive places may be tracked using sensor networks. However, the domain of military applications could imply hostile environment and thus monitoring of the nodes of the deployed WSN could be practically impossible. It is thus required that the deployed WSN ensures low energy consumption to give a high network life such that the cost of deployment may be suitably amortized. In this paper we propose an energy efficient solution for detecting intrusions through unmanned borders and other sensitive places with prolonged network lifetime using two routing schemes: KPS and Loop Free (LF)-KPS. We have compared these two schemes with LEACH and TEEN, and have shown how data transfer through KPS and LF-KPS protocols would ensure an enhanced lifetime for the deployed network. At the end we have also shown the effect of looping on the lifetime of the deployed network.
Article
Full-text available
Wireless sensor networks (WSNs) can be used in agriculture to provide farmers with a large amount of information. Precision agriculture (PA) is a management strategy that employs information technology to improve quality and production. Utilizing wireless sensor technologies and management tools can lead to a highly effective, green agriculture. Based on PA management, the same routine to a crop regardless of site environments can be avoided. From several perspectives, field management can improve PA, including the provision of adequate nutrients for crops and the wastage of pesticides for the effective control of weeds, pests, and diseases. This review outlines the recent applications of WSNs in agriculture research as well as classifies and compares various wireless communication protocols, the taxonomy of energy-efficient and energy harvesting techniques for WSNs that can be used in agricultural monitoring systems, and comparison between early research works on agriculture-based WSNs. The challenges and limitations of WSNs in the agricultural domain are explored, and several power reduction and agricultural management techniques for long-term monitoring are highlighted. These approaches may also increase the number of opportunities for processing Internet of Things (IoT) data.
Article
Full-text available
Structural health monitoring (SHM) using wireless sensor networks (WSNs) has gained research interest due to its ability to reduce the costs associated with the installation and maintenance of SHM systems. SHM systems have been used to monitor critical infrastructure such as bridges, high-rise buildings, and stadiums and has the potential to improve structure lifespan and improve public safety. The high data collection rate of WSNs for SHM pose unique network design challenges. This paper presents a comprehensive survey of SHM using WSNs outlining the algorithms used in damage detection and localization, outlining network design challenges and future research directions. Solutions to network design problems such as scalability, time synchronization, sensor placement, and data processing are compared and discussed. This survey also provides an overview of testbeds and real-world deployments of WSNs for SH.
Article
Full-text available
Signal strength-based positioning in wireless sensor networks is a key technology for seamless, ubiquitous localization, especially in areas where Global Navigation Satellite System (GNSS) signals propagate poorly. To enable wireless local area network (WLAN) location fingerprinting in larger areas while maintaining accuracy, methods to reduce the effort of radio map creation must be consolidated and automatized. Gaussian process regression has been applied to overcome this issue, also with auspicious results, but the fit of the model was never thoroughly assessed. Instead, most studies trained a readily available model, relying on the zero mean and squared exponential covariance function, without further scrutinization. This paper studies the Gaussian process regression model selection for WLAN fingerprinting in indoor and outdoor environments. We train several models for indoor/outdoor- and combined areas; we evaluate them quantitatively and compare them by means of adequate model measures, hence assessing the fit of these models directly. To illuminate the quality of the model fit, the residuals of the proposed model are investigated, as well. Comparative experiments on the positioning performance verify and conclude the model selection. In this way, we show that the standard model is not the most appropriate, discuss alternatives and present our best candidate.
Article
Full-text available
The synergy of computational and physical network components leading to the Internet of Things, Data and Services has been made feasible by the use of Cyber Physical Systems (CPSs). CPS engineering promises to impact system condition monitoring for a diverse range of fields from healthcare, manufacturing, and transportation to aerospace and warfare. CPS for environment monitoring applications completely transforms human-to-human, human-to-machine and machine-to-machine interactions with the use of Internet Cloud. A recent trend is to gain assistance from mergers between virtual networking and physical actuation to reliably perform all conventional and complex sensing and communication tasks. Oil and gas pipeline monitoring provides a novel example of the benefits of CPS, providing a reliable remote monitoring platform to leverage environment, strategic and economic benefits. In this paper, we evaluate the applications and technical requirements for seamlessly integrating CPS with sensor network plane from a reliability perspective and review the strategies for communicating information between remote monitoring sites and the widely deployed sensor nodes. Related challenges and issues in network architecture design and relevant protocols are also provided with classification. This is supported by a case study on implementing reliable monitoring of oil and gas pipeline installations. Network parameters like node-discovery, node-mobility, data security, link connectivity, data aggregation, information knowledge discovery and quality of service provisioning have been reviewed.
