Resource Reservation for Mixed Criticality Systems
|
|
- Job Cannon
- 5 years ago
- Views:
Transcription
1 Resource Reservaton for Mxed Crtcalty Systems Guseppe Lpar 1, Gorgo C. Buttazzo 2 1 LSV, ENS - Cachan, France 2 Scuola Superore Sant Anna, Italy Abstract. Ths paper presents a reservaton-based approach to schedule mxed crtcalty systems n a way that guarantees the schedulablty of hgh-crtcalty tasks ndependently of the behavour of low-crtcalty tasks. Two key deas are presented: frst, to reduce the system uncertanty and advance the tme at whch a hgh-crtcalty task reveals ts actual executon tme, the ntal porton of ts code s handled by a dedcated server wth a bandwdth reserved for the worst-case, but wth a shorter deadlne; second, to avod the pessmsm related to off-lne budget allocaton, an effcent reclamng mechansm, namely the GRUB algorthm [6], s used to explot the budget left by hgh-crtcalty tasks n favor of those low-crtcalty tasks that can stll complete wthn ther deadlne. 1 Introducton Wth the progress of computer archtectures, embedded computng systems are requred to execute more and more concurrent actvtes on the same hardware platform. In msson-crtcal systems, computatonal actvtes may have dfferent levels of crtcalty, and therefore dfferent guarantee requrements mposed by certfcaton authortes. In partcular, more crtcal tasks are requred to have more conservatve estmatons for ther computatonal requrements, wth respect to less crtcal actvtes. Such more conservatve estmatons ncrease system predctablty by over allocatng computatonal resources to more crtcal tasks, but also decrease the overall effcency. To partally compensate for such pessmstc estmatons, Vestal [16] proposed a new task model where each task can be specfed wth dfferent levels of crtcalty, each charactersed by a dfferent computaton tme estmate, dependng on the crtcalty level: the hgher the crtcalty level, the hgher the computaton tme estmate. Slghtly dfferent models have been also proposed by other authors. In ths paper, we consder a system wth two crtcalty levels that must execute a task set Γ of n perodc or sporadc tasks. Each task τ s charactersed The research leadng to these results has receved fundng from the European Unon Seventh Framework Programme (FP7/ ) under grant agreement No
2 by a crtcalty level X, whch can be ether hgh (HI) or low (LO), a worst-case computaton tme (WCET) C (whch depends on ts crtcalty level), a perod (or mnmum nterarrval tme) T, and a relatve deadlne D. Tasks wth low crtcalty,denoted aslo-tasks(τ LO ),haveasnglewcetestmatec,whereas tasks wth hgh crtcalty, denoted as HI-tasks (τ HI ), have a normal WCET estmate C and a more conservatve one, C ov, to take overruns nto account, where C ov > C. Each task generates an nfnte sequence of jobs, τ,1, τ,2,..., where each job τ,j s charactersed by a release tme r,j, a computaton tme c,j, and an absolute deadlne d,j. The actual computaton tme requested by a job τ,j s denoted by e,j. Accordng to such a model, the task set Γ s parttoned n two subsets Γ LO and Γ HI and the mxed crtcalty (MC) feasblty problem s formulated as follows: Defnton 1. A task set Γ s MC-feasble f and only f both the followng condtons are verfed: 1. If all HI-tasks execute for no more than ther optmstc computaton tme C, then there exsts a schedule where all tasks n Γ complete wthn ther deadlne. 2. If one or more HI-tasks exceed ther optmstc computaton tme C (but not ther conservatve estmate C ov ), then there exsts a schedule where all tasks n Γ HI complete wthn ther deadlne. The problem of optmally schedulng such mxed-crtcalty systems has been shown to be hghly ntractable even under very smple system models [2]. The complexty comes from the fact that optmal schedulng decsons depend on the knowledge of the actual tasks executon tmes, whch are not known off-lne, but wll be avalable only when a task completes or exceeds ts optmstc estmate. Gven such a dependency of schedulng decsons from future knowledge, t s clear that optmalty can only be acheved by an deal clarvoyant scheduler. To better clarfy ths ssue, consder the task set reported n Table 1. X C C ov T D τ 1 LO τ 2 HI Table 1. Sample mxed crtcalty task set. As llustrated n Fgure 1, the task set s MC-feasble, snce both condtons stated n Defnton 1 are verfed. Nevertheless, Fgure 2 llustrates that no onlne algorthm can guarantee the MC-schedulablty of the task set, because for each decson taken at tme t = 0 (to schedule τ 1 or τ 2 ), there exsts a stuaton n whch a task msses ts deadlne. Ths example shows that the correct decson that produces an M C- feasble schedule can be taken only by a clarvoyant scheduler that knows (at tme t = 0) how much task τ 2 wll execute.
