Abstract In today ’s distributed computing paradigm , achieving resilience against unpredictable faults is still a challenging task. Several state -of-the -art fault tolerance algorithms are static and rule -based, lacking the ability to foresee future events. This work investigates a novel self -healing framework that empl oys Long S hort -Term Memory (LSTM) networks for fault prediction and a Q -learning -based Reinforcement Learning (RL ) methodology for adaptive fault recovery. In this paper, we model the recovery decision process as a Markov Decision Process (MDP) whereby the framework learns the best actions to take autonomously from the system state transitions, as well as any feedback received. A high -fidelity simulation environment was utilized to inject realistic errors so that we could rigorously evaluate the model ’s performance. The comparative experimental results demonstrate significant improvements in accuracy (96.3%) and Mean Time to Recovery (MTTR) (34 sec) as well as improved availability (97.2%) compared to baseline and state -of-the -art models. This work highlights the oppo rtunity for hybrid Artificial Intelligence ( AI) models to promote resilience in cloud -edge distributed systems. The modular design of this framework also presents learning opportunities for deploying as a lesson plan in courses that focus on both distributed systems and intelligent automation in comput er science curricula.

References [1] Abdel Aziz N., Fouad G., Al -Saeed S., and Fawzy A., “Deep Q -Network (DQN) Model for Disease Prediction Using Electronic Health Records (EHRs),” Sci , vol. 7, no. 1, pp. 1 -15, 2025. https://doi.org/10.3390/sci7010014

[2] Ahmad F., Haroon M., and Siddiqui Z., “Evaluating Fault Tolerance in Distributed Systems Using Predictive Analytics with Gated Recurrent Unit and Long Short -Term Memory Models,” Journal of Information Systems Engineering and Management , vol. 10 , no. 27s , pp. 378 -399 , 2025. https://doi.org/10.52783/jisem.v10i27s.4421

[3] Ahmad F., Haroon M., and Siddiqui Z., “Predictive Analytics for Fault Tolerance in Distributed Systems Using Logistic Regression and Support Vector Machine,” International Journal of Environment al Sciences , vol. 11 , no. 22s, pp. 5275 -5289, 2025. https://doi.org/10.64252/yed3d546

[4] Ahmed W. and Wu Y., “A Survey on Reliability in Distributed Systems,” Journal of Computer and System Sciences , vol. 79, no. 8, pp. 1243 -1255, 2013. https://doi.org/10.10 16/j.jcss.2013.02.006

[5] Amin Z., Sethi N., and Singh H., “Review on Fault Tolerance Techniques in Cloud Computing,” International Journal of Computer Applications , vol. 116, no. 18, pp. 11 -17, 2015. https://research.ijcaonline.org/volume116/number 18/pxc39027 68.pdf

[6] Bampoula X., Nikolakis N., and Alexopoulos K., “Condition Monitoring and Predictive Maintenance of Assets in Manufacturing Using LSTM -Autoencoders and Transformer Encoders,” Sensors , vol. 24, no. 10, 1 2-5, 2024. https://doi.org/10.3390/s24103215

[7] Ban k D., Koenigstein N., and Giryes R., Data Mining and Knowledge Discovery Handbook , Springer International Publishing, 2023. https://doi.org/10.1007/978 -3-031 -24628 -9_16

[8] Bappy F., Islam T., Zaman T., Hasan R., and Caicedo C., “A Deep Dive into the Google Cl uster Workload Traces: Analyzing the Application Failure Characteristics and User Behaviors,” in Proceedings of the 10 th IEEE International Conference on Future Internet of Things and Cloud , Marrakesh, pp. 103 -108, 2023. DOI: 10.1109/FiCloud58648.2023.0002 3

[9] Berman R., Azar O., and Lichtman Y., “Simulated Fault Injection Environments for Teaching System Dependability,” Computer Applications in Engineering Education , vol. 29, no. 6, pp. 1532 - 1545, 2021. DOI: 10.1002/cae.22411

[10] Carmona R., Lauriere M., and Tan Z., “Model - Free Mean -Field Reinforcement Learning: Mean - Field MDP and Mean -Field Q -Learning,” The Annals of Applied Probability , vol. 33, no. 6B, 5334 -5381, 2023. DOI: 10.1214/23 -AAP1949

[11] Cheng Y., Elsayed E., and Huang Z., “Systems Resilience Assessments: A Review, Framework and Metrics,” International Journal of Production Research , vol. 60, no. 2, pp. 595 -622, 2022. https://doi.org/10.1080/00207543.2021.1971789

[12] Choe C., Baek S., Woon B., and Kong S., “Deep Q learning with LSTM for Traffic Light Control,” in Proceedings of the IEEE 24 th Asia -Pacific Conference on Communications , Ningbo, pp. 331 - 336. 2018. DOI: 10.1109/APCC.2018.8633520

[13] Clifton J. and Laber E., “Q -Learning: Theory and Applications,” Annual Review of Statistics and its Application , vol. 7, no . 1, pp. 279 -301, 2020. https://doi.org/10.1146/annurev -statistics - 031219 -041220

[14] Coulouris G., Dollimore J., Kindberg T., and Gordon B., Distributed Systems: Concepts and Design , Pearson Education, 2011. https://books.google.co.in/books?id=3ZouAAAA QBAJ

[15] Der Kiureghian A., Ditlevsen O., and Song J., “Availability, Reliability and Downtime of Systems with Repairable Components,” Reliability Engineering and System Safety , vol. 92, no. 2, pp. 231 -242, 2007. https://doi.org/10.1016/j.ress.2005.12.003

[16] Fan J., Wang Z., Xie Y., and Yang Z., “A Theoretical Analysis of Deep Q -Learning,” in Proceedings of the 2 nd Conference on Learning for Dynamics and Control PMLR , Berkeley, pp. 486 - 489, 2020. https://proceedings.mlr.press/v120/yang20a.html

[17] Gogineni A., “Ar tificial Intelligence -Driven Fault Tolerance Mechanisms for Distributed Systems Using Deep Learning Model,” Journal of Artificial Intelligence, Machine Learning and Data Science , vol. 1, no 4, pp. 1 -6, 2023. doi.org/10.51219/JAIMLD/anila -gogineni/519

[18] Goyal A. and Lavenberg S., “Modeling and Analysis of Computer System Availability,” IBM Journal of Research and Development , vol. 31, no. 6, pp. 651 -664, 1987. DOI: 10.1147/rd.316.0651

[19] Heda L. and Sahare P., “QLGWYB: Design of an Efficient Model for Analyzing C rowd Behavior Through Quad LSTM and Quad GRU Fusion Enhanced by Q -Learning and YOLO,” Iran Journal of Computer Science , vol. 8, pp. 1463 - 1483, 2025. https://doi.org/10.1007/s42044 -025 - 00272 -6

[20] Kambala G., “Intelligent Fault Detection and Self - Healing Archit ectures in Distributed Software Systems for Mission -Critical Applications,” International Journal of Scientific Research and Management , vol. 12, no. 10, 1647 -1657, 2024. DOI: 10.18535/ijsrm/v12i10.ec11

[21] Karamzadeh A. and Shameli -Sendi A., “Reducing Cold St art Delay in Serverless Computing Using Lightweight Virtual Machines,” Journal of Network and Computer Applications , vol. 232, no. c, pp 104030, 2024. https://doi.org/10.1016/j.jnca.2024.104030

[22] Khowaja S. and Khuwaja P., “Q -Learning and LSTM Based Deep Act ive Learning Strategy for Malware Defense in Industrial IOT Applications,” Multimedia Tools and Applications , vol. 80, pp. 14637 -14663, 2021. https://doi.org/10.1007/s11042 -020 -10371 -0

[23] Liang Y., Ruan N., Yi L., and Su X., “An Approach to Workload Generatio n for Modern Data Centers: A View from Alibaba Trace,” BenchCouncil Transactions on Benchmarks, Standards and Evaluations , vol. 4, no. 1, pp. 100 - 164, 2024. https://doi.org/10.1016/j.tbench.2024.1001

[24] Liu Y., Young R., and Jafarpour B., “Long -Short - Term Mem ory Encoder -Decoder with Regularized Hidden Dynamics for Fault Detection in Industrial Processes,” Journal of Process Control , vol. 124, pp. 166 -178, 2023. https://doi.org/10.1016/j.jprocont.2023.01.015

[25] Pasham S., “Fault -Tolerant Distributed Computing for Real -Time Applications in Critical Systems,” The Computertech , vol. 6, pp. 1 -29, 2020. https://www.yuktabpublisher.com/index.php/TCT/a rticle/view/142/127

[26] Punia S., Nikolopoulos K., Singh S., Madaan J., and Litsiou K., “Deep Learning with Long Short - Term Me mory Networks and Random Forests for Demand Forecasting in Multi -Channel Retail,” International Journal of Production Research , vol. 58, no. 16, pp. 4964 -4979, 2020. https://doi.org/10.1080/00207543.2020.1735666

[27] Puterman M., Handbooks in Operations Research and Management Science , Elsevier, 1990. https://doi.org/10.1016/S0927 -0507(05)80172 -0

[28] Reiss C., Wilkes J., and Hellerstein J., Google Clu ster -Usage Traces: Format+ Schema, White Paper, Google Inc, 2011. https://www.researchgate.net/profile/Auday -Al- Dulaimy/post/Are -there -any -datasets -for - cloudSim/attachment/59d61de379197b807797be3e/A S%3A273823268573184%401442295962733/downlo ad/Google+cluster+usage+traces.pdf

[29] Salman H., Kalakech A., and Steiti A., “Random Forest Algorithm Overview,” Babylonian Journa l of Machine Learning , vol. 2024. pp, 69 -79, 2024. https://doi.org/10.58496/BJML/2024/007

[30] Samir A., Dagenborg H., and Johansen D., “QMConn: A Self -Healing Controller for Microservices Using Q -Learning and Markov Decision Processes,” in Proceedings of the IEEE/ACM 17 th International Conference on Utility and Cloud Computing , Sharjah, pp. 389 - 398, 2024. DOI: 10.1109/UCC63386.2024.00060

[31] Sekar J. and Aquilanz L., “Autonomous Cloud Management Using AI: Techniques for Self - Healing and Self -Optimization,” Journal of Emerging Technologies and Innovative Research , vol. 10, no. 5, pp. 571 -580, 2023. http://www.jetir.org/papers/JETIR2305G78.pdf

[32] Sen P., Hajra M., and Ghosh M., Emerging Technology in Modelling and Graphics , Springer, 2020. https://doi.org/10.1007/978 -98 1-13 -7403 - 6_11

[33] Shah H. and Patel J., “Self -Healing AI: Leveraging Cloud Computing for Autonomous Software Recovery,” International Journal of Intelligent Systems and Applications in Engineering , vol. 10, no. 3s, pp. 341 -351, 2022. https://ijisae.org/index. php/IJISAE/article/view/7502/ 6515

[34] Shahid M., Islam N., Alam M., Mazliham M. , and Musa S., “Towards Resilient Method: An Exhaustive Survey of Fault Tolerance Methods in the Cloud Computing Environment,” Computer Science Review , vol. 40, no. 100398, pp. 1 -38 , 2021. DOI: 10.1016/j.cosrev.2021.100398

[35] Shukla A., Kumar S., and Singh H, “Fault Tolerance Based Load Balancing Approach for Web Resources in Cloud Environment,” The International Arab Journal of Information Technology , vol. 17, no. 2, pp. 225 -232, 2020. https://www.iajit.org/portal/PDF/Vol%2017,%20No .%202/17514.pdf

[36] Sun X., Cui B., and Cai Z, “Deep Q -Learning Based Circuit Breaking Method for Micro - Services in Cloud Native Systems,” in Proceedings of the CCF Conference on Computer Supported Cooperative Wo rk and Social Computing , Singapore, pp. 348 -362, 2024. https://doi.org/10.1007/978 -981 -99 -9637 -7_26

[37] Syed A. and Anazagasty E., “AI -Driven Infrastructure Automation: Leveraging AI and ML for Self -Healing and Auto -Scaling Cloud Environments,” International J ournal of Artificial Intelligence Data Science, and Machine Learning , vol. 5, no. 1, pp. 32 -43, 2024. https://doi.org/10.63282/3050 -9262.IJAIDSML - V5I1P104

[38] Torabi H., Mirtaheri S., and Greco S. “Practical Autoencoder Based Anomaly Detection by Using Vector Reconstruction Error,” Cybersecurity , vol. 6, no. 1, pp. 1 -13, 2023. https://doi.org/10.1186/s42400 -022 -00134 -9

[39] Vadisetty R., Polamarasetti A., Butani J., Prajapati S., and et al., “AI -Powered Self -Healing and Fault -Tolerant Cloud Infrastructures for Impro ved Resilience and Reliability,” SSRN Electronic Journal , vol. 9, no. 1, pp. 1 -10, 2024. http://dx.doi.org/10.2139/ssrn.5286332

[40] Varma S., “Artificial Intelligence in Cloud Computing: Building Intelligent, Distributed, and Fault -Tolerant Systems,” Internati onal Journal of AI, BigData, Computational and Management Studies , vol. 3, no. 1, pp. 37 -45, 2022 . DOI: 10.63282/3050 -9416.IJAIBDCMS -V3I1P105

[41] Wiering M. and Otterlo M., Reinforcement Learning: State -of-the -Art, Adaptation, Learning, and Optimization , Sprin ger, 2012. https://doi.org/10.1007/978 -3-642 -27645 -3

[42] Xiao W., Fang X., Liu B., Wang J., and Zhu X., “UNION: Fault -Tolerant Cooperative Computing in Opportunistic Mobile Edge Cloud,” ACM Transactions on Internet Technology , vol. 23, no. 4, pp. 1 -27, 2023. https://doi.org/10.1145/3617994

[43] Yang Y., Gao Y., Ding Z., Wu J., and et al., “Advancements in Q‐learning Meta‐Heuristic Optimization Algorithms: A Survey,” Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery , vol. 14, no. 6, pp. 1 -37, 2024. https://doi.org/10.1002/widm.1548

[44] Yang Z., Jin Y., Liu J., and Xu X., “An Intelligent Fault Self -Healing Mechanism for Cloud AI Systems via Integration of Large Language Models and Deep Reinforcement Learning,” arXiv Preprint , vol. arXiv:2506.07411v1, pp. 1-6, 2025. https://doi.org/10.48550/arXiv.2506.07411

[45] Zadakbar O., Imtiaz S. , and Khan F., “Dynamic Risk Assessment and Fault Detection Using a Multivariate Technique,” Process Safety Progress , vol. 32, no. 4, pp. 365 -375, 2013. https://doi.org/10.1002/prs .11609

[46] Zhang L., Jia T., Jia M., Wu Y., and et al., “A Survey of AIOps for Failure Management in the Era of Large Language Models,” arXiv Preprint , vol. arXiv:2406.11213v4, 2024. https://doi.org/10.48550/arXiv.2406.11213

[47] Zhang S., Huang Z., and Lang Y., “A pplication of Video Game Algorithm Based on Deep Q - Network Learning in Music Rhythm Teaching,” The International Arab Journal of Information Technology , vol. 22, no. 1, pp. 124 -138, 2025. https://doi.org/10.34028/iajit/22/1/10

[48] Zhang S., Wang Y., Liu M., an d Bao Z., “Data - Based Line Trip Fault Prediction in Power Systems Using LSTM Networks and SVM,” IEEE Access , vol. 6, pp. 7675 -7686, 2017. DOI: 10.1109/ACCESS.2017.2785763

[49] Zhao Z., Chen W., Wu X., Chen P., and Liu J., “LSTM Network: A Deep Learning Approach for Short‐Term Traffic Forecast,” IET Intelligent Transport Systems , vol. 11, pp. 68 -75, 2017. https://doi.org/1 0.1049/iet -its.2016.0208