Article
Full-text available
The intrusion detection application in a homogeneous wireless sensor network is defined as a mechanism to detect unauthorized intrusions or anomalous moving attackers in a field of interest. The quality of deterministic sensor nodes deployment can be determined sufficiently by a rigorous analysis before the deployment. However, when random deployment is required, determining the deployment quality becomes challenging. An area may require that multiple nodes monitor each point from the sensing area; this constraint is known as k-coverage where k is the number of nodes. The deployment quality of sensor nodes depends directly on node density and sensing range; mainly a random sensor nodes deployment is required. The major question is centred around the problem of network coverage, how can we guarantee that each point of the sensing area is covered by the required number of sensor nodes and what a sufficient condition to guarantee the network coverage? To deal with this, probabilistic intrusion detection models are adopted, called single/multi-sensing detection, and the deployment quality issue is surveyed and analysed in terms of coverage. We evaluate the capability of our probabilistic model in homogeneous wireless sensor network, in terms of sensing range, node density, and intrusion distance.
Conference Paper
Full-text available
Distributed prognostics architecture design is an enabling step for efficient implementation of health management systems. <sup>12</sup>A major challenge encountered in such design is formulation of optimal distributed prognostics algorithms. In this paper, we present a distributed GPR based prognostics algorithm whose target platform is a wireless sensor network. In addition to challenges encountered in a distributed implementation, a wireless network poses constraints on communication patterns, thereby making the problem more challenging. The prognostics application that was used to demonstrate our new algorithms is battery prognostics. In order to present trade-offs within different prognostic approaches, we present comparison with the distributed implementation of a particle filter based prognostics for the same battery data.
Conference Paper
Full-text available
Previous work on the coverage of mobile sensor networks focuses on algorithms to reposition sensors in order to achieve a static configuration with an enlarged covered area. In this paper, we study the dynamic aspects of the coverage of a mobile sensor network that depend on the process of sensor movement. As time goes by, a position is more likely to be covered; targets that might never be detected in a static sensor network can now be detected by moving sensors. We characterize the area coverage at specific time instants and during time intervals, as well as the time it takes to detect a randomly located static target. Our results show that sensor mobility can be exploited to compensate for the lack of sensors and improve the network coverage. For mobile targets, we take a game theoretic approach and derive optimal mobility strategies for sensors and the targets from their own perspectives.
Conference Paper
Full-text available
In this paper we address the problem of producing a high-resolution image from a single low-resolution image without any external training set. We propose a framework for both magnification and deblurring using only the original low-resolution image and its blurred version. In our method, each pixel is predicted by its neighbors through the Gaussian process regression. We show that when using a proper covariance function, the Gaussian process regression can perform soft clustering of pixels based on their local structures. We further demonstrate that our algorithm can extract adequate information contained in a single low-resolution image to generate a high-resolution image with sharp edges, which is comparable to or even superior in quality to the performance of other edge-directed and example-based super-resolution algorithms. Experimental results also show that our approach maintains highquality performance at large magnifications.
Article
Full-text available
The likelihood is defined for a state space model with incompletely specified initial conditions by transforming the data to eliminate the dependence on the unspecified conditions. This approach is extended to obtain estimates of the state vectors and predictors and interpolators for missing observations. It is then shown that this method is equivalent to placing a diffuse prior distribution on the unspecified part of the initial state vector, and modified versions of the Kalman filter and smoothing algorithms are derived to give exact numerical procedures for diffuse initial conditions. The results are extended to continuous time models, including smoothing splines and continuous time autoregressive processes.
