Vol.
4 No.
March 2026 .
ISSN: 2987-6990 DOI: https://doi.
org/10.
60005/coreid.
Multi-Layer PHP Source Code Protection Model Using Encoding with Integrity Hashing and Anti-Reverse Engineering Sarmuni1.
Suhaeri Rumdani2.
Anna Rosdiana3.
Khaeroni4.
Ahmad Faturohman5.
Devid Sujianto6.
Deden Rifki7
1,2,3,4,5,6,7
Center for Information Technology and Data Management.
Universitas Islam Negeri Sultan Maulana Hasanuddin Banten
Article Info
ABSTRACT
Article history:
This study proposes a unified multi-layer PHP source code protection architecture that integrates obfuscation, layered encoding.
AES-256 encryption.
SHA-256 integrity verification, auto-restore, and Telegram-based security alerts.
Unlike prior PHP protection approaches that generally focus solely on code concealment or encryption, the proposed model integrates protection, tamper detection, and incident response into a single workflow.
The system was evaluated using a research-and-development approach, including prototype implementation, security testing, and performance benchmarking on PHP datasets of varying complexity.
The proposed method was compared with conventional protection mechanisms and commercial PHP encoders to assess resistance to deobfuscation, integrity-checking capability, and runtime overhead.
The results indicate that the proposed architecture improves code readability, reliably detects unauthorized file modifications, and maintains acceptable performance, with an average execution-time overhead of 15Ae20 ms, a CPU load increase of 8Ae12%, and a memory usage increase of 5Ae8 MB.
These findings suggest that the model provides a practical balance between stronger source-code protection and operational efficiency for PHP-based web applications.
Received March 08, 2026 Revised April 06, 2026 Accepted April 08, 2026 Keywords:
AES-256 encryption Anti-tamper Auto-restore mechanism Multi-layer encoding PHP source code protection SHA-256 hashing This is an open access article under the CC BY-SA license.
Corresponding Author:
Khaeroni Center for Information Technology and Data Management.
Universitas Islam Negeri Sultan Maulana Hasanuddin Banten.
Jl.
Syech Nawawi Al-Bantani.
Kp.
Curug.
Andamui.
Kota Serang.
Prov.
Banten.
Indonesia Email: khaeroni@uinbanten.
INTRODUCTION
Web-based applications developed with PHP are generally distributed in source code rather than compiled binaries or bytecode, making them highly vulnerable to threats such as piracy, unauthorized modifications, reverse engineering, and the exploitation of vulnerabilities introduced by the original code .
Unprotected PHP source code allows third parties to read, copy, modify, or even claim ownership of the code, resulting in intellectual property rights violations and potential business losses or application security breaches .
Although solutions such as PHP encoders .
SourceGuardian.
IonCub.
and obfuscators exist, many still use relatively simple techniques such as base64 encoding or static obfuscation, which can be easily dismantled using online deobfuscator tools .
uch as UnPHP) or static analysis .
, .
Additionally, commercial solutions often depend on specific PHP extensions, making them less flexible and expensive for small developers or open-source projects .
This research addresses the need for PHP source code protection that combines multi-layer obfuscation/encoding/encryption, hash-based integrity verification, and automated incident response .
utorestore and notification.
, improving both security complexity and operational resilience .
, .
The CoreID Journal | Vol.
No.
March 2026: 22-31 scientific urgency of this research lies in the need for a PHP source code protection model that: .
Uses a multilayer approach .
bfuscation encoding symmetric encryptio.
to increase reverse engineering complexity .
Ensures file integrity with hash mechanisms and change detection .
nti-tampe.
, .
Is equipped with automatic response mechanisms .
uto-restore and notificatio.
when illegal modifications occur .
, thereby improving availability and accountability aspects.
Recent studies confirm that combining obfuscation and encryption provides stronger protection .
, .
, .
Research by Rahmatulloh and Munir demonstrated that obfuscation techniques using the RC4 algorithm and ASCII encoding can prevent reverse-engineering threats to PHP source code .
However, this approach still has limitations in terms of encryption strength and integrity verification mechanisms.
Aprianto and Winarno developed a PHP 5 Encoder using Blowfish encryption, but it lacks hashing for integrity checks and automatic response mechanisms .
The integration of AES-256 encryption with SHA-256 hashing has been proven effective for protecting file documents .
and ensuring data integrity in cloud storage systems .
However, it has not been specifically adapted for PHP with runtime monitoring.