3 τ LO 1 τ LO 1 τ HI 2 (a) τ HI 2 (b) Fg.1. If τ 2 does not exceed C 1, Γ s feasble (a), and f τ 2 executes for C ov 2, then Γ HI s feasble. τ 1 τ 1 τ 2 τ 2 (a) (b) Fg.2. If τ 1 s scheduled at t = 0, τ 2 can mss ts deadlne (a); and f τ 2 s scheduled at t = 0, τ 1 wll mss ts deadlne (b). Contrbuton In ths paper we propose a reservaton-based approach to schedule mxed crtcalty systems n a way that guarantees the schedulablty of HI-tasks ndependently of the behavor of LO-tasks. Two key deas are presented: frst, to reduce the system uncertanty and advance the tme at whch a HI-task reveals ts actual executon tme, the ntal porton of ts code s handled by a dedcated server wth a bandwdth reserved for the worst-case, but wth a shorter deadlne; second, to avod the pessmsm related to off-lne budget allocaton, an effcent reclamng mechansm, namely the GRUB algorthm [6], s used to explot the budget left by HI-tasks n favor of those LO-tasks that can stll complete wthn ther deadlne. Paper structure The rest of the paper s organzed as follows. Secton 2 formally presents the general approach. Secton 3 presents the detals of the reservaton server. Secton 4 llustrates the smulaton results obtaned wth the proposed approach. Secton 5 presents some related work. Secton 6 concludes the paper and presents some future work. 2 General approach We consder a reservaton-based real-tme system where each task (or group of tasks) can be assgned a reservaton server that allocates a budget Q every perod P. To better explot the avalable computatonal resources, we assume that all the servers are scheduled by the Earlest Deadlne Frst(EDF) schedulng algorthm [11].
4 Snce HI-tasks must be guaranteed under all operatng condtons and we cannot know n advance how much they wll execute, each HI-task τ HI s assgned a dedcated reservaton server (HI-server) wth bandwdth α HI suffcent to satsfy ts more conservatve executon tme C ov. Therefore, each HI-task s handled by a perodc reservaton (Q,P ), where Q = C ov and P = T, thus havng a bandwdth α HI = Cov. (1) T The bandwdth remanng for servng all the LO-tasks s then: α LO = 1 τ Γ HI α HI. (2) Let U LO be the total bandwdth requred by the LO-tasks, that s, U LO = τ Γ LO C T. (3) Note that, f U LO α LO, then there s enough bandwdth to complete all LO-tasks wthn ther deadlne, even when the system runs n hgh-crtcalty mode, hence the problem s trvally solved. In ths paper, we consder the more nterestng case where U LO > α LO, so that not all LO-tasks can be guaranteed to complete before ther deadlnes n all modes. Nevertheless, when the system runs n low-crtcalty mode, that s, when all the HI-tasks do not exceed ther optmstcestmatec,weproposetouseareclamngmechansmfordstrbutng the spare bandwdth saved by HI-task to LO-tasks. In partcular, whenever a HI-job τ,j HI completes before ts optmstc estmate (e,j < C ), the followng bandwdth can be reclamed: U,j rec = Cov e,j T. (4) When the system swtches to hgh-crtcalty mode, there are two ways of dealng wth LO-tasks: they may be dropped, as proposed n most of the prevous research on mxed-crtcalty schedulng, or they can contnue executng as soft real-tme tasks, wthout nterferng wth the HI-tasks. In ths paper, we adopt ths second approach. To schedule LO-tasks, we can follow two alternatve approaches: 1. All LO-tasks are assgned a sngle reservaton server (LO-server) wth bandwdth α LO. 2. Each LO-task s assgned a dfferent server such that the sum of the bandwdths of these servers does not exceed α LO. In ths paper, these two approaches are compared to see whether any one domnates the other.
5 3 Server mechansm As dscussed n the prevous secton, we assgn each HI-task a dedcated server wth bandwdth α HI suffcent to cover the largest computatonal requrements C ov of the task. At the same tme, as soon as each job of a HI-task τ completes, the remanng budget s reclamed and assgned to the LO-tasks. To be able to reclam such an extra budget n advance, we are nterested n advancng the completon tme of HI-tasks as soon as possble. To acheve ths goal, we use a technque smlar to the EDF-VD algorthm, proposed n [4,3]. Ths technque conssts n usng two dfferent deadlnes and budgets for a HI-task, dependng on whether t completes before or after ts normal worst-case executon tme C. The followng secton descrbes how to modfy the GRUB server to acheve ths objectve. 3.1 Hgh-crtcalty servers A HI-server for a HI-task τ s defned as a tuple (Q,Q ov,p ), where Q = C, Q ov = C ov, and P = T. The server s assgned a bandwdth α HI = Q ov /P suffcent to guarantee the more conservatve executon tme. At run tme, the server manages a capacty q, a vrtual tme v, a schedulng deadlne d, and a crtcalty mode γ. Moreover, the server has an nternal state whch can be IDLE, READY, EXECUTING, RECHARGING and RELEASING. The server s ntally n the IDLE state and ts crtcalty mode s γ = LO. The state dagram for the server s shown n Fgure 3. RECHARGING Budget exhausted Recharged Arrval dspatch IDLE READY EXEC preempton Vrtual fnshng fnshng RELEASING Fg. 3. State dagram for the HI-Server. A server that s not IDLE s sad to be actve. Let A be the set of actve servers. The algorthm mantans a global varable U act = α HI. (5) The server uses the followng rules: S A