Article
Full-text available
In the performance evaluation of a protocol for an ad hoc network, the protocol should be tested under realistic conditions including, but not limited to, a sensible transmission range, limited buffer space for the storage of messages, representative data traffic models, and realistic movements of the mobile users (i.e., a mobility model). This paper is a survey of mobility models that are used in the simulations of ad hoc networks. We describe several mobility models that represent mobile nodes whose movements are independent of each other (i.e., entity mobility models) and several mobility models that represent mobile nodes whose movements are dependent on each other (i.e., group mobility models). The goal of this paper is to present a number of mobility models in order to offer researchers more informed choices when they are deciding upon a mobility model to use in their performance evaluations. Lastly, we present simulation results that illustrate the importance of choosing a mobility model in the simulation of an ad hoc network protocol. Specifically, we illustrate how the performance results of an ad hoc network protocol drastically change as a result of changing the mobility model simulated.
Article
In order to solve the critical issues in Wireless Sensor Networks (WSNs), with concern for limited sensor lifetime, nature-inspired algorithms are emerging as a suitable method. Getting optimal network coverage is one of those challenging issues that need to be examined critically before any network setup. Optimal network coverage not only minimizes the consumption of limited energy of battery driven sensors but also reduce the sensing of redundant information. In this paper, we focus on nature-inspired optimization algorithms concerning the optimal coverage in WSNs. In the first half of the paper, we have briefly discussed the taxonomy of the optimization algorithms along with the problem domains in WSNs. In the second half of the paper, we have compared the performance of two nature-inspired algorithms for getting optimal coverage in WSNs. The first one is a combined Improved Genetic Algorithm and Binary Ant Colony Algorithm (IGA-BACA), and the second one is Lion Optimization (LO). The simulation results confirm that LO gives better network coverage, and the convergence rate of LO is faster than that of IGA-BACA. Further, we observed that the optimal coverage is achieved at a lesser number of generations in LO as compared to IGA-BACA. This review will help researchers to explore the applications in this field as well as beyond this area.
Article
Machine Learning is capable of providing real-time solutions that maximize the utilization of resources in the network thereby increasing the lifetime of the network. It is able to process automatically without being externally programmed thus making the process more easy, efficient, cost-effective, and reliable. ML algorithms can handle complex data more quickly and accurately. Machine Learning is used to enhance the ability of the Wireless Sensor Network environment. Wireless Sensor Networks (WSN) is a combination of several networks and it is decentralized and distributed in nature. WSN consists of sensor nodes and sinks nodes which have a property of self-organizing and self-healing. WSN is used in other applications, such as biodiversity and ecosystem protection, surveillance, climate change tracking, and other military applications.Now-a-days, a huge development is seen in WSNs due to the advancement of electronics and wireless communication technologies, several drawbacks like low computational capacity, small memory, and limited energy resources infrastructure needs physical vulnerability to require source measures where privacy plays a key role.WSN is used to monitor the dynamic environments and to adapt to such situation sensor networks need Machine Learning techniques to avoid unnecessary redesign. Machine learning techniques survey for WSNs provide a wide range of applications in which security is given top priority. To secure data from attackers the WSNs system should be able to delete the instruction if any hackers/attackers are trying to steal data.
Article
Intrusion detection is the identification of unauthorized access of a computer network. This paper proposes a novel algorithm for a network intrusion detection system (NIDS) using an improved feature subset selected directly by a genetic algorithm (GA)-based exhaustive search and fuzzy C-means clustering (FCM). The algorithm identifies the bagging (BG) classifier and the convolutional neural network (CNN) model as an effective extractor by implementing the GA in combination with 5-fold cross validation (CV) to select the CNN model structure. The deep feature subset extracted by the selected CNN model is put into the BG classifier to validate the performance with the 5-fold CV. The high quality feature set obtained by the three-layered feature construction using the GA, FCM, CNN extractor, and a hybrid CNN and BG learning method significantly improves the final detection performance. Moreover, the highly reliable validation performance results achieved by the 5-fold CV procedure for the proposed algorithm imply a well-fitted application in a practical computer network environment NIDS.
