[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-121649-en":3,"doc-seo-121649-105":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":4,"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},121649,1099513958762,"Logic","https://ap-avatar.wpscdn.com/avatar/1000023916a998db790?x-image-process=image/resize,m_fixed,w_180,h_180&k=1784791008015729253",8,"Research & Report","Similarity-Based Logic Locking Against Machine Learning Attacks - Preprint","Logic locking protects integrated-circuit intellectual property during outsourcing by inserting key-controlled logic such as XOR/XNOR gates, multiplexers, and LUTs. Recent oracle-less attacks using graph neural networks (GNNs) have broken several multiplexer-based locking methods by reformulating key recovery as link prediction. This work proposes SimLL, a similarity-based multiplexer locking technique that clusters topologically and functionally similar gates or wires and inserts key-controlled MUXes to equalize true and false subgraph structures, reducing attack accuracy to about 50%.","PREPRINT - accepted at Design Automation Conference (DAC), 2023 .  \nSimilarity-Based Logic Locking Against Machine  \nLearning Attacks  \nSubhajit Dutta Chowdhury, Kaixin Yang, Pierluigi Nuzzo  \nMing Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA  \nfduttacho, kaixinya, [nuzzo](nuzzog@usc.edu)[g](nuzzog@usc.edu)[@usc.edu](nuzzog@usc.edu)  \narXiv :2305 .05870v1 [ cs .CR] 10 May 2023  \nAbstract—Logic locking is a promising technique for protecting integrated circuit designs while outsourcing their fabrication. Recently, graph neural network (GNN)-based link prediction attacks have been developed which can successfully break all the multiplexer-based locking techniques that were expected to be learning-resilient. We present SimLL, a novel similaritybased locking technique which locks a design using multiplexersand shows robustness against the existing structure-exploiting oracle-less learning-based attacks. Aiming to confuse the machine learning (ML) models, SimLL introduces key-controlled multiplexers between logic gates or wires that exhibit high levels of topological and functional similarity. Empirical results show that SimLL can degrade the accuracy of existing ML-based attacks to approximately 50%, resulting in a negligible advantage over random guessing.  \nIndex Terms—Topological similarity, graph neural networks, machine learning, link prediction, hardware security  \nI. INTRODUCTION  \nLogic locking (LL) is a promising solution to protect a design's intellectual property (IP) from various threats throughout the integrated circuit (IC) supply chain [1] . LL performs functional and structural design alterations through the insertion of additional key-controlled logic such as XOR/XNOR gates, multiplexers (MUXes), or look-up tables (LUTs) [2]–[5] . Figure 1 shows examples from different LL methods. The security of LL has, however, been challenged by the development of various attacks which can be broadly classified into two categories, namely, oracle-guided (OG) and oracleless (OL) . In OG attacks [6], attackers access the locked design netlist and a functional chip, i.e., the oracle. In OL attacks [7],[8], the attackers only have the locked design netlist at their disposal. OL attacks pose greater threats to LL as they can be mounted even if an oracle is not available.  \nAmong OL attacks, those based on machine learning (ML) [7], [8] aim to predict the correct key by exploiting the information leakage occurring through the structural signatures induced by the different LL schemes. In fact, in XOR/XNORbased and AND/OR-based LL, there usually exists a direct mapping between the type of the key gate and the key value. Attacks like SAIL [7], SnapShot [8], and OMLA [9] leverage this information along with the surrounding circuitry of a key gate to uncover the key value using deep learning models. MUX-based LL tends to leak less information about the key value via the structure of the key gate. However, other attacks, like the constant propagation attacks SWEEP [10] and SCOPE [11] and the structural analysis attack on MUX-based  \nFig. 1. Logic locking techniques: (a) Overview; (b) Original netlist; (c) XOR/XNOR-based logic locking; (d) MUX-based logic locking.  \nLL (SAAM), can successfully break MUX-based locking, calling for the development of learning-resilient LL techniques, such as deceptive MUX-based (D-MUX) [12] LL and symmetric MUX-based LL [11] .  \nUnfortunately, however, a new attack based on graph neural networks (GNNs), called MuxLink [13], has been recently developed that can break D-MUX and symmetric MUX-based locking and recover the correct key with 100% accuracy for many designs from the ISCAS-85 and ITC-99 benchmarks. MuxLink maps the task of finding the correct key to a link-prediction problem, i.e., the problem of predicting the likelihood for a link (or wire) to be present in a design. In link prediction [14], a local subgraph is extracted around each (true ","cbCain46kUpizZUI","https://ap.wps.com/l/cbCain46kUpizZUI","pdf",1655989,1,6,"English","en",105,"# Introduction\n## Logic locking overview\n## Oracle-guided vs oracle-less attacks\n## Machine learning and GNN link-prediction attacks\n## Proposed SimLL similarity-based MUX locking","[{\"question\":\"What is the main security goal of logic locking in integrated circuits?\",\"answer\":\"Logic locking aims to protect a design’s intellectual property throughout the IC supply chain by altering functionality and structure using additional key-controlled logic.\"},{\"question\":\"Why can GNN-based attacks like MuxLink break multiplexer-based locking?\",\"answer\":\"MuxLink turns key recovery into a link-prediction problem, extracting local subgraphs around candidate links and training GNN layers to distinguish true from false wires.\"},{\"question\":\"How does SimLL improve robustness against learning-based attacks?\",\"answer\":\"SimLL inserts key-controlled multiplexers between gates or wires with high topological and functional similarity, so true and false links produce similar subgraph structures and learned models lose accuracy.\"}]","Similarity-Based Logic Locking Against Machine Learning Attacks - Preprint | PDF",1785805932,15,{"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},"similarity-based-logic-locking-against-machine-learning-attacks-preprint","",{"@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/similarity-based-logic-locking-against-machine-learning-attacks-preprint/121649/",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","2026-08-04",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main security goal of logic locking in integrated circuits?","Question",{"text":75,"@type":76},"Logic locking aims to protect a design’s intellectual property throughout the IC supply chain by altering functionality and structure using additional key-controlled logic.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why can GNN-based attacks like MuxLink break multiplexer-based locking?",{"text":80,"@type":76},"MuxLink turns key recovery into a link-prediction problem, extracting local subgraphs around candidate links and training GNN layers to distinguish true from false wires.",{"name":82,"@type":73,"acceptedAnswer":83},"How does SimLL improve robustness against learning-based attacks?",{"text":84,"@type":76},"SimLL inserts key-controlled multiplexers between gates or wires with high topological and functional similarity, so true and false links produce similar subgraph structures and learned models lose accuracy.","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"]