6 1. When a HI-task s actvated at tme t, the server moves from the IDLE to the READY state. Correspondngly, ts budget s replenshed at q = Q and ts deadlne s set at: d = t+ Q Q ov P = t+ Q α HI = t+ q α HI Note that such a deadlne s shorter than the usual server deadlne (t+p ). Moreover, the capacty (Q = C ) s also less than the maxmum one (Q ov = C ov ). However, the server bandwdth s stll α HI. Also, the server moves to the actve set A, therefore the value of U act s updated to U act +α HI. 2. When n READY, the HI-server wth the earlest schedulng deadlne s executed and moves to EXECUTING. 3. When n EXECUTING, the server capacty s decreased as: dq = U act dt If the system s fully utlsed and all servers are actve (U act = 1), ths translates n reducng the capacty of the server at unt rate. When the system s not fully utlzed, or when some server s IDLE, the capacty s decreased as a lower rate, so that the server can actually reclam the free system bandwdth. 4. If, whle n EXECUTING, the server s preempted by another server wth earler deadlne, t moves back to READY, and ts capacty s not decremented anymore. 5. If, whle n EXECUTING, the capacty s exhausted, then the server behaves accordng to the values of the crtcalty mode: (a) Ifthecrtcaltymodeγ = LO,thenthecapactysmmedatelyrecharged to q = Q ov Q, and the deadlne s postponed to d d + Qov Q α HI = d + q α HI (Notce that ths corresponds to the deadlne of the orgnal HI-task). Also, the crtcalty level s rased to γ HI. (b) If the crtcalty mode s γ = HI, a system excepton s rased, as a HI-task should never exceed ts budget Q ov. 6. If, whle n EXECUTING, the task completes the executon of the current job, the server moves to the RELEASING state. Correspondngly, the vrtual tme s computed for the server as follows: v d q α HI 7. The server remans n state RELEASING untl t v. At that pont, the server moves to IDLE state and the crtcalty level s set to γ LO. If t > v, then an extra capacty of α HI (v t) s donated to the frst server n the ready queue. Also, the server s removed from the set of actve servers A, and the overall utlzaton s decreased to U act U act α HI...
7 3.2 Low-crtcalty servers A smlar algorthm s used for mplementng the LO-server for servng LOtasks, wth a few modfcatons to the rules. A LO-Server s defned by only two parameters, the budget Q and the perod P. As stated n Secton 2, the LO-server s assgned a bandwdth α LO gven by Equaton (2). At run-tme, the servermanagesa capactyq, a schedulng deadlne d, and has a state. The state dagram for a LO-server s the same as a HI-server, and s shown n Fgure 3. The followng rules change: 1 If the server s IDLE, when one of the LO-tasks served by the server s actvated at tme t, the server moves from IDLE to the READY state. Correspondngly, ts budget s replenshed at the value q = Q and the server deadlne s set at: d = t+p = t+ Q α LO = t+ q α LO Notce that ths s the same equaton as the HI-server. Also, the server moves to the actve set A, therefore the value of U act s updated to U act U act +α LO. 5 If, whle n EXECUTING, the capacty s exhausted, then the server moves to the RECHARGING state and a rechargng tme s set at d. The server s suspended from executon untl the capacty replenshment. 6 If, whle n EXECUTING, the task completes the executon of the current job, and there are no more tasks n the server local queue, the server moves to the RELEASING state. Correspondngly, the vrtual tme s computed for the server as follows: v d q. α LO 8 If, whle the servers n RECHARGING state, t = d, then the serverbudget s replenshed at q Q, the server deadlne s postponed at d d +P, and the server s moved to the READY state. 9 If, whle the server s n RELEASING state, a new LO-task s actvated locally n the server, then the server moves to the READY state, wth the same capacty and deadlne An example Consder a smple MC task set consstng of only two tasks: a LO-task τ 1 = (C 1 = 4,T 1 = 6) and a HI-task τ2 HI = (C 2 = 2,C2 ov = 4,T 2 = 8). To schedule ths task set, we start by defnng a HI-server S 2 = (Q 2 = 2,Q ov 2 = 4,P 2 = 8); the server bandwdth s α HI 2 = 0.5. The frst queston s: how much budget we can reserve for τ 1? Clearly, we cannot reserve (Q 1 = 4,P 1 = 6), as there s not enough bandwdth left. Therefore, the LO-server s defned wth a bandwdth α LO 1 = 1 α HI 2 = 0.5, a perod P 1 = T 1 = 6, and a budget Q 1 = α LO 1 P 1 = 3. The schedule produced by the proposed algorthm s shown n Fgure 4.
