Figure 1:
Figure 2:
Figure 3:
Figure 4:
Figure 5:
Figure 6:
Figure 7:
Figure 8:
Figure 9:
Figure 10:
Figure 11:
j_ijssis-2025-0043_tab_003
| The components are shown below | |
|---|---|
| CBP | Convergent Billing Point implements rating, charging, and accounting functions and supports both online charging and offline charging, also providing real-time QoS control, and this is triggered based on the threshold configured in the tariff. |
| DCC Proxy | The function supports the specific demands of Diameter Charging—a dedicated online mediation instance, to take control of any internal routing of Diameter traffic based on subscriber. |
| GGSN | Manages data sessions and integrates with Online Charging System for real-time data rating and charging. |
| F5 | F5 Distributed Cloud Services Billing service enables a subscriber to understand usage reports, quotas, and pricing, obtain usage reports, and switch between subscription plans. |
Fraud detection methods
| Classification method | Description | Limitation | Reference |
|---|---|---|---|
| XGBoost | The method uses a gradient-boosting framework to build full-scale decision trees and implement parallel decision trees. | Trends to overfit the data. | Sheng and Yu (2022), Bao, (2020) |
| ADABoost | The technique handles binary classification problems and improves predictability by the conversion of a larger number of weak learners into strong learners. | The requirement of a dataset devoid of most of the noise. | Yulita et al. (2021), Chang and Fan (2019) |
| Naive Bayes | The method is based on Bayes’ theorem to predict the outcome by the probability of occurrence. | The downfall is faced by the zero-frequency issue, which is the missing variable as zero. | Vijay and Verma (2023), Hairani et al. (2021) |
| Decision Tree Classifier | The algorithm uses classification and regression problems, and it works on a tree-based structure, which serves as the classifier of the dataset. | High in computation of resources and changes in data affect the outcome. | Zulfikar et al. (2018), Indumathi et al. (2021) |
| Random Forest classifier | The technique allows aggregation of several Decision Tree classifiers to improve the predictive capability of the algorithms. | The issue on the massive amount of computational resources and the time required between periods. | Mishra et al. (2020), Lu et al. (2019) |
| KNN | The method is used for classification on a distance-based approach to locate all unknown data points. | The downfall does not work on high dimensionality or large records. | Lu et al. (2015), Altay (2022) |
| Logistic Regression | The techniques are used for dichotomous and dependent variables. | The issue is on overfitting the count of features and recorded observations. | Doss and Gunasekaran (2023), Bheemesh and Deepa (2023) |
Overview of cloud key services
| Service | Description | AWS | Azure | GCP |
|---|---|---|---|---|
| Computing | Virtual machines and scalable computing resources | EC2 | Virtual machines | Compute engine |
| Object storage | Storage for unstructured data in objects | S3 | Blob storage | Cloud storage |
| Block storage | Storage for data in blocks, similar to traditional hard drives | EBS | Disk storage | Persistent disk |
| Database (relational) | Managed relational database services | RDS | SQL database | Cloud SQL |
| Database (NoSQL) | Managed NoSQL database services | DynamoDB | Cosmos DB | Firestore/datastore |
| CDN | Global distribution of content to reduce latency and improve performance | CloudFront | Azure CDN | Cloud CDN |
| Serverless computing | Running code without managing server infrastructure | Lambda | Functions | Cloud functions |
| Big data processing | Processing and analyzing large datasets | EMR | HDInsight | Dataproc |
| Machine learning | Provision of services and tools for machine learning | SageMaker | Machine Learning Studio | AI platform |
| Identity and access management | Management of users and permissions | IAM | Azure AD | Cloud IAM |
| Monitoring and logging | Monitoring and logging of applications and infrastructure | CloudWatch | Azure monitor | Stackdriver (operations) |
| Networking | Management of networks and their security | VPC | Virtual network | VPC |
| Container orchestration | Management and orchestration of containers | EKS | AKS | GKE |
| Data warehousing | Storage and analysis of large amounts of structured data | Redshift | Synapse analytics | BigQuery |
| Backup and disaster recovery | Backup and recovery of data | Backup | Azure backup | Backup |