AES-256 provides robust confidentiality protection, while SHA-256 generates unique digital fingerprints that can detect even minor unauthorized alterations .
, .
However, the application of this combined approach specifically for PHP source code protection with automated incident response mechanisms has not been extensively explored in the literature.
Existing PHP protection solutions, such as SourceGuardian, ionCube, and PHPDefender, primarily focus on code obfuscation or single-layer encryption, which can be defeated through static analysis or online deobfuscation tools .
, .
While some approaches incorporate hashing for integrity verification .
, .
, they lack integrated runtime tamper detection and automated response mechanisms.
This study addresses these gaps by proposing a unified multi-layer protection architecture that combines obfuscation, layered encoding.
AES-256 encryption.
SHA-256 integrity verification, and automated incident response .
uto-restore and Telegram alert.
in a single PHP protection pipeline.
The main contribution is the end-to-end workflow that not only protects source code but also actively monitors and responds to tampering attempts during runtime, providing comprehensive security beyond conventional encoders.
Thus, this research aims to fill the gap between simple PHP source code protection and expensive commercial solutions, and to contribute to the field of web-based software security by proposing a comprehensive multi-layer protection model that integrates obfuscation, multi-layer encoding.
AES-256 encryption, integrity hashing, anti-tamper detection, auto-restore capabilities, and real-time notification systems through Telegram Bot API .
The research questions addressed in this study are: .
How to design and implement a PHP source code protection model that combines obfuscation techniques, multi-layer encoding, and AES encryption to prevent reverse engineering and code piracy? .
How to integrate hash-based integrity checking and antitamper mechanisms to detect illegal changes to protected PHP files? .
How to design auto-restore and Telegram Bot notification mechanisms that can be activated automatically when illegal modifications are detected? .
How to evaluate the effectiveness and performance of the proposed protection model, both from the security side .
esistance to reverse engineerin.
and the overhead side .
xecution time and file siz.
? METHOD This research employs a Research and Development (R&D) methodology with a prototyping approach to develop and validate a PHP source code protection model .
The R&D stages follow the Borg & Gall model: literature review, system design, prototype development, security experimentation, performance benchmarking, testing, revision, and limited implementation through controlled experiments .
Research Procedure Literature Study and Requirements Analysis Comprehensive review of PHP security literature focusing on obfuscation, multi-layer encoding.
AES encryption, cryptographic hashing, and anti-reverse engineering techniquesAirequirements analysis conducted from the developer and organizational security perspectives.
Protection Model Design This stage focuses on designing the proposed protection model by defining the system architecture and the mechanisms required to secure PHP source code.
The process begins by designing a multi-layer PHP encoder architecture comprising an encoder component and a runtime loader that decodes and executes protected scripts.
A multi-layer encoding scheme is then designed by combining several transformation techniques with AES encryption to ensure that the protected source code remains unreadable without proper In addition, a hash-based integrity checking mechanism is incorporated to detect unauthorized modifications by comparing the runtime hash values of protected files with their original references.
Multi-Layer PHP Source Code Protection Model Using Encoding.
(Sarmuni, et a.
- 23 CoreID Journal | Vol.
No.
March 2026: 22-31 enhance system resilience, an auto-restore mechanism recreates or replaces protected files when integrity violations are detected.
Prototype Development This stage focuses on developing and implementing the proposed PHP source code protection system based on the previously designed architecture.
The process begins with the development of a PHP encoder tool that transforms original PHP source code into a protected format via obfuscation, multi-layer encoding, and In parallel, a PHP loader module is developed to handle the runtime decryption and execution of protected code in a controlled environment.
Additional security modules are implemented to support the protection mechanism, including hash-based integrity verification to detect unauthorized modifications, runtime decryption routines to restore executable code, and a panic mode module that responds to detected tampering attempts.
Furthermore, the system includes an auto-restore feature that automatically recreates or replaces protected files when integrity violations occur.
This mechanism is complemented by a Telegram Bot notification system that delivers real-time alerts to administrators, enabling rapid response to potential security Security Experiments and Performance Benchmarking Security experiments were conducted to evaluate the effectiveness and resilience of the proposed multi-layer PHP source code protection model.
The developed encoder was subjected to a series of security tests and compared with several widely used PHP code protection tools.
These included commercial encoders, such as ionCube Encoder and SourceGuardian, as well as open-source encoders, such as PHPDefender (PDW) and FOPO PHP Obfuscator.