8 τ 1 τ Fg. 4. Example of reclamaton. At tme 0, task τ 2 starts executng nsde ts server. Both servers are actve (because they both arrve at tme t = 0), so the capacty s updated at the followng rate whle t executes: dq 2 dt = Uact = 1 Therefore, when task τ 2 completes, the server has budget q 2 = 0. The server moves to state RELEASING and vrtual tme s computed as v 2 = 4. Therefore, the server remans actve untl tme t = 4, and from then t becomes nactve untl tme t = 8. At tme t = 2, task τ 1 starts executng. Its budget s ntally q 1 = 3. Agan, all servers are actve (remember that Server S 2 wll become nactve at tme 4), hence capacty s decreased at unt rate. At tme t = 4, the server S 2 becomes IDLE. Therefore, the capacty rate for task τ 1 changes. Frst of all, let us observe that at tme t = 4 ts value s q 1 = 1. The new rate s: dq 2 dt = Uact = 0.5. Ths means that, at the current rate, the task can stll execute for 2 unts of tme. That s exactly what we need to complete the task at tme t = 6. At tme t = 6, the frst nstance of task τ 1 completes, but the second one s actvated. Therefore, the server rechargests budget at q 1 = 3, and contnues to execute wth deadlne at d 1 = 12. Snce the other server s stll nactve, the rates for the vrtual tme and the deadlne do not change. At tme t = 8, task τ 2 s actvated agan and S 2 becomes actve. The current value of the capacty s q 1 = 2. Suppose the scheduler decdes to contnue executng τ 1. Therefore, the new rates s now: dq 2 dt = 1 And ths means that τ 1 can execute for two more unts of tme, completng ts executng at tme t = 10. In ths example we have seen that τ 1 s able to complete executon of all ts jobs wthn ts deadlne, even f the assgned budget s less than ts executon requrements, thanks to the reclamng mechansm.
9 However, t s stll to be understood how much capacty can be reclamed by LO-tasks. In fact, t s easy to buld an example n whch the LO-server does not receve enough extra capacty to complete all ts tasks before ther deadlnes. For example, f the LO-tasks have very short relatve deadlnes compared to the HI-tasks, they wll executed before them most of the tmes, and hence they wll not be able to reclam any capacty. Computng the amount of capacty reclamed by LO-tasks s a dffcult and open problem that wll be the subject of our future research. In ths paper we just compare two methods for schedulng LO-tasks nsde a LO-server: usng a sngle server for servng all the LO-tasks; or usng a dedcated LO-server for each LO-task. 4 Expermental results To evaluate the performance of the reclamaton algorthm presented n Secton 3, we performed a set of smulaton experments. The server algorthms have been mplemented n RTSm [12], a schedulng smulaton tool for modelng real-tme systems. In each smulaton run, we generated 4 HI-tasks and 4 LO-tasks. Computaton tmes and the perods of the HI-tasks have been randomly selected to get a cumulatve utlzaton equal to 50%: Cov = 0.5. T τ ΓHI In the frst set of experments, the perods of the HI-tasks have been chosen accordng to a unform dstrbuton n the range [1000, 5000], n multples of 100. In the second set of experments, they were chosen n the nterval [6000, 10000]. Computaton tmes n low-crtcalty mode were computed as a fxed fracton of the computaton tmes n hgh-crtcalty mode: τ Γ HI C = r C ov where r was vared between [0.2, 0.75]. Computaton tmes and perods of the LO-tasks were chosen to acheve a cumulatve utlzaton equal to In ths way, for all values of parameter r, we have: C + C 1. T T τ HI τ LO In other words, we made sure that the system s never overloaded n lowcrtcalty mode. In the frst set of experments, the perods of the LO-tasks were chosen n the range [6000, 10000], whereas n the second set of experments perods were chosen n [1000, 5000]. For each HI-task we prepared a HI-server wth bandwdth α HI = C ov /T. For the LO-tasks, we made two dfferent choces: assgnng all LO-tasks to a
10 sngle LO-server wth utlzaton equal to 50% and perod P = 100; or assgn each LO-task to a dfferent LO-server wth bandwdth proportonal to ts computatonal requrements, scaled down so that the sum of the bandwdth assgned to the LO-server was 50%. In the sngle server case, LO-tasks nsde the server were scheduled by the EDF local scheduler. For each combnaton of parameters, we generated 30 dfferent task sets wth random perods and computaton tmes. Each smulaton was run for unts of smulaton tme. We only analysed the low-crtcalty mode, to test the effectveness of the reclamng algorthm n that condton. Therefore, HI-tasks never execute more than ther optmstc computaton tme C. We measured the average number of deadlnes mssed and the tardness of the LO-tasks. Smulaton results are reported n the followng sectons. 4.1 Frst set of experments In ths frst set of experments, the perods of the HI-tasks have been chosen to be alllowerthanthe perodsofthe LO-tasks.Therefore,atleastatthebegnnng of the schedule, LO-tasks execute after HI-tasks have been completed, and so they can mmedately take advantage of the reclamed bandwdth. The deadlne mss percentage of the LO-tasks s shown n Fgure 5 for the case of separate servers, and sngle server. It s evdent that the deadlne mss percentage s very low n all cases. Even for the case of r = 0.75 (where the total system utlzaton s very close to 100%), t never goes above 5% of the total number of deadlnes. Also notce that puttng all LO-tasks n a sngle server s very effectve, as we observed 0 deadlnes mssed n all the experments Deadlne Msses of LO-tasks Separate Servers Sngle Server % Deadlne Mss Rato of executon LO/HI of HI-tasks Fg.5. Deadlne mss percentage of LO-tasks, usng dedcated servers, or a sngle cumulatve server. The LO-tasks have larger perods than the HI-tasks.
11 A very smlar trend can be observed for the tardness reported n Fgure 6. Such small values of the tardness ndcate that the reclamng mechansm s very effectve, even wth very hghly loaded systems Tardness of LO-tasks Separate Servers Sngle Server Tardness Rato of executon LO/HI of HI-tasks Fg. 6. Tardness of LO-tasks, usng dedcated servers or a sngle cumulatve server. The LO-tasks have larger perods than the HI-tasks. 4.2 Second set of experments In ths second set of experments, the perods of the HI-tasks have been chosen to be all hgher than the perods of the LO-tasks. Therefore, at least at the begnnng of the schedule, LO-tasks execute before HI-tasks, so they cannot mmedately take advantage of the reclamed bandwdth. The deadlne mss percentage of the LO-tasks s shown n Fgure 7 for the case of separate servers and sngle server. Onceagan,averysmlartrendcan be observedforthe tardnessn Fgure8. Notce that n ths case we detected very small values of the tardness for r 0.6 due to a very small number of deadlne msses (not vsble n the graphs). Whle these smulatons cannot conclusvely establsh the theoretcal performance guarantees for LO-tasks, they ndcate that t s ndeed worthwhle to perform further nvestgaton on reclamaton technques for mxed crtcalty systems. 5 Related work The problem of schedulng mxed crtcalty systems has been addressed by several authors under slghtly dfferent models and assumptons. Lakshmanan et al.[5,8,9] proposed a slack-aware approach on top of fxed prorty schedulng, provdng a schedulablty test that guarantees that all deadlnes
12 Deadlne Msses of LO-tasks Separate Servers Sngle Server 0.1 % Deadlne Mss Rato of executon LO/HI of HI-tasks Fg.7. Deadlne mss percentage of LO-tasks, usng dedcated servers, or a sngle cumulatve server. The LO-tasks have smaller perods than the HI-tasks Tardness of of LO-tasks Separate Servers Sngle Server Tardness Rato of executon LO/HI of HI-tasks Fg. 8. Tardness of LO-tasks, usng dedcated servers or a sngle cumulatve server. The LO-tasks have smaller perods than the HI-tasks.
13 of HI-tasks are met regardless of the runtme behavor of LO-tasks, provded the executon of at most one HI-task overruns ts lower WCET estmate. Baruah et. al. [1] proposed an effectve algorthm, called OCBP (Own Crtcalty Based Prorty), to schedule a set of non-recurrent jobs. Although OCBP s able to acheve the hghest speedup factor among all the fxed-job-prorty algorthms, t cannot be appled to recurrent tasks. LandBaruah[10]proposedanalgorthm,referredtoasLB,toextendOCBP to sporadc tasks, but t reles on very pessmstc schedulablty tests based on load bound condtons. Moreover, t ntroduces large run-tme overhead as t needs on-lne pseudo-polynomal prorty assgnment recomputaton. To overcome the lmtatons oflb,guan et al. [7] presentedanew algorthm,referredto as PLRS(Prorty Lst Reuse Schedulng) to schedule certfable mxed-crtcalty sporadc task systems. Pellzzon et al. [13] proposed a reservatons-based approach to ensure strong solaton among subsystems of dfferent crtcalty. Petters et. al. [14] also consdered the use of temporal solaton of subsystems for mxed-crtcalty systems, and addressed many practcal ssues n buldng such systems n realty. In general, the drawback of the resource/temporal solaton approach s that t reles on severely over-provsonng computng resources, whch may result n sgnfcant cost and energy waste. Baruah et al. proposed the EDF-VD algorthm [3,4], whch s smlar to the server mechansm proposed n ths paper. In partcular, they propose to antcpate the deadlnes of HI-tasks so that when executng n LO-crtcalty mode ther completon s antcpated. Ther formula for computng the antcpated deadlne s global,.e. t reles on the global utlsaton of HI-tasks, whereas n ths paper we propose a dfferent formula that accounts for the bandwdth of each dfferent HI-task separately and ndependently. In [3], the authors also propose a test for checkng the schedulablty of LO-tasks n LO-crtcalty mode, and compute the speed-up factor of ther algorthm to be 4/3. Santy et al. [15] propose an algorthm for lettng some of the LO-crtcalty task execute even after the system has swtched to HI-crtcalty mode, as long as ther executon does not compromse the schedulablty of HI-tasks. Also, they propose a method to reset the system crtcalty level at certan specfed dle ntervals. In ths paper, we also propose to contnue executng LO-crtcalty tasks as soft real-tme tasks even after the system swtches to HI-crtcalty: the temporal solaton mechansm guarantees that the HI-tasks wll not be nfluenced. 6 Conclusons We presented a reservaton-based approach to schedule mxed crtcalty systems n a way that guarantees the schedulablty of hgh-crtcalty tasks ndependently of the behavor of low-crtcalty tasks. Pessmsm related to off-lne budget allocaton s avoded by an effcent reclamng mechansm that explots the budget left by hgh-crtcalty tasks for those actve low-crtcalty tasks that can stll complete wthn ther deadlnes.
14 We are currently nvestgatng a test for guaranteeng the schedulablty of LO-tasks n LO-crtcalty mode, by computng a lower bound on the amount of reclamng avalable n any tme nterval. We wll then compare the schedulablty test wth the one proposed n [3] to evaluate the performance of the two approaches. References 1. S. Baruah, H. L, and L. Stouge. Towards the desgn of certfable mxed-crtcalty systems. In Proceedngs of the 16th IEEE Real-Tme and Embedded Technology and Applcatons Symposum (RTAS 2010), Sanjoy Baruah, Vncenzo Bonfac, Ganlorenzo D Angelo, Haohan L, Alberto Marchett-Spaccamela, Ncole Megow, and Leen Stouge. Schedulng real-tme mxed-crtcalty jobs. IEEE Transactons on Computers, 61(8): , Sanjoy K. Baruah, Vncenzo Bonfac, Ganlorenzo D Angelo, Haohan L, Alberto Marchett-Spaccamela, Suzanne van der Ster, and Leen Stouge. The preemptve unprocessor schedulng of mxed-crtcalty mplct-deadlne sporadc task systems. In ECRTS 12, pages , Sanjoy K. Baruah, Vncenzo Bonfac, Ganlorenzo D Angelo, Alberto Marchett- Spaccamela, Suzanne Van Der Ster, and Leen Stouge. Mxed-crtcalty schedulng of sporadc task systems. In Proceedngs of the 19th European conference on Algorthms, ESA 11, pages , Berln, Hedelberg, Sprnger-Verlag. 5. D. de Nz, K. Lakshmanan, and R. Rajkumar. On the schedulng of mxedcrtcalty real-tme task sets. In Proceedngs of the 30th IEEE Real-Tme Systems Symposum (RTSS 2009), December G.Lpar and S.K. Baruah. Greedy reclamaton of unused bandwdth n constant bandwdth servers. In IEEE Proceedngs of the 12th Euromcro Conference on Real-Tme Systems, Stokholm, Sweden, June Nan Guan, Pontus Ekberg, Martn Stgge, and Wang Y. Effectve and effcent schedulng of certfable mxed-crtcalty sporadc task systems. In Proceedngs of the 32nd IEEE Real-Tme Systems Symposum (RTSS 2011), K. Lakshmanan, D. de Nz, and R. Rajkumar. Resource allocaton n dstrbuted mxed-crtcalty cyber-physcal systems. In Proceedngs of the 30th Internatonal Conference on Dstrbuted Computng Systems (ICDCS 2010), K. Lakshmanan, D. de Nz, and R. Rajkumar. Mxed-crtcalty task synchronzaton n zero-slack schedulng. In Proceedngs of the 17th IEEE Real-Tme and Embedded Technology and Applcatons Symposum (RTAS 2011), H. L and S. Baruah. An algorthm for schedulng certfable mxedcrtcalty sporadc task systems. In Proceedngs of the 31st IEEE Real-Tme Systems Symposum (RTSS 2010), December C.L. Lu and J.W. Layland. Schedulng algorthms for multprogrammng n a hard-real-tme envronment. Journal of the Assocaton for Computng Machnery, 20(1):46 61, January Lug Palopol, Guseppe Lpar, Luca Aben, Marco D Natale, Paolo Anclott, and Fabo Contcell. A tool for smulaton and fast prototypng of embedded control systems. In Seongsoo Hong and Santosh Pande, edtors, LCTES/OM, pages ACM, R. Pellzzon, P. Meredth, M.Y. Nam, M. Sun, M. Caccamo, and L. Sha. Handlng mxed crtcalty n soc-based real-tme embedded systems. In Proceedngs of the ACM Internatonal Conference on Embedded Software (EMSOFT 2009), 2009.
15 14. S. Petters, M. Lawtzky, R. Heffernan, and K. Elphnstone. Towards real multcrtcalty schedulng. In 15th IEEE Internatonal Conference on Embedded and Real-Tme Computng Systems and Applcatons (RTCSA 2009), Franços Santy, Laurent George, Ph. Therry, and Joël Goossens. Relaxng mxedcrtcalty schedulng strctness for task sets scheduled wth fp. In IEEE Computer Socety, edtor, ECRTS, pages , July Steve Vestal. Preemptve schedulng of mult-crtcalty systems wth varyng degrees of executon tme assurance. In Proceedngs of the 28th IEEE Real-Tme Systems Symposum (RTSS 2007), pages , December 2007.
Real-Time Systems. Multiprocessor scheduling. Multiprocessor scheduling. Multiprocessor scheduling
Real-Tme Systems Multprocessor schedulng Specfcaton Implementaton Verfcaton Multprocessor schedulng -- -- Global schedulng How are tasks assgned to processors? Statc assgnment The processor(s) used for
More informationEmbedded Systems. 4. Aperiodic and Periodic Tasks
Embedded Systems 4. Aperodc and Perodc Tasks Lothar Thele 4-1 Contents of Course 1. Embedded Systems Introducton 2. Software Introducton 7. System Components 10. Models 3. Real-Tme Models 4. Perodc/Aperodc
More informationPartitioned Mixed-Criticality Scheduling on Multiprocessor Platforms
Parttoned Mxed-Crtcalty Schedulng on Multprocessor Platforms Chuanca Gu 1, Nan Guan 1,2, Qngxu Deng 1 and Wang Y 1,2 1 Northeastern Unversty, Chna 2 Uppsala Unversty, Sweden Abstract Schedulng mxed-crtcalty
More informationTwo Methods to Release a New Real-time Task
Two Methods to Release a New Real-tme Task Abstract Guangmng Qan 1, Xanghua Chen 2 College of Mathematcs and Computer Scence Hunan Normal Unversty Changsha, 410081, Chna qqyy@hunnu.edu.cn Gang Yao 3 Sebel
More informationThe Schedulability Region of Two-Level Mixed-Criticality Systems based on EDF-VD
The Schedulablty Regon of Two-Level Mxed-Crtcalty Systems based on EDF-VD Drk Müller and Alejandro Masrur Department of Computer Scence TU Chemntz, Germany Abstract The algorthm Earlest Deadlne Frst wth
More informationLimited Preemptive Scheduling for Real-Time Systems: a Survey
Lmted Preemptve Schedulng for Real-Tme Systems: a Survey Gorgo C. Buttazzo, Fellow Member, IEEE, Marko Bertogna, Senor Member, IEEE, and Gang Yao Abstract The queston whether preemptve algorthms are better
More informationFixed-Priority Multiprocessor Scheduling with Liu & Layland s Utilization Bound
Fxed-Prorty Multprocessor Schedulng wth Lu & Layland s Utlzaton Bound Nan Guan, Martn Stgge, Wang Y and Ge Yu Department of Informaton Technology, Uppsala Unversty, Sweden Department of Computer Scence
More informationQuantifying the Sub-optimality of Uniprocessor Fixed Priority Pre-emptive Scheduling for Sporadic Tasksets with Arbitrary Deadlines
Quantfyng the Sub-optmalty of Unprocessor Fxed Prorty Pre-emptve Schedulng for Sporadc Tasksets wth Arbtrary Deadlnes Robert Davs, Sanjoy Baruah, Thomas Rothvoss, Alan Burns To cte ths verson: Robert Davs,
More informationImproving the Sensitivity of Deadlines with a Specific Asynchronous Scenario for Harmonic Periodic Tasks scheduled by FP
Improvng the Senstvty of Deadlnes wth a Specfc Asynchronous Scenaro for Harmonc Perodc Tasks scheduled by FP P. Meumeu Yoms, Y. Sorel, D. de Rauglaudre AOSTE Project-team INRIA Pars-Rocquencourt Le Chesnay,
More informationLast Time. Priority-based scheduling. Schedulable utilization Rate monotonic rule: Keep utilization below 69% Static priorities Dynamic priorities
Last Tme Prorty-based schedulng Statc prortes Dynamc prortes Schedulable utlzaton Rate monotonc rule: Keep utlzaton below 69% Today Response tme analyss Blockng terms Prorty nverson And solutons Release
More informationResource Allocation with a Budget Constraint for Computing Independent Tasks in the Cloud
Resource Allocaton wth a Budget Constrant for Computng Independent Tasks n the Cloud Wemng Sh and Bo Hong School of Electrcal and Computer Engneerng Georga Insttute of Technology, USA 2nd IEEE Internatonal
More informationImproved Worst-Case Response-Time Calculations by Upper-Bound Conditions
Improved Worst-Case Response-Tme Calculatons by Upper-Bound Condtons Vctor Pollex, Steffen Kollmann, Karsten Albers and Frank Slomka Ulm Unversty Insttute of Embedded Systems/Real-Tme Systems {frstname.lastname}@un-ulm.de
More informationLecture 4: November 17, Part 1 Single Buffer Management
Lecturer: Ad Rosén Algorthms for the anagement of Networs Fall 2003-2004 Lecture 4: November 7, 2003 Scrbe: Guy Grebla Part Sngle Buffer anagement In the prevous lecture we taled about the Combned Input
More informationModule 9. Lecture 6. Duality in Assignment Problems
Module 9 1 Lecture 6 Dualty n Assgnment Problems In ths lecture we attempt to answer few other mportant questons posed n earler lecture for (AP) and see how some of them can be explaned through the concept
More informationEDF Scheduling for Identical Multiprocessor Systems
EDF Schedulng for dentcal Multprocessor Systems Maro Bertogna Unversty of Modena, taly As Moore s law goes on Number of transstor/chp doubles every 18 to 24 mm heatng becomes a problem Power densty (W/cm
More informationMixed-Criticality Scheduling with I/O
Mxed-Crtcalty Schedulng wth I/O Erc Mssmer, Katherne Mssmer and Rchard West Computer Scence Department Boston Unversty Boston, USA Emal: mssmer,kzhao,rchwest}@cs.bu.edu Abstract Ths paper addresses the
More informationEnergy-Efficient Scheduling Fixed-Priority tasks with Preemption Thresholds on Variable Voltage Processors
Energy-Effcent Schedulng Fxed-Prorty tasks wth Preempton Thresholds on Varable Voltage Processors XaoChuan He, Yan Ja Insttute of Network Technology and Informaton Securty School of Computer Scence Natonal
More informationOn the Scheduling of Mixed-Criticality Real-Time Task Sets
On the Schedulng of Mxed-Crtcalty Real-Tme Task Sets Donso de Nz, Karthk Lakshmanan, and Ragunathan (Raj) Rajkumar Carnege Mellon Unversty, Pttsburgh, PA - 15232 Abstract The functonal consoldaton nduced
More informationReal-Time Operating Systems M. 11. Real-Time: Periodic Task Scheduling
Real-Tme Operatng Systems M 11. Real-Tme: Perodc Task Schedulng Notce The course materal ncludes sldes downloaded from:! http://codex.cs.yale.edu/av/os-book/! and! (sldes by Slberschatz, Galvn, and Gagne,
More informationModule 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur
Module 3 LOSSY IMAGE COMPRESSION SYSTEMS Verson ECE IIT, Kharagpur Lesson 6 Theory of Quantzaton Verson ECE IIT, Kharagpur Instructonal Objectves At the end of ths lesson, the students should be able to:
More informationECE559VV Project Report
ECE559VV Project Report (Supplementary Notes Loc Xuan Bu I. MAX SUM-RATE SCHEDULING: THE UPLINK CASE We have seen (n the presentaton that, for downlnk (broadcast channels, the strategy maxmzng the sum-rate
More informationGlobal EDF Scheduling for Parallel Real-Time Tasks
Washngton Unversty n St. Lous Washngton Unversty Open Scholarshp Engneerng and Appled Scence Theses & Dssertatons Engneerng and Appled Scence Sprng 5-15-2014 Global EDF Schedulng for Parallel Real-Tme
More informationFixed-Priority Multiprocessor Scheduling with Liu & Layland s Utilization Bound
Fxed-Prorty Multprocessor Schedulng wth Lu & Layland s Utlzaton Bound Nan Guan, Martn Stgge, Wang Y and Ge Yu Department of Informaton Technology, Uppsala Unversty, Sweden Department of Computer Scence
More informationQuantifying the Sub-optimality of Uniprocessor Fixed Priority Non-Pre-emptive Scheduling
Quantfyng the Sub-optmalty of Unprocessor Fxed Prorty Non-Pre-emptve Schedulng Robert I Davs Real-Tme Systems Research Group, Department of Computer Scence, Unversty of York, York, UK robdavs@csyorkacuk
More informationUtilization-Based Scheduling of Flexible Mixed-Criticality Real-Time Tasks
1 Utlzaton-Based Schedulng of Flexble Mxed-Crtcalty Real-Tme Tasks Gang Chen, Nan Guan, D Lu, Qngqang He, Ka Huang, Todor Stefanov, Wang Y arxv:1711.00100v1 [cs.dc] 29 Sep 2017 Abstract Mxed-crtcalty models
More informationAN EXTENDIBLE APPROACH FOR ANALYSING FIXED PRIORITY HARD REAL-TIME TASKS
AN EXENDIBLE APPROACH FOR ANALYSING FIXED PRIORIY HARD REAL-IME ASKS K. W. ndell 1 Department of Computer Scence, Unversty of York, England YO1 5DD ABSRAC As the real-tme computng ndustry moves away from
More informationEnergy and Feasibility Optimal Global Scheduling Framework on big.little platforms
Energy and Feasblty Optmal Global Schedulng Framework on bg.little platforms Hoon Sung Chwa, Jaebaek Seo, Hyuck Yoo Jnkyu Lee, Insk Shn Department of Computer Scence, KAIST, Republc of Korea Department
More informationHandling Overload (G. Buttazzo, Hard Real-Time Systems, Ch. 9) Causes for Overload
PS-663: Real-Te Systes Handlng Overloads Handlng Overload (G Buttazzo, Hard Real-Te Systes, h 9) auses for Overload Bad syste desgn eg poor estaton of worst-case executon tes Sultaneous arrval of unexpected
More informationA Robust Method for Calculating the Correlation Coefficient
A Robust Method for Calculatng the Correlaton Coeffcent E.B. Nven and C. V. Deutsch Relatonshps between prmary and secondary data are frequently quantfed usng the correlaton coeffcent; however, the tradtonal
More informationImproving the Quality of Control of Periodic Tasks Scheduled by FP with an Asynchronous Approach
Improvng the Qualty of Control of Perodc Tasks Scheduled by FP wth an Asynchronous Approach P. Meumeu Yoms, L. George, Y. Sorel, D. de Rauglaudre AOSTE Project-team INRIA Pars-Rocquencourt Le Chesnay,
More informationCollege of Computer & Information Science Fall 2009 Northeastern University 20 October 2009
College of Computer & Informaton Scence Fall 2009 Northeastern Unversty 20 October 2009 CS7880: Algorthmc Power Tools Scrbe: Jan Wen and Laura Poplawsk Lecture Outlne: Prmal-dual schema Network Desgn:
More informationProblem Set 9 Solutions
Desgn and Analyss of Algorthms May 4, 2015 Massachusetts Insttute of Technology 6.046J/18.410J Profs. Erk Demane, Srn Devadas, and Nancy Lynch Problem Set 9 Solutons Problem Set 9 Solutons Ths problem
More informationarxiv: v3 [cs.os] 12 Mar 2016
Mxed-Crtcalty Schedulng wth I/O Erc Mssmer, Katherne Zhao and Rchard West Computer Scence Department Boston Unversty Boston, MA 02215 Emal: mssmer,kzhao,rchwest}@cs.bu.edu arxv:1512.07654v3 [cs.os] 12
More informationChapter 13: Multiple Regression
Chapter 13: Multple Regresson 13.1 Developng the multple-regresson Model The general model can be descrbed as: It smplfes for two ndependent varables: The sample ft parameter b 0, b 1, and b are used to
More informationCOMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS
