Concurrency and Computation: Practice and Experience

Wang, Jin

doi: 10.1002/cpe.7002pmid: N/A

Owing to the limited computing capability of edge cloudlets, use of cloud platforms to provide computation‐intensive services as an assistant should be studied. Offloading, among the key techniques in the cloud‐edge environment, is plagued by the serious problem of privacy disclosure. In addition, the edge does not have access control of terminals, and the controller working the edge that makes offloading decisions, must not be allowed access to any information regarding the contents of the terminal. Thus, this study introduced a system architecture wherein the data was separated from the control plane to minimize systemic risk and optimize the offloading process towards makespan and cost reduction. Traditional methods of privacy preservation are dependent on encryption of total or content awareness. Edge privacy preservation renders maintaining edge delay advantages a challenge because traditional architectures result in edges being engaged with data encryption and decryption. Moreover, in content applications such as video compression and intelligent collaborative applications, users are unaware of content considered private data, and the relative privacy sensitivity and security priority of tasks. The proposed solution avoids side‐channel attacks by eliminating the individual characteristics via risk‐free constraints under a certain amount of entropy at the security level. Further, a smart optimum approach based on MOEA/D was employed to strike the balance between privacy preservation and the cost/latency of activity. This reduces the burden on edge servers, enhances security based on a better offloading strategy, and protects against threats and attacks using a reasonable system architecture. Through experiments on different types of workflow parameters the validity and effectiveness of the offloading algorithm with privacy preservation was demonstrated.

Ramesh Kumar, Aarthi; Kuttiappan, Helenprabha

doi: 10.1002/cpe.7016pmid: N/A

Nowadays, brain tumor (BT), which is the abnormal growth of the mass of tissues in the human brain, is the main cause of death among kids and adults. The existing methodologies concentrate only on the BT classification and could not detect the BT size. This work proposes to detect the BT size utilizing Modified Deep Learning (MDNN) and multilevel thresholding (MT) utilizing modified dragonfly optimization (MD‐MT) algorithm. Afterward, the Histogram Clipping (HC)‐based Contrast Limited Adaptive Histogram Equalization (CLAHE) approach is implemented for enhancing the contrast of the inputted image, and certain features are extracted as of those contrast‐enhanced images. Next, the image is classified as (i) tumor image and (ii) nontumor image with the aid of MDNN. The tumor part is segmented as the classified tumor affected image utilizing MD‐MT. Here, the proposed HC‐CLAHE attains a 99.74% Spearman Rank Correlation, which is greater among other methods while the proposed MD‐MT attains a higher accuracy (99.73%). For classification, the proposed MDNN and the existing Artificial Neural Network, K‐Nearest Neighbor, Support Vector Machine, and Naïve Bayes are contrasted grounded on their performance in respect of precision, f‐measure, and recall. The proposed work shows excellent performance during the experimental evaluation.

Shahzad, Usman; Ahmad, Ishfaq; Hanif, Muhammad; Al‐Noor, Nadia H.

doi: 10.1002/cpe.7006pmid: N/A

Assessment of coefficient of variation (CV) is of major importance in numerous examinations. However, the appearance of extreme observations raises concerns about the outcomes of CV estimates based on conventional moments. So, motivated by some recent developments in finite sampling theory, we propose some new estimators of CV based on the properties of linear moments (L‐moments and Trimmed L‐moments), which are highly robust whenever outliers or extreme observations appear in a dataset. The proposed estimators are initially established on the premise that the variable of interest is nonsensitive which deals with the subjects that do not embarrass respondents when asked about them explicitly. These estimators are also applied to situations where the variable of interest is associated with sensitive issues that cause measurement errors resulting from nonresponse or unreliable reporting where such issues can be mitigated by increasing respondent participation by scrambled response models which obscure the true value of the sensitive variable. Four models are considered for this article: additive, multiplicative, mixed, and combined additive‐multiplicative models. Finally, in both nonsensitive and sensitive settings, real‐life data sets are employed to undertake simulation‐based analysis. In all cases, the proposed estimators considerably increased efficiency.