Article
Wireless sensor networks are widely applicable in surveillance applications for intruder detection. Many research articles in the literature have focused on to generate an inference about the intruder detection and evaluated this in terms of detection probability and false alarm rate. In this paper, we have solved the problem of passive mobile intruder detection using two modalities i.e. acoustic signal model and sensing probability model. A three-level hierarchy is proposed to generate an inference about the presence of a mobile intruder. At base level the deployed sensor nodes are grouped together using k-mean clustering. For both modalities the distance between the sensor nodes and intruder is computed. At cluster head, the received signal strengths or sensing probabilities reported by the sensor nodes are employed for binary hypothesis testing. The cluster heads transmit their decisions to the fusion center for the generation of accurate inference about the intruder detection after performing Likelihood Ratio Test (LRT) on the fused decisions. Numerical analysis of received signals identify the optimum threshold value for detection probability computation. The derived fusion rule maximizes detection probability with respect to the allowable false alarm rates. The accuracy of proposed fusion rule is depending on the number of actual detections reported by the deployed sensor nodes. Simulation results also show that the derived fusion rule performs better as compared to other fusion rules in terms of sensor count, speed of mobile intruder, detection probability and detection accuracy.
Article
The volume of network and Internet traffic is expanding daily, with data being created at the zettabyte to petabyte scale at an exceptionally high rate. These can be characterized as big data, because they are large in volume, variety, velocity, and veracity. Security threats to networks, the Internet, websites, and organizations are growing alongside this growth in usage. Detecting intrusions in such a big data environment is difficult. Various intrusion-detection systems (IDSs) using artificial intelligence or machine learning have been proposed for different types of network attacks, but most of these systems either cannot recognize unknown attacks or cannot respond to such attacks in real time. Deep learning models, recently applied to large-scale big data analysis, have shown remarkable performance in general but have not been examined for detection of intrusions in a big data environment. This paper proposes a hybrid deep learning model to efficiently detect network intrusions based on a convolutional neural network (CNN) and a weight-dropped, long short-term memory (WDLSTM) network. We use the deep CNN to extract meaningful features from IDS big data and WDLSTM to retain long-term dependencies among extracted features to prevent overfitting on recurrent connections. The proposed hybrid method was compared with traditional approaches in terms of performance on a publicly available dataset, demonstrating its satisfactory performance.
Article
Wireless sensor networks (WSNs) found application in many diverse fields, starting from environment monitoring to machine health monitoring. The sensor in WSNs senses information. Sensing and transmitting this information consume most of the energy. Also, this information requires proper processing before final usages. This paper deals with minimising the redundant information sensed by the sensors in WSNs to reduce the unnecessary energy consumption and prolong the network lifetime. The redundant information is expressed in terms of the overlapping sensing area of the working sensors set. A mathematical model is proposed to find the redundant information in terms of the overlapping area. A combined meta-heuristic approach is used to achieve the optimal coverage, and the effect of the overlapping area is considered in the objective function to reduce the amount of redundant information sensed by the working sensors set. Improved genetic algorithm (IGA) and Binary ant colony algorithm (BACA) are used as heuristic tools to optimise the multi-objective function. The objective was to find the minimum number of sensors that cover a complete scenario with minimum overlapping sensing region. The results show that optimal coverage with the minimum working sensor set is achieved and then by incorporating the concept of overlapping area in the objective function, sensing of redundant information is further reduced.
Article
Nickel-based superalloys, used extensively in advanced gas turbine engines, exhibit complex microstructures that evolve during exposure to high temperatures (i.e., aging treatments). In this work, we examine critically if the principal component (PC) representation of rotationally invariant 2-point spatial correlations can adequately capture the salient features of the microstructure evolution in the thermal aging of the superalloys. For this purpose, an experimental study involving microstructure characterization of 27 differently aged (i.e., different combinations of temperature and time of exposure) samples was designed and conducted. Of these, 23 samples were employed for training a Gaussian Process Regression (GPR) model that took the aging temperature and the aging time as inputs, and predicted the microstructure statistics as output. The viability of the approach described above was evaluated critically by comparing the predictions for the four samples that were not used in the training of the GPR model. Furthermore, a new strategy was developed and implemented to generate digital microstructures corresponding to the predicted microstructure statistics. The predicted microstructures were found to be in good agreement with the experimentally measured one, validating the novel framework presented in this work.