The evaluation focused on analyzing the resistance of each encoder to reverseengineering attempts, including code extraction, decoding, and structural analysis of the protected scripts.
addition to security evaluation, performance benchmarking was performed to measure the computational overhead introduced by the proposed protection mechanism.
The benchmarking process assessed several key performance indicators, including script execution time.
CPU utilization, and runtime memory consumption.
Results Analysis and Report Writing The results obtained from the security experiments and performance benchmarking were systematically analyzed to evaluate the effectiveness of the proposed protection model.
Quantitative analysis used descriptive statistics, comparative tables, and one-way ANOVA .
<0.
Qualitative analysis included a thematic analysis of reverse-engineering attempts and tampering case studies.
This analysis examined the outcomes of the conducted tests and compared them with those produced by existing PHP source code protection solutions.
Based on the analysis findings, the research outcomes were compiled and documented in a structured academic report.
The report was prepared in accordance with standard scientific writing conventions, encompassing sections such as introduction, research methodology, results, discussion, and System Design The encoder system proposed in this research is designed using a modular architecture to ensure flexibility, maintainability, and security.
The architecture consists of several interconnected modules that collectively perform code protection, integrity verification, and runtime execution, i.
Encoder: The module that is responsible for transforming the original PHP source code through a series of protection mechanisms, including obfuscation, encoding, and encryption, ultimately generating protected files with the .
php extension.
Loader: A lightweight runtime component that is responsible for decrypting and executing the protected code during application execution.
Hash: This module ensures that the protected files have not been altered or tampered with.
It calculates and verifies the integrity hash of PHP files using the SHA-256 hashing algorithm.
Decrypt: This module performs the decryption process of protected data using the Advanced Encryption Standard with a 256-bit key in Cipher Block Chaining (CBC) mode.
Panic Mode: a security response mechanism triggered when unauthorized modifications are detected.
When activated, this module initiates an automatic file restoration process and sends a notification to the system administrator via Telegram.
Key Management: AES keys stored in secure environment variables.
backups use separate keys in an encrypted config file.
The proposed protection mechanism employs a multi-layer encoding scheme to enhance the security of the PHP source code:
Multi-Layer PHP Source Code Protection Model Using Encoding.
(Sarmuni, et a.
- 24 CoreID Journal | Vol.
No.
March 2026: 22-31 Obfuscation: replaces original variable names, function names, and class names with randomly generated This process aims to obscure the code's logical structure and complicate reverse-engineering Multi-Layer Encoding, consisting of several sequential transformations, i.
- Layer 1: gzdeflate .
to reduce and transform the data structure - Layer 2: base64_encode .
to convert the compressed binary data into an ASCII-safe - Layer 3: String scrambling .
everse base.
to combine string reversal operations with another Base64 transformation to further obscure the encoded content.
AES-256 Encryption: Encrypting encoded data using AES-256 with Cipher Block Chaining (CBC) mode.
This encryption stage ensures that the encoded data cannot be directly interpreted without the corresponding decryption mechanism embedded in the loader component.
The overall system architecture is illustrated in Figure 1, which presents the complete workflow of the proposed protection model.
The architecture diagram describes the transformation pipeline starting from the original PHP source code input, passing through multiple protection layers including obfuscation, encoding, and encryption, and finally producing a protected output file.
Additionally, the architecture highlights the integration of an integrity verification mechanism that ensures the authenticity of protected files at runtime.
Figure 1.
Multi-layer architecture for PHP source code protection.
To ensure file integrity and prevent unauthorized modifications, the system incorporates a hashing and anti-tamper mechanism.
The hashing and anti-tamper mechanism is designed as follows:
Calculating the SHA-256 hash value of the original PHP file prior to the protection process.
The resulting hash values are then stored in a reference file named __hash_map.
During runtime, the loader module recalculates the hash value of the executed file and compares it with the reference hash stored in the mapping file.
If the calculated hash value differs from the stored reference, the system interprets this condition as a potential tampering attempt and activates the security response mechanism.
The system further enhances its defensive capability by implementing an auto-restore and panic notification mechanism.
The auto-restore and panic notification features are designed as follows:
Auto-Restore: In the auto-restore mechanism, backup copies of the original PHP files are securely stored in a separate directory.
Panic Notification: When unauthorized modifications are detected, the system sends a notification message through the Telegram API to a predefined chat channel.
Attacker Model The security evaluation assumes the following threat model:
Multi-Layer PHP Source Code Protection Model Using Encoding.
(Sarmuni, et a.
- 25 CoreID Journal | Vol.
No.
March 2026: 22-31 Attacker capabilities: Read access to protected files, static/dynamic analysis tools, online deobfuscators (UnPHP, etc.
), file modification privileges on the web server Attacker goals: Recover source code logic, bypass protection, and inject malicious code Attacker limitations: No access to encryption keys, no kernel-level privileges, no physical server access.
Experimental Environment Environment Specifications:
OS: Ubuntu 22.
04 LTS
PHP: 8.
6 (CLI & FPM)
Web Server: Nginx 1.
0 PHP-FPM
CPU: Intel i7-12700 .
cores @ 2.
1GH.
RAM: 32GB DDR4 Storage: NVMe SSD 1TB Benchmark Tool: Apache Bench .
Testing Dataset The testing dataset used in this study consists of several PHP-based applications with varying levels of complexity.
The purpose of using applications with different structural and functional characteristics is to evaluate the robustness, compatibility, and performance impact of the proposed multi-layer source code protection model under diverse development scenarios.
By testing across multiple levels of complexity, the evaluation can more accurately reflect real-world usage conditions, where PHP applications range from small scripts to large modular systems.
Table 1.
Testing dataset characteristics Dataset
Description Simple CRUD scripts Mini blog application MVC framework
E-commerce prototype File Count
Lines of Code
1,000
3,000
10,000
Complexity Low Medium High Very High Baseline Comparisons Protected files were compared against:
Commercial: SourceGuardian 15, ionCube PHP Encoder 15 Open-source: PHPDefender.
FOPO encoder Control: Unprotected PHP baseline Evaluation Metrics The evaluation of the proposed multi-layer PHP source code protection model is conducted using several parameters and metrics designed to measure both security effectiveness and system performance.
These metrics are intended to provide a comprehensive assessment of how well the protection mechanism safeguards the source code while maintaining acceptable runtime efficiency.
Table 2.
Evaluation parameters and metrics Parameter Readability Score Cracking Time Execution Time CPU Load Memory Usage Tamper Detection File Size Telegram Notification Unit min/hours yes/no yes/no Explanation Readability score based on structure complexity Time required to read/understand the main logic Average execution time per request Average CPU usage during testing Average memory usage per request False positive/negative rate .
test case.
Increase vs original Whether the Telegram notification is successfully sent Data collection was conducted using three main techniques: document and literature review, security experiments, and performance benchmarking.
Security testing involved attempts by security assessors to reverse-engineer the system and evaluate its readability and cracking time.
Performance benchmarking used standard tools such as Apache Bench, top, and free commands to measure execution time.
CPU load, and memory usage .
Data analysis was conducted using quantitative and qualitative techniques.
Quantitative analysis included descriptive analysis to calculate average, median, and standard deviation of performance metrics.
comparative analysis to compare metrics between the developed encoder and other encoders using comparison Multi-Layer PHP Source Code Protection Model Using Encoding.
(Sarmuni, et a.
- 26 CoreID Journal | Vol.
No.
March 2026: 22-31 and inferential statistical analysis using one-way ANOVA to compare runtime performance across encoders .
, .
Qualitative analysis included thematic analysis of security testing results from reverse engineering assessors and case analysis of specific tampering cases.
RESULTS AND DISCUSSION
Prototype Implementation The multi-layer encoder successfully processed all 83 test files across datasets D1AeD4, producing php files with an average increase in size of 28.
4% (E = 4.
2%).
The loader executed all protected files without errors, confirming functional compatibility.
The encoder module processes PHP source code through obfuscation, multi-layer encoding, and AES-256 encryption stages.
The implementation uses the OpenSSL library for AES-256-CBC encryption with randomly generated initialization vectors (IV) for each file to enhance security .
The obfuscation process replaces all variable names, function names, and class names with random alphanumeric strings .
, $userData becomes $a7f2e, function processPayment() becomes function b9d4k()), making the code structure difficult to understand while maintaining its functionality.
The multi-layer encoding applies three sequential transformations: gzip compression, base64 encoding, and string scrambling .
everse operation combined with additional base64 encodin.
, .
The hash verification module calculates SHA-256 hashes for all protected files and stores them in a JSON-formatted hash map.
The loader verifies the hash before decrypting and executing the protected code.
Testing confirmed that even a single byte modification triggers hash-mismatch detection, demonstrating the effectiveness of the integrity-checking mechanism .
, .
The auto-restore functionality maintains secure backups of original files in an encrypted directory When tampering is detected, the system automatically restores the original file within an average time of 45 ms .
tandard deviation: 8.
3 m.
, ensuring rapid recovery from unauthorized modifications.
The Telegram Bot notification system successfully delivers real-time alerts with a 99.
7% success rate .
successful notifications out of 1000 tampering event.
, with an average notification delivery time of 1.
Figure 2.
Sample PHP code transformation through a multi-layer protection pipeline.
Figure 2 illustrates a sample PHP code transformation through the multi-layer protection pipeline, showing the transformation from readable source code into obfuscated, encoded, and encrypted protected code with a readability score of 12/100.
Security Evaluation Results Security experiments were conducted to compare the developed multi-layer encoder against commercial and open-source encoders across the four dataset categories .
The readability scores were assessed by expert evaluators using a standardized evaluation rubric that considers code structure visibility, identifier comprehensibility, and ease of logic extraction.
Table 3.
Security comparison across different encoders Encoder Multi-layer .
IonCube v12.
SourceGuardian v13.
PDW v3.
Readability Score Avg Cracking Time 285 minutes 165 minutes 142 minutes 78 minutes Protection Level Very High High High Medium Multi-Layer PHP Source Code Protection Model Using Encoding.
(Sarmuni, et a.
- 27 CoreID Journal | Vol.
No.
March 2026: 22-31 FOPO v2.
Base64 only 65 minutes 8 minutes Medium Very Low The proposed multi-layer encoder achieved the lowest readability score .
3 out of .
, indicating the highest level of code obfuscation.
The average cracking time of approximately 4.
75 hours demonstrates significantly stronger protection than commercial solutions.
IonCube and SourceGuardian showed strong protection levels, with cracking times of 165 and 142 minutes, respectively, while open-source encoders PDW and FOPO were more easily reversed, with cracking times of 78 and 65 minutes, respectively .
The integrity-checking mechanism achieved 100% accuracy in detecting file modifications across all test scenarios.
Three types of tampering were tested: content modification, file injection, and partial deletion.
All attempts were immediately detected due to hash mismatches, triggering the panic mode response.
The falsepositive rate was 0%, and the false-negative rate was also 0% .
ll tampering attempts were detecte.
, .
, .
, .
Reverse engineering attempts using common deobfuscation tools failed to produce readable code from the multi-layer protected files.
Manual static analysis revealed that the combination of obfuscation, multi-layer encoding, and encryption creates sufficient complexity to deter casual attackers.
However, highly skilled attackers may still extract partial logic through dynamic analysis, highlighting that no protection is absolute .
Performance Benchmark Results Performance benchmarking was conducted in a standardized test environment using Apache Bench.
using Apache Bench with 10,000 requests .
oncurrency level: .
for each dataset category .
Table 4.
Performance comparison for dataset D3 .
0-3000 lines of cod.
Encoder No protection .
Multi-layer .
IonCube v12.
SourceGuardian v13.
PDW v3.
FOPO v2.
Exec Time .
CPU Load (%) Memory (MB) The proposed multi-layer encoder introduces an average execution time overhead of 16.
compared to unprotected code, which is competitive with commercial encoders IonCube .
overhead, 44.
7% increas.
and SourceGuardian .
1ms overhead, 40.
4% increas.
The CPU load increased 5 percentage points .
8% increas.
, and memory usage increased by 5.
8 MB .
5% increas.
compared to baseline .
Statistical analysis using one-way ANOVA revealed significant differences in execution time across encoders .
< 0.
However, post-hoc analysis showed that the proposed encoderAos performance does not differ significantly from IonCube and SourceGuardian, offering comparable runtime efficiency while providing stronger security protection .
The performance overhead scales proportionally with code complexity.
For simple applications, the overhead is minimal.
for complex applications, it increases but remains acceptable for typical web response times .
The auto-restore mechanism adds negligible overhead during normal operation, and the Telegram notification system operates asynchronously without affecting application performance.
The auto-restore mechanism adds negligible overhead during normal operation .
ess than 1ms per reques.
since hash verification is computationally efficient.
The Telegram notification system operates asynchronously, sending alerts without blocking the main execution thread, ensuring that incident response does not impact application availability .
Discussion and Implications The experimental results demonstrate that the proposed multi-layer protection model successfully addresses the research questions posed in the introduction.
The combination of obfuscation, multi-layer encoding, and AES-256 encryption creates a robust defense against reverse engineering, providing superior protection compared to both commercial and open-source alternatives .
The integration of hash-based integrity checking with automated incident response .
uto-restore and Telegram notificatio.
represents an improvement over conventional PHP encoders that focus only on obfuscation or encryption.
This proactive security approach enables immediate detection and remediation of tampering attempts and provides real-time security monitoring .
The acceptable performance overhead makes the solution practical for production environments.
The performance characteristics are comparable to those of commercial solutions, but without licensing costs, making the approach suitable for individual developers and small organizations.
The performance Multi-Layer PHP Source Code Protection Model Using Encoding.
(Sarmuni, et a.
- 28 CoreID Journal | Vol.
No.
March 2026: 22-31 characteristics are comparable to commercial solutions that cost hundreds of dollars in licensing fees, making this approach viable for individual developers and small organizations with limited budgets .
However, several limitations should be acknowledged.
First, like all client-side protection mechanisms, the proposed model cannot provide absolute security against determined attackers with unlimited time and resources.
Skilled reverse engineers can eventually extract code logic through dynamic analysis techniques such as runtime debugging and memory dumping.
Second, the loader module remains visible in the protected files, potentially providing clues about the protection mechanism.
Third, the system relies on the security of the encryption key management, and key compromise would undermine the entire protection scheme .
Future enhancements may include domain/IP binding, license key verification, hardware-based key storage, integration with intrusion detection systems, and machine learning-based anomaly detection for advanced tampering detection .
Overall, this research contributes to software protection research by demonstrating a practical implementation of a multi-layer defense strategy specifically designed for PHP applications and for interpreted programming languages such as Python.
Ruby, or JavaScript, thereby extending its impact beyond PHP applications .
CONCLUSION
This research has successfully designed, developed, and evaluated a multi-layer PHP source code protection model that integrates obfuscation techniques, three-layer encoding .
zdeflate Ie base64 Ie scrambl.
AES-256-CBC encryption.
SHA-256 integrity verification, and automated response mechanisms .
uto-restore and Telegram alert.
The system demonstrates a unified protectionAedetectionAeresponse workflow that enhances software security beyond conventional approaches.
The findings indicate that the proposed model provides superior protection against reverse engineering compared to both commercial encoders (IonCube.
SourceGuardia.
and open-source alternatives (PDW.
FOPO).
This result is evidenced by the lowest readability score .
ICC = 0.
and the longest average cracking time .
, reflecting significantly increased analysis complexity.
Additionally, the tamper detection mechanism achieved a 98% detection rate, with an average latency of 12 ms and a falsepositive rate of 2%, confirming the effectiveness of the integrity verification process.
From a performance perspective, the system introduces an acceptable overhead, with an average increase of 16.
4 ms in execution time, 8.
5% in CPU load, and 5.
8 MB in memory usage.
Statistical analysis shows that this overhead is comparable to commercial solutions, indicating that enhanced security can be achieved without significant performance degradation.
In terms of reliability, the system demonstrates robust operational performance, including 100% file recovery via the auto-restore mechanism .
verage recovery time of 45 m.
and a 99.
7% success rate in delivering real-time Telegram notifications.
These results highlight the practical feasibility of deploying the model in real-world production environments.
This research successfully bridges the gap between simplistic protection tools and costly commercial solutions by providing a cost-effective, enterprise-grade security model that requires no licensing fees or proprietary extensions.
The key contributions of this study include: .
an end-to-end protection pipeline combining code concealment, runtime monitoring, and automated remediation.
a practical and accessible solution for PHP developers.
real-time incident response through integrated notification mechanisms.
Despite these contributions, several limitations remain, particularly regarding potential key exposure risks and the possibility of bypass via advanced attack techniques such as memory dumping.
Therefore, future research should focus on enhancing the security framework by leveraging hardware-based key management .
HSM.
, domain/IP binding mechanisms, and machine learning-based anomaly detection to address advanced persistent threats.
ACKNOWLEDGEMENTS
The authors would like to express their sincere gratitude to the reviewers for their valuable feedback and constructive suggestions that contributed to the improvement of this research.
This study was conducted using the authorsAo personal funding without external financial support.
The authors also extend their appreciation to all parties who supported the completion of this research.
In particular, the authors would like to thank the Unit Pelaksana Teknis (UPT) Pusat Teknologi Informasi dan Pangkalan Data Universitas Islam Negeri Sultan Maulana Hasanuddin Banten for providing the research facilities and technical environment necessary for conducting the experiments and implementing the system.
The support and infrastructure provided greatly contributed to the successful completion of this study.
REFERENCES