Bokade, Gayatri Umakant; Kanphade, Rajendra D.

doi: 10.1002/cpe.7011pmid: N/A

One of the basic requirements of our modern day society is personal authentication. Biometric recognition should make a human‐like identity determination by identifying its physiological and/or behavioral characteristics. In comparison to traditional knowledge‐based approaches, biometric identification systems have the potential to bring benefits. However, because of the difficulties in extracting non‐class discriminative features, the lack of protection during storage of extracted features, and poor recognition accuracy, most frequently used biometric systems lack model protection and robustness. This research proposed a Mimus multimodal biometric system focused on the combination of multiple modalities and optimal level fusion of features to resolve these problems. Initially, the novel Blob‐funk method extracts the complementary non‐class discriminatory information among different modalities, which accomplishes the biometric data enrollment. Thus, it extracts the different properties by comparing surrounding regions based on finding the local maxima and minima of the function. After extracting the features, they need to be stored in a secure manner in a database. Therefore, the paper incorporates the new code block protection strategy to achieve an effectual protection of continuous monitoring via the generation of non‐invertible features, which is used to create the templates, thus storing them in a database. Finally, the novel Lucynomial logistic regression system incorporates user authentication and thus achieves greater recognition accuracy through estimation of threshold value with confrontation of spoof attacks. Hence, compared to the existing techniques such as SVM, PCA, and DBN, the outcome of the proposed work attains 97.53% accuracy, 0.020% FAR, 96.44% recall, and 97.85% precision, thus exemplifying the competence of the novel system.

Sivalingam, Nithya Priya; Chinnasamy, Sundar; Suruli Muniyandi, Thanabal

doi: 10.1002/cpe.7012pmid: N/A

The leukemia represents a blood cancer that begins from bone marrow and spread to bloodstream and other limbs. The leukocytes pose various morphologies and size that makes it complex to detect and segment exactly. To consider this issue, this article develops a novel and efficient method for chronic lymphocytic leukemia (CLL) detection with blood smear images. Here, the preprocessing is done with Type 2 fuzzy and cuckoo search‐based filter for discarding artifacts and noise contained in the images. The cell segmentation is done with deep fuzzy clustering for producing the segments from the preprocessed blood smear images. The discovery of CLL is done with random multimodal deep learning (RMDL). The RMDL training is done with proposed Jaya‐competitive swarm optimization (Jaya‐CSO) algorithm. The developed Jaya‐CSO is devised by combining Jaya optimization and competitive swarm optimizer (CSO). Hence, the proposed Jaya‐CSO‐based RMDL is employed for discovering lymphocytic leukemia. The proposed Jaya‐CSO‐based RMDL offered improved performance with highest accuracy of 95.1%, highest true negative rate of 95.4% and highest true positive rate of 95.2%.

Konakoglu, Berkant; Aydemir, Salih Berkan; Kutlu Onay, Funda

doi: 10.1002/cpe.7017pmid: N/A

In the present article, the efficiency of a new hybrid learning method, named artificial neural network with gradient‐based optimizer algorithm (ANN‐GBO), is investigated to determine a local geoid. The outcomes of the assessed method are compared with classical ANN (without GBO), some metaheuristic‐based ANN models and other study results (interpolation methods). Four commonly used performance metrics, root mean square error (RMSE), mean absolute error (MAE), mean absolute relative error (MARE) and coefficient of determination (R2$$ {R}^2 $$) have been used to assess the applied methods. Assessment of predictions revealed that ANN‐GBO yielded the best results (RMSE: 10.51 cm, MAE: 8.26 cm, MARE: 0.27 cm, and R2$$ {R}^2 $$: 0.9622). The outcomes of the work clearly show that the ANN‐GBO approach has the lowest prediction error compared with ANN. It can be said that the GBO algorithm enhances the ANN capability in local geoid modeling. It may be recommended to optimize the weights of the ANN with GBO for high prediction accuracy. This research can be extended to other regions.

