{"id":"https://openalex.org/W4417293528","doi":"https://doi.org/10.1109/infocom59046.2026.11571335","title":"GoodSpeed: Optimizing Fair Goodput with Adaptive Speculative Decoding in Distributed Edge Inference","display_name":"GoodSpeed: Optimizing Fair Goodput with Adaptive Speculative Decoding in Distributed Edge Inference","publication_year":2026,"publication_date":"2026-05-18","ids":{"openalex":"https://openalex.org/W4417293528","doi":"https://doi.org/10.1109/infocom59046.2026.11571335"},"language":null,"primary_location":{"id":"doi:10.1109/infocom59046.2026.11571335","is_oa":false,"landing_page_url":"https://doi.org/10.1109/infocom59046.2026.11571335","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE INFOCOM 2026 - IEEE Conference on Computer Communications","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2512.09963","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5091627176","display_name":"Phuong T. Tran","orcid":"https://orcid.org/0000-0002-1448-8882"},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Phuong Tran","raw_affiliation_strings":["The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006","institution_ids":["https://openalex.org/I129604602","https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107895337","display_name":"T. Liu","orcid":null},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Tzu-Hao Liu","raw_affiliation_strings":["The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006","institution_ids":["https://openalex.org/I129604602","https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103325289","display_name":"Long Tan Le","orcid":"https://orcid.org/0000-0003-3284-1990"},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Long Tan Le","raw_affiliation_strings":["The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006","institution_ids":["https://openalex.org/I129604602","https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011434006","display_name":"Tung-Anh Nguyen","orcid":"https://orcid.org/0009-0000-0664-0218"},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Tung-Anh Nguyen","raw_affiliation_strings":["The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006","institution_ids":["https://openalex.org/I129604602","https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004129641","display_name":"Van Quan La","orcid":null},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Van Quan La","raw_affiliation_strings":["The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006","institution_ids":["https://openalex.org/I129604602","https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5009622388","display_name":"Eun-Jeong Yu","orcid":null},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Eason Yu","raw_affiliation_strings":["The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006","institution_ids":["https://openalex.org/I129604602","https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102675778","display_name":"Shu Han","orcid":null},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Han Shu","raw_affiliation_strings":["The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006","institution_ids":["https://openalex.org/I129604602","https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034052371","display_name":"Choong Seon Hong","orcid":"https://orcid.org/0000-0003-3484-7333"},"institutions":[{"id":"https://openalex.org/I35928602","display_name":"Kyung Hee University","ror":"https://ror.org/01zqcg218","country_code":"KR","type":"education","lineage":["https://openalex.org/I35928602"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Choong Seon Hong","raw_affiliation_strings":["Kyung Hee University,Department of Computer Science and Engineering,Yongin-si,Korea,17104"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Kyung Hee University,Department of Computer Science and Engineering,Yongin-si,Korea,17104","institution_ids":["https://openalex.org/I35928602"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5027558961","display_name":"Nguyen H. Tran","orcid":"https://orcid.org/0000-0001-7323-9213"},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"The University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Nguyen H. Tran","raw_affiliation_strings":["The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of Sydney,School of Computer Science,Darlington,NSW,Australia,2006","institution_ids":["https://openalex.org/I129604602","https://openalex.org/I31746571"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"10"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.13439999520778656,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.13439999520778656,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T14347","display_name":"Big Data and Digital Economy","score":0.08100000023841858,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.07909999787807465,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/goodput","display_name":"Goodput","score":0.9502999782562256},{"id":"https://openalex.org/keywords/server","display_name":"Server","score":0.521399974822998},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.48559999465942383},{"id":"https://openalex.org/keywords/speculative-multithreading","display_name":"Speculative multithreading","score":0.4697999954223633},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.43939998745918274},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.42910000681877136},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.35359999537467957},{"id":"https://openalex.org/keywords/computational-complexity-theory","display_name":"Computational complexity