[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119534-en":3,"doc-seo-119534-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},119534,1374391974564,"Clementine","https://ap-avatar.wpscdn.com/avatar/14000253aa45c000a9e?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779874745381141002",8,"Research & Report","Enhancing Row-Sampling-Based Rowhammer Defense Methods with Machine Learning Approach","This paper investigates integrating machine learning into the Row-Sampling technique to improve its effectiveness against Rowhammer attacks in DRAM systems. A multidimensional multilabel predictor dynamically estimates and updates probability thresholds using real-time memory access patterns, refining row selection for targeted refresh. Experimental results indicate improved security by reducing Rowhammer-induced bit flips while preserving energy efficiency and limiting performance overhead. The proposed ML-enhanced method offers a scalable, adaptive defense for modern DRAM memory vulnerabilities.","UDC 004.33  \nEnhancing Row-Sampling-Based Rowhammer defense methods with Machine Learning approach  \nValentyn Mazurok1, Volodymyr Lutsenko 1  \n1 National Technical University of Ukraine «Igor Sikorsky Kyiv Polytechnic Institute», Cybersecurity Department of Institute of Physics and Technology  \nAbstract  \nThis paper investigates the integration of machine learning into the Row-Sampling technique to enhance its effectiveness in mitigating Rowhammer attacks in DRAM systems. A multidimensional multilabel predictor model is employed to dynamically predict and adjust probability thresholds based on real-time memory access patterns, improving the precision of row selection for targeted refresh. The approach demonstrates significant improvements in security, reducing Rowhammer-induced bit flips, while also maintaining energy efficiency and minimizing performance overhead. By leveraging machine learning, this work refines the Row-Sampling method, offering a scalable and adaptive solution to memory vulnerabilities in modern DRAM architectures.  \nKeywords: DRAM, Rowhammer, memory defense, machine learning  \nIntroduction  \nDynamic Random Access Memory (DRAM) is a critical component in modern computing systems, providing high-density and low-cost storage for a wide range of applications. However, as DRAM technology scales to smaller form factors, it becomes increasingly vulnerable to security threats such as the Rowhammer attack [1] . Rowhammer exploits physical vulnerabilities in DRAM by inducing bit flips in adjacent memory rows through frequent and aggressive row activations, potentially leading to data corruption or security breaches.  \nTo mitigate Rowhammer attacks, a common approach is to refresh vulnerable memory rows ata higher frequency. [2] However, static refresh strategies can impose significant performance and energy penalties, as they do not adapt to runtime memory access patterns or inherent variability in DRAM hardware. This calls for an intelligent, adaptive mechanism to optimize refresh rates for individual DRAM rows based on their susceptibility to Rowhammer and runtime usage characteristics.  \nIn this work, we propose a machine learningbased approach leveraging Multidimensional predictor to dynamically generate and update refresh probabilities for DRAM rows.  \nMultidimensional Predictors is an ensemble learning method, are particularly suited for this task due to their ability to handle highdimensional data, robustness to overfitting, and interpretability. By training the model on access patterns, row activation frequencies, and hardware-specific features, the algorithm can predict optimal refresh intervals for each row, minimizing the risk of  \nRowhammer while balancing performance and energy efficiency. By introducing machine learning into the DRAM refresh process, we aim to bridge the gap between security and efficiency, paving the way for more resilient memory systems in future computing architectures.  \n1. Background  \nThe challenge of mitigating Rowhammer attacks and improving DRAM refresh strategies has garnered significant attention in recent years.  \n[3] As DRAM scaling continues, innovative techniques are needed to address both the security vulnerabilities and the performance trade-offs inherent in traditional memory management strategies. Figure 1 illustrates the typical structure of a modern DRAM system. DRAM is arranged as a hierarchical array  \ncontaining billions of DRAM cells, each storing a single bit of data. In contemporary systems, the CPU chip incorporates multiple memory controllers, with each controller connected to a DRAM channel. These controllers handle read, write, and maintenance operations (such as refresh) via a dedicated I/O bus that operates independently of other channels in the system, as shown in Figure 1.  \nEach DRAM channel can support one or more DRAM modules, and each module is composed of one or more DRAM ranks. A rank consists of several DRAM chips that function in unison, while","cbCaikUzu30h1NMQ","https://ap.wps.com/l/cbCaikUzu30h1NMQ","pdf",924895,1,6,"English","en",105,"# Introduction\n## Mitigation Techniques\n# Background\n## DRAM organization and RowHammer cause\n## Single-sided and double-sided access patterns","[{\"question\":\"How does the proposed method enhance the Row-Sampling RowHammer defense in DRAM?\",\"answer\":\"It integrates a multidimensional multilabel predictor that uses real-time memory access patterns to dynamically predict and adjust probability thresholds for targeted refresh row selection.\"},{\"question\":\"What data does the machine learning model use to refine refresh behavior?\",\"answer\":\"The model is trained on access patterns, row activation frequencies, and hardware-specific features to estimate appropriate refresh intervals per row.\"},{\"question\":\"Why can static refresh strategies be inefficient for Rowhammer mitigation?\",\"answer\":\"Static strategies do not adapt to runtime memory access patterns or hardware variability, which can cause unnecessary performance and energy penalties.\"}]","Enhancing Row-Sampling-Based Rowhammer Defense Methods with Machine Learning Approach | PDF",1785724825,15,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"enhancing-row-sampling-based-rowhammer-defense-methods-with-machine-learning-approach","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/enhancing-row-sampling-based-rowhammer-defense-methods-with-machine-learning-approach/119534/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"How does the proposed method enhance the Row-Sampling RowHammer defense in DRAM?","Question",{"text":75,"@type":76},"It integrates a multidimensional multilabel predictor that uses real-time memory access patterns to dynamically predict and adjust probability thresholds for targeted refresh row selection.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What data does the machine learning model use to refine refresh behavior?",{"text":80,"@type":76},"The model is trained on access patterns, row activation frequencies, and hardware-specific features to estimate appropriate refresh intervals per row.",{"name":82,"@type":73,"acceptedAnswer":83},"Why can static refresh strategies be inefficient for Rowhammer mitigation?",{"text":84,"@type":76},"Static strategies do not adapt to runtime memory access patterns or hardware variability, which can cause unnecessary performance and energy penalties.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]