Demir, Alparslan Serhat; Mumcu, Ebru

doi: 10.1002/cpe.7019pmid: N/A

Automated guided vehicles (AGV), which are frequently used to reduce the workload of employees in production systems, are now being used in service systems. AGV has been widely used in the field of healthcare and is well suited to transporting. This study aims to provide AGV is being used in different areas of healthcare today within the scope of Industry 4.0. In recent years, the use of AGV has started to increase to reduce the workload and contamination risk of healthcare workers. In this study, it was aimed that AGVs would distribute autonomous drugs by processing patient information such as age, drug type, frequency of use and so forth. The information obtained from the internal diseases department of a university hospital was transferred to a computer simulation environment and the current situation and the proposed situation were analyzed separately, and the results were compared. The statistically verified results show that an autonomous AGV that will process patient information and distribute drugs reduced the workload of nurses by 14% in approximately a month.

Puri, Vartika; Kaur, Parmeet; Sachdeva, Shelly

doi: 10.1002/cpe.7020pmid: N/A

Recent years have witnessed the wide availability of an array of transactional datasets for mining and other research activities. A primary concern related to the public sharing of transactional datasets is identifying individuals whose data is being published. Data anonymization is a commonly utilized privacy preservation method for preventing user identification. However, the existing anonymization models such as km$$ {k}^m $$‐anonymity, ρ$$ \rho $$‐uncertainty, and (h, k, p)‐coherence for privacy preservation of transactional data do not provide complete protection from the various types of possible privacy attacks. Therefore, this article proposes a novel privacy model called (k, m, t)‐anonymity to effectively prevent identity and attribute disclosure as well as skewness attack on transactional data. A genetic algorithm‐based implementation of the model is also presented. The genetic algorithm clusters transactional data based on the similarity among the transactions for effective km$$ {k}^m $$‐anonymization with low information loss. The clustering algorithm simultaneously aims to minimize the skewness of data distribution in the obtained clusters for preventing skewness attack on anonymized data. Experimental results have verified that the (k, m, t)‐anonymity model ensures transactional data anonymization without significant information loss. The proposed privacy model is implemented using the proposed approach on two real‐world datasets (health domain and click‐stream data) and an enormous dataset generated synthetically (health domain consisting of 5,00,000 records). The relative error is less as compared to the relative privacy and disassociation technique for all test case scenarios. Hence, the proposed anonymization model maintains the data utility.

Fayyaz, Mohsin; Shah, Yasir Ali; Fayyaz, Ahmed; Mujtaba, Ghulam

doi: 10.1002/cpe.7018pmid: N/A

An architecture which promised very low computational complexity for optical interconnects in DCN was previously proposed, the very low computational complexity (VLCC) architecture. In this article, we present enhancement over VLCC architecture called very low computational complexity optical interconnect architecture with queueing (VLCC‐Q) architecture which represents M/D/64 output queue. Contention occurs at the destination if the source nodes which access the same destination node exceeds 64. The performance analysis is carried out through mathematical analysis, TCP simulation in NS2 and eye diagram of optical signal. The mathematical analysis gives theoretical bounds of optical switch, whereas TCP analysis gives practical situation results and eye diagram shows that optical signal is practically recoverable after the optical signal passes through various optical components of the optical interconnect. Results clearly show that VLCC‐Q outperforms VLCC architecture in terms of throughput increase, delay reduction and back pressure probability reduction.

Mustafa, Sidra; Amin, Muhammad; Akram, Muhammad Nauman; Afzal, Nimra

doi: 10.1002/cpe.7005pmid: N/A

The beta regression model (BRM) is appropriate when the response variable is continuous and is in the form of ratios and proportions. For the estimation of the BRM, the maximum likelihood estimation (MLE) method is used with a specific link function. However, the MLE provides unstable results when the explanatory variables are correlated. In this study, we consider some ridge parameters for the beta ridge regression estimator (BRRE) under different link functions. However, mostly the researchers do not pay much attention to the suitable link function. So, we consider five link functions to see the performance of ridge parameters in the BRRE. For the performance assessment of ridge parameters and different link functions, a Monte Carlo simulation and a real application are considered, where mean squared error is used as the evaluation criterion. Both the simulation and example findings demonstrate that the BRRE with the log–log link function provides efficient results.