{"id":{"repo_id":"bu","oai_identifier":"oai:open.bu.edu:2144/53565"},"canonical_url":"https://search.ndltd.org/etd/bu/oai:open.bu.edu:2144/53565","repository":{"repo_id":"bu","name":"Boston University","base_url":"https://open.bu.edu/oai/request"},"display":{"title":"Fault tolerance in a system-of-systems","abstract":"Fault tolerance is essential in safety-critical systems, where even minor faults are able to lead to severe consequences such as system failure or loss of life. One such non-malicious source of faults are environment-induced hardware faults, particularly transient errors such as Single Event Upsets (SEUs) (both Single Bit Upsets (SBUs) and Multiple Bit Upsets (MBUs)) and Multiple Event Upsets (MEUs), which are increasingly significant in harsh environments with cosmic radiation, solar radiation or thermal effects. Existing approaches, such as Error-Correcting Codes (ECC) and centralized hypervisor-based supervision, face limitations including vulnerability to complex faults, higher costs, and single points of failure. This thesis proposes a software-based alternative: a decentralized, lock-free fault tolerance mechanism across both hypervisor and Real-Time Operating System (RTOS) kernel layers. The systems are managed by a partitioning hypervisor, such that each system gets its own private Virtual Machine Manager (VMM) monitor, responsible for running the fault detection mechanism by generating integrity summary hashes of code/text sections at both RTOS kernel and hypervisor level. The fault recovery, though not evaluated in this thesis, works on the concept of writing over a faulty text section with a shared validated text section. The proposed approach detects both single and multiple bit errors while eliminating reliance on a central supervisor. The evaluations show the fault detection mechanism to introduce modest overhead while maintaining timely fault detection. Specifically, for a system of 3 Quest RTOSes of 1 core each (at 2.4GHz processor speed) and 620 KB text sections, the end-to-end detection overhead was approximately 5.27ms per sandbox per periodic detection with the period being customizable and potentially could be split between different hypervisor traps, demonstrating that the mechanism remains practical for safety-critical systems.","abstract_html":"Fault tolerance is essential in safety-critical systems, where even minor faults are able to lead to severe consequences such as system failure or loss of life. One such non-malicious source of faults are environment-induced hardware faults, particularly transient errors such as Single Event Upsets (SEUs) (both Single Bit Upsets (SBUs) and Multiple Bit Upsets (MBUs)) and Multiple Event Upsets (MEUs), which are increasingly significant in harsh environments with cosmic radiation, solar radiation or thermal effects. Existing approaches, such as Error-Correcting Codes (ECC) and centralized hypervisor-based supervision, face limitations including vulnerability to complex faults, higher costs, and single points of failure. This thesis proposes a software-based alternative: a decentralized, lock-free fault tolerance mechanism across both hypervisor and Real-Time Operating System (RTOS) kernel layers. The systems are managed by a partitioning hypervisor, such that each system gets its own private Virtual Machine Manager (VMM) monitor, responsible for running the fault detection mechanism by generating integrity summary hashes of code/text sections at both RTOS kernel and hypervisor level. The fault recovery, though not evaluated in this thesis, works on the concept of writing over a faulty text section with a shared validated text section. The proposed approach detects both single and multiple bit errors while eliminating reliance on a central supervisor. The evaluations show the fault detection mechanism to introduce modest overhead while maintaining timely fault detection. Specifically, for a system of 3 Quest RTOSes of 1 core each (at 2.4GHz processor speed) and 620 KB text sections, the end-to-end detection overhead was approximately 5.27ms per sandbox per periodic detection with the period being customizable and potentially could be split between different hypervisor traps, demonstrating that the mechanism remains practical for safety-critical systems.","abstract_has_math":false,"creators":["Iyer Rajkumar, Varsha"],"institution":"Boston University","degree_name":"Master of Science","degree_level":"masters","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":[],"advisors":["West, Richard"],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026","date_published":"2026","updated_at":"2026-11-18T20:23:14Z","subjects":["Computer science","Distributed systems","Fault tolerance","Operating systems","Real-time systems"],"languages":["en_US"],"rights":["Attribution 4.0 International"],"rights_urls":["http://creativecommons.org/licenses/by/4.0/"],"identifier_entries":[],"person_identifier_entries":[{"key":"dc:identifier.orcid","label":"ORCID","values":["0009-0001-7678-8741"],"render_values":[{"text":"0009-0001-7678-8741","href":"https://orcid.org/0009-0001-7678-8741","code":true}]}]},"links":{"outbound_url":"https://hdl.handle.net/2144/53565","outbound_label":"Handle","outbound_source":"dc:identifier.uri","record_url":"https://hdl.handle.net/2144/53565","record_label":"Repository record","record_kind":"handle","record_source":"dc:identifier.uri","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://open.bu.edu/oai/request?verb=GetRecord&metadataPrefix=dim&identifier=oai%3Aopen.bu.edu%3A2144%2F53565","prefix":"dim"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["West, Richard"]},{"key":"dc:creator","label":"Author","values":["Iyer Rajkumar, Varsha"]},{"key":"dc:identifier.orcid","label":"ORCID","values":["0009-0001-7678-8741"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-09-28T18:22:44Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2026-09-28T18:22:44Z"]},{"key":"dc:date.issued","label":"Date","values":["2026"]},{"key":"dc:date.updated","label":"Dc Date Updated","values":["2026-09-10T22:15:29Z"]},{"key":"dc:type","label":"Dc Type","values":["Thesis/Dissertation"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["masters"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["Boston University"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Computer science","Distributed systems","Fault tolerance","Operating systems","Real-time systems"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en_US"]},{"key":"dc:rights","label":"Dc Rights","values":["Attribution 4.0 International"]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://creativecommons.org/licenses/by/4.