抄録
This paper deals with a quantized feedback stabilization problem of nonlinear networked control systems via linearization. In particular, we study circumstances where the communication channel is interrupted by Denial-of-Service (DoS) attacks and its data rate is limited. We employ a deterministic DoS attack model which constraints the amount of attacks only by their frequency and duration, allowing us to capture a large class of potential attacks. To achieve asymptotic stabilization, we propose a resilient dynamic quantizer in the sense that it does not saturate in the presence of packet losses caused by DoS attacks. A sufficient condition for stability is derived by restricting the average frequency and duration of attacks. Since our result only guarantees local stability, we explicitly investigate an estimate of the region of attraction, which may be reduced by attacks. A simulation example is presented for demonstration of our results.
| 本文言語 | English |
|---|---|
| ページ（範囲） | 3054-3059 |
| ページ数 | 6 |
| ジャーナル | IFAC-PapersOnLine |
| 巻 | 53 |
| 号 | 2 |
| DOI | |
| 出版ステータス | Published - 2020 |
| 外部発表 | はい |
| イベント | 21st IFAC World Congress 2020 - Berlin, Germany 継続期間: 2020 7月 12 → 2020 7月 17 |
ASJC Scopus subject areas
- 制御およびシステム工学
フィンガープリント
「DoS-aware quantized control of nonlinear systems via linearization」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。引用スタイル
- APA
- Standard
- Harvard
- Vancouver
- Author
- BIBTEX
- RIS
DoS-aware quantized control of nonlinear systems via linearization. / Kato, Rui; Cetinkaya, Ahmet; Ishii, Hideaki.
In: IFAC-PapersOnLine, Vol. 53, No. 2, 2020, p. 3054-3059.研究成果: Conference article › 査読
}
TY - JOUR
T1 - DoS-aware quantized control of nonlinear systems via linearization
AU - Kato, Rui
AU - Cetinkaya, Ahmet
AU - Ishii, Hideaki
N1 - Funding Information: Abstract: This paper deals wiˇh a quanˇized feedbaflk sˇabilizaˇion problem of nonlinear Abstract: This paper deals wiˇh a quanˇized feedbaflk sˇabilizaˇion problem of nonlinear neˇworked flonˇrol sysˇems via linearizaˇion. In parˇiflular, we sˇudy flirflumsˇanfles where ˇhe neˇworked flonˇrol sysˇems via linearizaˇion. In parˇiflular, we sˇudy flirflumsˇanfles where ˇhe flommuniflaˇion flhannel is inˇerrupˇed by DenialΛofΛServifle (DoS) aˇˇaflks and iˇs daˇa raˇe is flommuniflaˇion flhannel is inˇerrupˇed by DenialΛofΛServifle (DoS) aˇˇaflks and iˇs daˇa raˇe is flommuniflaˇion flhannel is inˇerrupˇed by DenialΛofΛServifle (DoS) aˇˇaflks and iˇs daˇa raˇe is only by ˇheir frequenfly and duraˇion, allowing us ˇo flapˇure a large fllass of poˇenˇial aˇˇaflks. only by ˇheir frequenfly and duraˇion, allowing us ˇo flapˇure a large fllass of poˇenˇial aˇˇaflks. only by ˇheir frequenfly and duraˇion, allowing us ˇo flapˇure a large fllass of poˇenˇial aˇˇaflks. ˇhaˇ iˇ does noˇ saˇuraˇe in ˇhe presenfle of paflkeˇ losses flaused by DoS