Article
In this letter, we present a comprehensive analysis of the use of machine and deep learning solutions for IDS systems in Wireless Sensor Networks (WSNs). To accomplish this, we introduce Restricted Boltzmann Machine-based Clustered IDS (RBC-IDS), a potential deep learning-based IDS methodology for monitoring critical infrastructures by WSNs. We study the performance of RBC-IDS, and compare it to the previously proposed adaptive machine learning-based IDS: The Adaptively Supervised and Clustered Hybrid IDS (ASCH-IDS). Numerical results show that RBC-IDS and ASCH-IDS achieve the same detection and accuracy rates, though the detection time of RBC-IDS is approximately twice that of ASCH-IDS.
Article
An effective security strategy for Wireless Sensor Networks (WSNs) is imperative to counteract security threats. Meanwhile, energy consumption directly affects the network lifetime of a wireless sensor. Thus, an attempt to exploit a low-consumption Intrusion Detection System (IDS) to detect malicious attacks makes a lot of sense. Existing Intrusion Detection Systems can only detect specific attacks and their network lifetime is short due to their high energy consumption. For the purpose of reducing energy consumption and ensuring high efficiency, this paper proposes an intrusion detection model based on game theory and an autoregressive model. The paper not only improves the autoregressive theory model into a non-cooperative, complete-information, static game model, but also predicts attack pattern reliably. The proposed approach improves on previous approaches in two main ways: (1) it takes energy consumption of the intrusion detection process into account, and (2) it obtains the optimal defense strategy that balances the system's detection efficiency and energy consumption by analyzing the model's mixed Nash equilibrium solution. In the simulation experiment, the running time of the process is regarded as the main indicator of energy consumption of the system. The simulation results show that our proposed IDS not only effectively predicts the attack time and the next targeted cluster based on the game theory, but also reduces energy consumption.
Article
Function estimation/approximation is viewed from the perspective of numerical optimization iti function space, rather than parameter space. A connection is made between stagewise additive expansions and steepest-descent minimization. A general gradient descent "boosting" paradigm is developed for additive expansions based on any fitting criterion. Specific algorithms are presented for least-squares, least absolute deviation, and Huber-M loss functions for regression, and multiclass logistic likelihood for classification. Special enhancements are derived for the particular case where the individual additive components are regression trees, and tools for interpreting such "TreeBoost" models are presented. Gradient boosting of regression trees produces competitives highly robust, interpretable procedures for both regression and classification, especially appropriate for mining less than clean data. Connections between this approach and the boosting methods of Freund and Shapire and Friedman, Hastie and Tibshirani are discussed.
Article
The optimal design problem is tackled in the framework of a new model and new objectives. A regression model is proposed in which the regression function is permitted to take any form over the space of independent variables. The design objective is based on fitting a simplified function for prediction. The approach is Bayesian throughout. The new designs are more robust than conventional ones. They also avoid the need to limit artificially design points to a predetermined subset of . New solutions are also offered for the problems of smoothing, curve fitting and the selection of regressor variables.
Article
Barrier coverage problem in mobile sensor networks has been an interesting research issue. Existing solutions are mainly concerned with deciding one-time movement for individual sensors to construct as many barriers as possible, which may not be suitable when there are no sufficient sensors to form a single barrier. In this paper, we aim to achieve barrier coverage in sensor scarcity scenario by dynamic sensor patrolling. In specific, we design a periodic monitoring scheduling (PMS) algorithm. Based on the insight from PMS, we then propose a coordinated sensor patrolling (CSP) algorithm to further improve the barrier coverage. By jointly exploiting sensor mobility and intruder arrival information, CSP is able to significantly enhance barrier coverage. We prove that the total distance the sensors move during each time slot in CSP is the minimum. Considering the decentralized nature of mobile sensor networks, we further introduce two distributed versions of CSP: S-DCSP and G-DCSP. We study the scenario where sensors are moving on two barriers and propose two heuristic algorithms to guide the movement of sensors. Through extensive simulations, we demonstrate that CSP has a desired barrier coverage performance and S-DCSP and G-DCSP have similar performance as that of CSP.
Article
This paper presents a method of using Gaussian process regression to model spatial functions for mobile wireless sensor networks. A distributed Gaussian process regression (DGPR) approach is developed by using a sparse Gaussian process regression method and a compactly supported covariance function. The resultant formulation of the DGPR approach only requires neighbor-to-neighbor communication, which enables each sensor node within a network to produce the regression result independently. The collective motion control is implemented by using a locational optimization algorithm, which utilizes the information entropy from the DGPR result. The collective mobility of sensor networks plus the online learning capability of the DGPR approach also enables the mobile sensor network to adapt to spatiotemporal functions. Simulation results are provided to show the performance of the proposed approach in modeling stationary spatial functions and spatiotemporal functions.
Article
This paper proposes a new variational approximation for infinite mixtures of Gaussian processes. As an extension of the single Gaussian process regression model, mixtures of Gaussian processes can characterize varying covariances or multimodal data and reduce the deficiency of the computationally cubic complexity of the single Gaussian process model. The infinite mixture of Gaussian processes further integrates a Dirichlet process prior to allowing the number of mixture components to automatically be determined from data. We use variational inference and a truncated stick-breaking representation of the Dirichlet process to approximate the posterior of hidden variables involved in the model. To fix the hyperparameters of the model, the variational EM algorithm and a greedy algorithm are employed. In addition to presenting the variational infinite-mixture model, we apply it to the problem of traffic flow prediction. Experiments with comparisons to other approaches show the effectiveness of the proposed model.
Article
We provide a new unifying view, including all existing proper probabilistic sparse approximations for Gaussian process regression. Our approach relies on expressing the effective prior which the methods are using. This allows new insights to be gained, and highlights the relationship between existing methods. It also allows for a clear theoretically justified ranking of the closeness of the known approximations to the corresponding full GPs. Finally we point directly to designs of new better sparse approximations, combining the best of the existing strategies, within attractive computational constraints.
Article
The partially improper prior behind the smoothing spline model is used to obtain a generalization of the maximum likelihood (GML) estimate for the smoothing parameter. Then this estimate is compared with the generalized cross validation (GCV) estimate both analytically and by Monte Carlo methods. The comparison is based on a predictive mean square error criteria. It is shown that if the true, unknown function being estimated is smooth in a sense to be defined then the GML estimate undersmooths relative to the GCV estimate and the predictive mean square error using the GML estimate goes to zero at a slower rate than the mean square error using the GCV estimate. If the true function is "rough" then the GCV and GML estimates have asymptotically similar behavior. A Monte Carlo experiment was designed to see if the asymptotic results in the smooth case were evident in small sample sizes. Mixed results were obtained for $n = 32$, GCV was somewhat better than GML for $n = 64$, and GCV was decidedly superior for $n = 128$. In the $n = 32$ case GCV was better for smaller $\sigma^2$ and the comparison close for larger $\sigma^2$. The theoretical results are shown to extend to the generalized spline smoothing model, which includes the estimate of functions given noisy values of various integrals of them.
Article
Function approximation is viewed from the perspective of numerical optimization in function space, rather than parameter space. A connection is made between stagewise additive expansions and steepest--descent minimization. A general gradient--descent "boosting" paradigm is developed for additive expansions based on any fitting criterion. Specific algorithms are presented for least--squares, least--absolute--deviation, and Huber--M loss functions for regression, and multi--class logistic likelihood for classification. Special enhancements are derived for the particular case where the individual additive components are decision trees, and tools for interpreting such "TreeBoost" models are presented. Gradient boosting of decision trees produces competitive, highly robust, interpretable procedures for regression and classification, especially appropriate for mining less than clean data. Connections between this approach and the boosting methods of Freund and Shapire 1996, and Fr...
The intrusion detection in mobile sensor network
• G Y Keung
• B Li
• Q Zhang
Keung, G. Y., Li, B., & Zhang, Q. (2012). The intrusion detection in mobile sensor network. IEEE/ACM Transactions on Networking, 20, 1152-1161.
Intrusion detection using dynamic feature selection and fuzzy temporal decision tree classification for wireless sensor networks
• P Nancy
• S Muthurajkumar
• S Ganapathy
• S S Kumar
• M Selvi
• K Arputharaj
Nancy, P., Muthurajkumar, S., Ganapathy, S., Kumar, S. S., Selvi, M., & Arputharaj, K. (2020). Intrusion detection using dynamic feature selection and fuzzy temporal decision tree classification for wireless sensor networks. IET Communications, 14, 888-895.
Efficient deployment quality analysis for intrusion detection in wireless sensor networks