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/2144/53565"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["2026"]},{"key":"dc:description.abstract","label":"Abstract","values":["Fault tolerance is essential in safety-critical systems, where even minor faults are able to lead to severe consequences such as system failure or loss of life. One such non-malicious source of faults are environment-induced hardware faults, particularly transient errors such as Single Event Upsets (SEUs) (both Single Bit Upsets (SBUs) and Multiple Bit Upsets (MBUs)) and Multiple Event Upsets (MEUs), which are increasingly significant in harsh environments with cosmic radiation, solar radiation or thermal effects. Existing approaches, such as Error-Correcting Codes (ECC) and centralized hypervisor-based supervision, face limitations including vulnerability to complex faults, higher costs, and single points of failure. This thesis proposes a software-based alternative: a decentralized, lock-free fault tolerance mechanism across both hypervisor and Real-Time Operating System (RTOS) kernel layers. The systems are managed by a partitioning hypervisor, such that each system gets its own private Virtual Machine Manager (VMM) monitor, responsible for running the fault detection mechanism by generating integrity summary hashes of code/text sections at both RTOS kernel and hypervisor level. The fault recovery, though not evaluated in this thesis, works on the concept of writing over a faulty text section with a shared validated text section. The proposed approach detects both single and multiple bit errors while eliminating reliance on a central supervisor. The evaluations show the fault detection mechanism to introduce modest overhead while maintaining timely fault detection. Specifically, for a system of 3 Quest RTOSes of 1 core each (at 2.4GHz processor speed) and 620 KB text sections, the end-to-end detection overhead was approximately 5.27ms per sandbox per periodic detection with the period being customizable and potentially could be split between different hypervisor traps, demonstrating that the mechanism remains practical for safety-critical systems."]},{"key":"dc:title","label":"Title","values":["Fault tolerance in a system-of-systems"]}]}],"canonical_facts":{"dc:contributor.advisor":["West, Richard"],"dc:creator":["Iyer Rajkumar, Varsha"],"dc:date.accessioned":["2026-09-28T18:22:44Z"],"dc:date.available":["2026-09-28T18:22:44Z"],"dc:date.issued":["2026"],"dc:date.updated":["2026-09-10T22:15:29Z"],"dc:description":["2026"],"dc:description.abstract":["Fault tolerance is essential in safety-critical systems, where even minor faults are able to lead to severe consequences such as system failure or loss of life. One such non-malicious source of faults are environment-induced hardware faults, particularly transient errors such as Single Event Upsets (SEUs) (both Single Bit Upsets (SBUs) and Multiple Bit Upsets (MBUs)) and Multiple Event Upsets (MEUs), which are increasingly significant in harsh environments with cosmic radiation, solar radiation or thermal effects. Existing approaches, such as Error-Correcting Codes (ECC) and centralized hypervisor-based supervision, face limitations including vulnerability to complex faults, higher costs, and single points of failure. This thesis proposes a software-based alternative: a decentralized, lock-free fault tolerance mechanism across both hypervisor and Real-Time Operating System (RTOS) kernel layers. The systems are managed by a partitioning hypervisor, such that each system gets its own private Virtual Machine Manager (VMM) monitor, responsible for running the fault detection mechanism by generating integrity summary hashes of code/text sections at both RTOS kernel and hypervisor level. The fault recovery, though not evaluated in this thesis, works on the concept of writing over a faulty text section with a shared validated text section. The proposed approach detects both single and multiple bit errors while eliminating reliance on a central supervisor. The evaluations show the fault detection mechanism to introduce modest overhead while maintaining timely fault detection. Specifically, for a system of 3 Quest RTOSes of 1 core each (at 2.4GHz processor speed) and 620 KB text sections, the end-to-end detection overhead was approximately 5.27ms per sandbox per periodic detection with the period being customizable and potentially could be split between different hypervisor traps, demonstrating that the mechanism remains practical for safety-critical systems."],"dc:identifier.orcid":["0009-0001-7678-8741"],"dc:identifier.uri":["https://hdl.handle.net/2144/53565"],"dc:language.iso":["en_US"],"dc:rights":["Attribution 4.0 International"],"dc:rights.uri":["http://creativecommons.org/licenses/by/4.0/"],"dc:subject":["Computer science","Distributed systems","Fault tolerance","Operating systems","Real-time systems"],"dc:title":["Fault tolerance in a system-of-systems"],"dc:type":["Thesis/Dissertation"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["masters"],"thesis:degree_name":["Master of Science"],"thesis:institution_name":["Boston University"]},"updated_at":"2026-11-18T20:23:14Z"}