theory","score":0.31690001487731934}],"concepts":[{"id":"https://openalex.org/C94022561","wikidata":"https://www.wikidata.org/wiki/Q1172393","display_name":"Goodput","level":4,"score":0.9502999782562256},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8529999852180481},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.521399974822998},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.48559999465942383},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4814999997615814},{"id":"https://openalex.org/C15296174","wikidata":"https://www.wikidata.org/wiki/Q7575343","display_name":"Speculative multithreading","level":4,"score":0.4697999954223633},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.43939998745918274},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.42910000681877136},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.35359999537467957},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.31690001487731934},{"id":"https://openalex.org/C2779971761","wikidata":"https://www.wikidata.org/wiki/Q629872","display_name":"Arbiter","level":2,"score":0.3165999948978424},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2935999929904938},{"id":"https://openalex.org/C75949130","wikidata":"https://www.wikidata.org/wiki/Q848010","display_name":"Database transaction","level":2,"score":0.28369998931884766},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.27219998836517334},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2702000141143799},{"id":"https://openalex.org/C72108876","wikidata":"https://www.wikidata.org/wiki/Q844565","display_name":"Transaction processing","level":3,"score":0.26649999618530273},{"id":"https://openalex.org/C196921405","wikidata":"https://www.wikidata.org/wiki/Q786431","display_name":"Online algorithm","level":2,"score":0.2648000121116638},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.26460000872612},{"id":"https://openalex.org/C22927095","wikidata":"https://www.wikidata.org/wiki/Q1784206","display_name":"Stateful firewall","level":3,"score":0.2578999996185303},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.2563000023365021},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.25429999828338623}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1109/infocom59046.2026.11571335","is_oa":false,"landing_page_url":"https://doi.org/10.1109/infocom59046.2026.11571335","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE INFOCOM 2026 - IEEE Conference on Computer Communications","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2512.09963","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2512.09963","pdf_url":"https://arxiv.org/pdf/2512.09963","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2512.09963","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2512.09963","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2512.09963","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2512.09963","pdf_url":"https://arxiv.org/pdf/2512.09963","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1,7,121],"models":[2,45],"(LLMs)":[3],"have":[4],"revolutionized":[5],"natural":[6],"processing,":[8],"yet":[9],"their":[10],"high":[11,63],"computational":[12],"demands":[13],"pose":[14],"significant":[15],"challenges":[16],"for":[17,225],"real-time":[18],"inference,":[19],"especially":[20],"in":[21,163,199,229],"multi-user":[22],"server":[23,82,108,173],"speculative":[24,101,125,157,227],"decoding":[25,30,228],"and":[26,71,174,181,202,222],"resource-constrained":[27],"environments.":[28],"Speculative":[29],"has":[31],"emerged":[32],"as":[33],"a":[34,55,79,91,105,111,119,134,145,219],"promising":[35],"technique":[36],"to":[37,46,123,149,194],"accelerate":[38],"LLM":[39,231],"inference":[40,94,232],"by":[41,54],"using":[42],"lightweight":[43],"draft":[44,75,115,161,176],"generate":[47,124],"candidate":[48],"tokens,":[49],"which":[50],"are":[51],"subsequently":[52],"verified":[53],"larger,":[56],"more":[57],"accurate":[58],"model.":[59],"However,":[60],"ensuring":[61],"both":[62,179],"goodput":[64,98,197],"(the":[65],"effective":[66],"rate":[67],"of":[68,113],"accepted":[69],"tokens)":[70],"fairness":[72,152],"across":[73,153],"multiple":[74],"servers":[76,162],"cooperating":[77],"with":[78,206],"central":[80,106],"verification":[81,107,142,172],"remains":[83],"an":[84],"open":[85],"challenge.":[86],"This":[87],"paper":[88],"introduces":[89],"GOODSPEED,":[90],"novel":[92],"distributed":[93,175,230],"framework":[95,166],"that":[96,109,138,191,216],"optimizes":[97],"through":[99],"adaptive":[100],"decoding.":[102],"GOODSPEED":[103,132,192,217],"employs":[104],"coordinates":[110],"set":[112],"heterogeneous":[114],"servers,":[116],"each":[117],"running":[118],"small":[120],"model":[122],"tokens.":[126],"To":[127],"manage":[128],"resource":[129],"allocation":[130,198],"effectively,":[131],"incorporates":[133],"gradient":[135],"scheduling":[136],"algorithm":[137],"dynamically":[139],"assigns":[140],"token":[141],"tasks,":[143],"maximizing":[144],"logarithmic":[146],"utility":[147],"function":[148],"ensure":[150],"proportional":[151],"servers.":[154],"By":[155],"processing":[156],"outputs":[158],"from":[159],"all":[160],"parallel,":[164],"the":[165,171,195],"enables":[167],"efficient":[168,223],"collaboration":[169],"between":[170],"generators,":[177],"streamlining":[178],"latency":[180],"throughput.":[182],"Through":[183],"rigorous":[184],"fluid":[185],"sample":[186],"path":[187],"analysis,":[188],"we":[189],"show":[190],"converges":[193],"optimal":[196],"steady-state":[200],"conditions":[201],"maintains":[203],"near-optimal":[204],"performance":[205],"provably":[207],"bounded":[208],"error":[209],"under":[210],"dynamic":[211],"workloads.":[212],"These":[213],"results":[214],"demonstrate":[215],"provides":[218],"scalable,":[220],"fair":[221],"solution":[224],"multi-server":[226],"systems.":[233]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-12-13T00:00:00"}