aˇˇaflks. A suffiflienˇ To aflhieve asympˇoˇifl sˇabilizaˇion, we propose a resilienˇ dynamifl quanˇizer in ˇhe sense flondiˇion for sˇabiliˇy is derived by resˇriflˇing ˇhe average frequenfly and duraˇion of aˇˇaflks. flondiˇion for sˇabiliˇy is derived by resˇriflˇing ˇhe average frequenfly and duraˇion of aˇˇaflks. Sinfle our resulˇ only guaranˇees loflal sˇabiliˇy, we explifliˇly invesˇigaˇe an esˇimaˇe of ˇhe Sinfle our resulˇ only guaranˇees loflal sˇabiliˇy, we explifliˇly invesˇigaˇe an esˇimaˇe of ˇhe region of aˇˇraflˇion, whiflh may be redufled by aˇˇaflks. A simulaˇion example is presenˇed for region of aˇˇraflˇion, whiflh may be redufled by aˇˇaflks. A simulaˇion example is presenˇed for region of aˇˇraflˇion, whiflh may be redufled by aˇˇaflks. A simulaˇion example is presenˇed for Copyright ©demonsˇraˇion2020 Thof oue Aur resuthors. Thlˇs. is is an open access article under the CC BY-NC-ND license demonsˇraˇion of our resulˇs. ) (Khettypw:/o/crrdesa:tiDveocSomaˇmˇaoflnkss.,oqrgu/alincˇeinzseeds/fbloyn-ˇnrco-ln, dn/o4n.0linear sysˇems, sˇabiliˇy analysis, linearizaˇion. Keywords: DoS aˇˇaflks, quanˇized flonˇrol, nonlinear sysˇems, sˇabiliˇy analysis, linearizaˇion. Keywords: DoS aˇˇaflks, quanˇized flonˇrol, nonlinear sysˇems, sˇabiliˇy analysis, linearizaˇion. 1. INTRODUCTION In ˇhis paper, we sˇudy sˇabilizaˇion of nonlinear flonˇrol 1. INTRODUCTION In ˇhis paper, we sˇudy sˇabilizaˇion of nonlinear flonˇrol 1. INTRODUCTION In ˇhis paper, we sˇudy sˇabilizaˇion of nonlinear flonˇrol Neˇworked flonˇrol sysˇems have been widely sˇudied over DoS aˇˇaflks. We follow a sampledΛdaˇa flonˇrol approaflh Neˇworkedflonˇrolsysˇemshavebeenwidelysˇudiedover sDysoSˇemaˇsˇaofvlkes.r daWˇae folralˇeowlima iˇesampd fllhedanneΛdaˇlsainfloˇhenˇrolpreapsepnflroeafolhf ˇhe pasˇ several deflades (Ishii and Franflis (2002) and based on linearizaˇion. Though linearizaˇionΛbased flonˇrol ˇhe pasˇ several deflades (Ishii and Franflis (2002) and based on linearizaˇion. Though linearizaˇionΛbased flonˇrol Bemporad eˇ al. (2010)). When a flommuniflaˇion flhannel design is a ˇypiflal meˇhod in praflˇifle, only few works deal Bemporad eˇ al. (2010)). When a flommuniflaˇion flhannel design is a ˇypiflal meˇhod in praflˇifle, only few works deal Bemporad eˇ al. (2010)). When a flommuniflaˇion flhannel design is a ˇypiflal meˇhod in praflˇifle, only few works deal informaˇion exflhanged over ˇhe flhannel need ˇo be quanΛ inˇeresˇ in ˇhe flonˇexˇ of DoS aˇˇaflks, sinfle ˇhey may bring is used in flonˇrol sysˇems, measuremenˇ and flonˇrol inpuˇ wiˇh ˇhis approaflh in ˇhe liˇeraˇure. Iˇ is of parˇiflular ˇized. Moreover, iˇ beflomes neflessary ˇo invesˇigaˇe how flriˇiflal issues when flommuniflaˇion may be inˇerrupˇed by ˇized. Moreover, iˇ beflomes neflessary ˇo invesˇigaˇe how flriˇiflal issues when flommuniflaˇion may be inˇerrupˇed by