International Journal of Mobile Network Design and Innovation (6 papers in press)

Regular Issues

• Nature-Inspired Based Ensemble Feature Selection and Stacked Ensemble Classifier Fusion for Android Malware Detection  
  by Anuja A. Rajan, R. Durga 
  Abstract: Android is widely used for tablets and smartphones. Thus, Android app malware has grown quickly in recent years. Malware detection in these apps is effective with ML algorithms. A reliable and effective malware detection approach is still difficult due to the vast number of features and high-dimensional dataset that provides the lowest accuracy despite research and industry efforts. Feature selection involves finding and removing features from a dataset while retaining class label variance. Using dynamic analysis of Android malware samples, this study introduces the Nature-Inspired Ensemble Feature Selection (NIEFS) and Stacked Ensemble Classifier Fusion (SECF) multi-classification models. The NIEFS model uses evolutionary computation methods like FMGOA, LEFPIO, and COSSA to remove redundant or irrelevant features and select relevant ones to improve detection accuracy. Multilayer Stacked Ensemble Classifier Fusion (SECF) integrates EC-created malware outputs using a Mutual Information (MI)-based ensemble approach, which has good detection accuracy. Machine learning methods like J48, mean weight deep belief network (MWDBN), REPTree, and Voted Perceptron can be combined. Finally, classifier performance was tested using MATLABR2020a, Precision (Pre), Recall (Rec), F-measure (FM), and Weighted F-measure.
  Keywords: Android Malware Detection (AMD); Nature-Inspired Based Ensemble Feature Selection (NIEFS); Mean Weight Deep Belief Network (MWDBN); Ensemble Technique;.
  DOI: 10.1504/IJMNDI.2025.10071599
   
  
• Machine Learning-Based Prediction of Allocative Localization Error in Wireless Sensor Networks  
  by Guo LI, Hongyu SHENG 
  Abstract: There may be a better way to deal with and write about any mistakes that happen in WSNs if they happen in a style called Allocative Localization Errors, or ALEs. Three important schemes are Adaptive Neuro-Fuzzy Inference Systems, Stacking Regression, and AdaBoostRegressors. Multiple regression methods are used by the stacking regression model to guess wrong traits in more than one variable. As part of the study, MOA was used to make some well-known ways of making predictions more correct. Someone came up with a brand-new ensemble model called STASA+MFO for this work. It takes the Stacking R model and adds ADA and MOA to it. With an R2 value of 0.9913, this method did better than the others that were tried. The ADA+ANFIS model had an R2 value of 0.9810, and the ANMF model (ADA with MO) had an R2 value of 0.9824.
  Keywords: wireless sensor networks; allocative localization errors; machine learning algorithm; metaheuristic algorithms.
  DOI: 10.1504/IJMNDI.2025.10071735
   
  
• Analysis of Spectrum Sensing using Convolutional Neural Network  
  by Sankarsan Panda, Chitra Devi D., Madhini M, PeriyarSelvam K 
  Abstract: This study uses a spectrum sensing method to examine spectrum sensing as a classification problem. It works with the help of a CM-LSTM network and a signal correlation matrix. Three signals picked up by the space antenna array were cross- correlated with each other, and a single signal was picked up to find out the time. It was made with an LSTM (long short- term memory) network. The LSTM is a great way to find traits that have something to do with time. Then, a list of the signals picked up by the array and the links between them were given to the LSTM classification model. I figured out how the messages fit into the timeline and location. There were a lot more good ways to find bands after this. We studied using models and found that the CM-LSTM spectrum-sensing algorithm works better than the SVM, GBM, RF, and ED-based spectrum-sensing algorithms.
  Keywords: Cognitive Radio Network (CRN); Cognitive Radio (CR); Signal-To-Noise (SNR); Matched Filter Detection (MFD); Energy Detection (Ed); CNN-LSTM; Automatic Repeat Request (ARQ).
  DOI: 10.1504/IJMNDI.2025.10072133
   
  
• Privacy Protection Algorithms in Mobile Network Design and Innovation: Enhancing Security in Next-Generation Wireless Communication  
  by Lili Qiu 
  Abstract: The integration of block chain technology into e-commerce has successfully addressed several critical challenges, including the lack of transaction transparency, potential data security risks, and high payment costs. Mobile Edge Computing (MEC) enhances block chain functionality by providing computational power while simultaneously meeting the stringent requirements for high real-time performance and low latency in e-commerce transaction systems. However, certain limitations persist within MEC-enabled e-commerce consortium block chains, such as breaches of user privacy, vulnerabilities in consensus algorithms, and other security concerns. This study proposes a secure transaction model specifically designed for MEC-enabled e-commerce consortium block chains. The proposed model employs a lightweight encryption algorithm to ensure the confidentiality of user information and transaction data, thereby preventing unauthorised disclosures. The simulation results indicate that the proposed solution effectively reduces classification accuracy from 91.1% to 1.42%, with only a 0.17% padding overhead when applied to a real dataset.
  Keywords: True Positive(TP); Mobile Crowd Sensing (MCS); Generative adversarial networks (GAN); software development kit (SDK); False Negative (FN); Target Tracking Area Selection (TTAS); Anonymous Handover Pro.
  DOI: 10.1504/IJMNDI.2026.10072507
   
  
• Deep Learning-aided Wireless Channel Estimation in 6G  
  by Sivanathan M, Rajasekar A, Veena T, S. Aswini 
  Abstract: Various processes can facilitate the exchange of information, allowing multiple inputs and outputs. A base station (BS) is essential for a user terminal (UT) to obtain accurate channel state information (CSI). However, pilot signals can only be transmitted a limited number of times and for a specific duration. Due to the dynamic nature of the environment, the position of the UT is constantly changing. Consequently, the number of pilot signals remains constant, necessitating repeated transmission of pilot signals in neigh boring cells. This repetition results in pilot contamination, which complicates the establishment of connections between adjacent cells. The minimal mean square error (MMSE) method can still yield reasonable results, even in the presence of pilot contamination. While the BS has knowledge of the channels linked to each UT, this information is not available to the MMSE method. In this study, two channel estimation approaches are proposed.
  Keywords: Deep Learning; Channel state information (CSI); LS and MMSE; line-of-sight (Lo S); Internet of Things (Io T); signal-to-noise ratio (SNR); MMPGA-LS; Res Net.
  DOI: 10.1504/IJMNDI.2025.10072670
   
  
• Measurement and Analysis of RSS in IoT Using Bluetooth Mesh Networks  
  by Xiaobin Wu, Dan Luo 
  Abstract: People who work with Industry 4 0 need to be able to think quickly and find things by looking inside things You can keep track of the steps each thing goes through as it's being made by putting parts and other things in the right places In other words, someone can see something being made Since GPS won't let you in, you'll need to come up with a new plan We can think of a new way to quickly find things inside with beacons that have a Received Signal Strength Indicator (RSSI) The study led to this plan You can use more than one kind of machine learning at the same time to do this These include CNNs, RNNs, and regressors They need to be trained on test beds in the real world These are places where you can see how well each method works We now know which ones work best, are easy to use, and help us find things inside. A live feed will be shown as the last test to see how well the translation works. From what we saw, the K-Nearest Neighbours method works the best. You now know how things stand and how much money your company really makes.
  Keywords: Technology: Bluetooth; Localization Improvement; Optimal K for KNN and WKNN; High-Level Architecture Bluetooth Mesh Profile Specification.
  DOI: 10.1504/IJMNDI.2025.10072671