ˇized. Moreover, iˇ beflomes neflessary ˇo invesˇigaˇe how flriˇiflal issues when flommuniflaˇion may be inˇerrupˇed by of ˇhe flhannel. Many researflhers ˇhus ˇaflkled suflh daˇa nonlinear sysˇem is loflally flonˇrolled under DoS aˇˇaflks. of ˇhe flhannel. Many researflhers ˇhus ˇaflkled suflh daˇa nonlinear sysˇem is loflally flonˇrolled under DoS aˇˇaflks. of ˇhe flhannel. Many researflhers ˇhus ˇaflkled suflh daˇa nonlinear sysˇem is loflally flonˇrolled under DoS aˇˇaflks. (see, e.g., Nair eˇ al. (2007) and ˇhe referenfles ˇherein). aˇˇraflˇion due ˇo DoS aˇˇaflks, ˇhen iˇ will noˇ flonverge ˇo (see, e.g., Nair eˇ al. (2007) and ˇhe referenfles ˇherein). aˇˇraflˇion due ˇo DoS aˇˇaflks, ˇhen iˇ will noˇ flonverge ˇo (Onsee,ˇhee.g.,oNaiˇherreˇhaalnd,.(2007)in reflanendˇˇyheearsref,erˇheenflesvieˇwhperoeiinnˇ).of ˇheaˇˇraequilibriumflˇion due ˇoevDoenSaafˇeˇˇarfˇhelks,flˇheomnmiˇunwiillflaˇiononˇ flisonrevesˇorgree ˇod. ˇhe equilibrium even afˇer ˇhe flommuniflaˇion is resˇored. Onflyberˇhesefluoˇheriˇyrhhaasnd,beflomein reimpflenorˇˇayneaˇrsfor, nˇheeˇwovierkwedpofloinnˇˇroolf Tˇhehisequilibriumpaper exˇendsevenˇheafˇefrarmˇheewoflorkmpremusneinflˇaeˇiod inn isKareˇosˇoeˇread.l. (2019)his paˇpoerˇaekxˇee qndsuanˇheˇizaˇfraionmeiwnˇoorkaflfloupresennˇˇ.eAld inso,Kaweˇofoleˇloawl. flyber sefluriˇy has beflome imporˇanˇ for neˇworked flonˇrol This paper exˇends ˇhe framework presenˇed in Kaˇo eˇ al. flysysˇberemssefluasrisuˇyflhhassybsˇeflomeems haimpve boreenˇanfouˇ forndneˇˇowboerkveudlnfleronabˇrolle (2019)e worˇkosˇbakyeHouquaneˇˇizaˇal.ion(1997)inˇo aflflouand Hunˇ. eˇAlso,al. w(1999)e folloiwn syˇosˇaˇemsˇaflkass (susee,flhe.syg.sˇ, Cemsárdhenavase eˇbeenal. fou(2008)nd ˇando bePvauslqnueralabeˇlˇei linearizaˇion analysis. We ˇhen derive a suffiflienˇ flondiˇion ˇeˇoaˇalˇ.a(fl2015)ks(see,fore.g.an,Coávredrveniewas)eˇ. Iˇal.h(2008)as beflomeandPfllasearqualˇeˇhaˇˇi ˇhe works by Hou eˇ al. (1997) and Hu eˇ al. (1999) in ˇoaˇˇaflks(see,e.g.,Cárdenaseˇal.(2008)andPasqualeˇˇi linearizaˇion analysis. We ˇhen derive a suffiflienˇ flondiˇion eˇboˇalh.fly(2015)ber anford panhysioflalvervaˇiewˇa)fl.ksIˇˇohasflonbˇeflomerol sysˇfllemsear mˇhaaˇy ˇhalineˇarizˇheaˇiosˇanˇeanaˇralysjeisfl.ˇoWryeˇheremnadeinsrivweiˇhinasuffiaflfleienrˇaˇflinonsˇadiˇioblen eˇ al. (2015) for an overview). Iˇ has beflome fllear ˇhaˇ ˇhaˇ ˇhe sˇaˇe ˇrajeflˇory remains wiˇhin a flerˇain sˇable boˇh flyber and physiflal aˇˇaflks ˇo flonˇrol sysˇems may ˇhaˇ ˇhe sˇaˇe ˇrajeflˇory remains wiˇhin a flerˇain sˇable bindufloˇheflyflbriˇifler analdinflphideysinflalˇs inaˇˇaˇheflksreˇoalflownoˇrld,rol syresˇsulˇingems minay, regionevenunderDoSaˇˇaflks. e.g., large finanflial losses. Aflflording ˇo Amin eˇ al. (2009), Similarly ˇo De Persis and Tesi (2015), we ˇreaˇ DoS indufle.g.,laregeflriˇiflfinanalfliinflalidelossesnˇs.inAflflorˇhedirengalˇowAmiorld,neˇresalulˇing.(2009)in,, Similarly ˇo De Persis and Tesi (2015), we ˇreaˇ DoS flyber aˇˇaflks on flonˇrol sysˇems are fllassified ˇo deception aˇˇaflks in a deˇerminisˇifl manner raˇher ˇhan a sˇoflhasˇifl flyber aˇˇaflks on flonˇrol sysˇems are fllassified ˇo deception aˇˇaflks in a deˇerminisˇifl manner raˇher ˇhan a sˇoflhasˇifl attacks, whiflh are flonduflˇed by flhanging ˇhe flonˇenˇs of one (see Ceˇinkaya eˇ al. (2019b) for more deˇailed disΛ attacks, whiflh are flonduflˇed by flhanging ˇhe flonˇenˇs of one (see Ceˇinkaya eˇ al. (2019b) for more deˇailed disΛ attacks, whiflh are flonduflˇed by flhanging ˇhe flonˇenˇs of one (see Ceˇinkaya eˇ al. (2019b) for more deˇailed disΛ refer ˇo flommuniflaˇion inˇerrupˇions inflluding jamming Persis and Tesi (2015), DoS aˇˇaflks were flharaflˇerized in paflkeˇ daˇa, and Denial-of-Service (DoS) attacks, whiflh flussion on various DoS aˇˇaflk models). In ˇhe work of De aˇˇaflks. DoS aˇˇaflks are parˇiflularly flriˇiflal as iˇ is easy ˇo ˇerms of average frequenfly and duraˇion. There, inpuˇΛˇoΛ aˇˇaflks. DoS aˇˇaflks are parˇiflularly flriˇiflal as iˇ is easy ˇo ˇerms of average frequenfly and duraˇion. There, inpuˇΛˇoΛ launflh suflh aˇˇaflks as menˇioned in Teixeira eˇ al. (2015). sˇaˇe sˇabiliˇy of linear sysˇems is invesˇigaˇed under DoS launflh suflh aˇˇaflks as menˇioned in Teixeira eˇ al. (2015). sˇaˇe sˇabiliˇy of linear sysˇems is invesˇigaˇed under DoS For ˇhis reason, we examine ˇhe effeflˇs of DoS aˇˇaflks aˇˇaflks and flondiˇions on allowable aˇˇaflk frequenfly and For ˇhis reason, we examine ˇhe effeflˇs of DoS aˇˇaflks aˇˇaflks and flondiˇions on allowable aˇˇaflk frequenfly and Fˇhroor ugˇhhois uˇreasonˇhe,pawpeer.examine ˇhe effeflˇs of DoS aˇˇaflks duraaˇˇaflˇioksnandwereflondiˇioobˇainensd.onTheallsoewflaoblendiˇioaˇˇansflkwfererequemnfladey alendss ˇhroughouˇ ˇhe paper. duraˇion were obˇained. These flondiˇions were made less ★ˇhTrohuisgwhoorukˇwˇahsespupappoerrt.edinpartbytheJSTCRESTGrantNo. flonduraserˇiovnaˇiwveereinoFbˇaenineg and.dTTheessie(2017)flondiˇiobynsusiwnegreampraeddeiflˇleorss ★ flonservaˇive in Feng and Tesi (2017) by using a prediflˇor ★ This work was supported in part by the JST CREST Grant No. ˇhaˇ esˇimaˇes inˇerrupˇed measuremenˇs. For nonlinear This work was supported in part by the JST CREST Grant No. ˇhaˇ esˇimaˇes inˇerrupˇed measuremenˇs. For nonlinear JPMJCRl5K3, by JSPS under Grant-in-Aid for Scientific Research sysˇems under DoS aˇˇaflks, De Persis and Tesi (2016) JPMJCRl5K3, by JSPS under Grant-in-Aid for Scientific Research sysˇems under DoS aˇˇaflks, De Persis and Tesi (2016) icsGrafnort NSyo.stems18H01460,DesignandProbjyectJST(No.ERJPATMOJEHRA1603).SUO Metamathemat- invesˇigaˇed a global sˇabilizaˇion problem wiˇh flerˇain ics for Systems Design Project (No. JPMJER1603). invesˇigaˇed a global sˇabilizaˇion problem wiˇh flerˇain ics for Systems Design Project (No. JPMJER1603). invesˇigaˇed a global sˇabilizaˇion problem wiˇh flerˇain Funding Information: This work was supported in part by the JST CREST Grant No. JPMJCRl5K3, by JSPS under Grant-in-Aid for Scientific Research Grant No. 18H01460, and by JST ERATO HASUO Metamathematics for Systems Design Project (No. JPMJER1603). Publisher Copyright: © 2020 Elsevier B.V.. All rights reserved.
PY - 2020
Y1 - 2020
N2 - This paper deals with a quantized feedback stabilization problem of nonlinear networked control systems via linearization. In particular, we study circumstances where the communication channel is interrupted by Denial-of-Service (DoS) attacks and its data rate is limited. We employ a deterministic DoS attack model which constraints the amount of attacks only by their frequency and duration, allowing us to capture a large class of potential attacks. To achieve asymptotic stabilization, we propose a resilient dynamic quantizer in the sense that it does not saturate in the presence of packet losses caused by DoS attacks. A sufficient condition for stability is derived by restricting the average frequency and duration of attacks. Since our result only guarantees local stability, we explicitly investigate an estimate of the region of attraction, which may be reduced by attacks. A simulation example is presented for demonstration of our results.
AB - This paper deals with a quantized feedback stabilization problem of nonlinear networked control systems via linearization. In particular, we study circumstances where the communication channel is interrupted by Denial-of-Service (DoS) attacks and its data rate is limited. We employ a deterministic DoS attack model which constraints the amount of attacks only by their frequency and duration, allowing us to capture a large class of potential attacks. To achieve asymptotic stabilization, we propose a resilient dynamic quantizer in the sense that it does not saturate in the presence of packet losses caused by DoS attacks. A sufficient condition for stability is derived by restricting the average frequency and duration of attacks. Since our result only guarantees local stability, we explicitly investigate an estimate of the region of attraction, which may be reduced by attacks. A simulation example is presented for demonstration of our results.
KW - DoS attacks
KW - Linearization
KW - Nonlinear systems
KW - Quantized control
KW - Stability analysis
UR - http://www.scopus.com/inward/record.url?scp=85105056554&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85105056554&partnerID=8YFLogxK
U2 - 10.1016/j.ifacol.2020.12.1003
DO - 10.1016/j.ifacol.2020.12.1003
M3 - Conference article
AN - SCOPUS:85105056554
SN - 2405-8963
VL - 53
SP - 3054
EP - 3059
JO - IFAC-PapersOnLine
JF - IFAC-PapersOnLine
IS - 2
T2 - 21st IFAC World Congress 2020
Y2 - 12 July 2020 through 17 July 2020
ER -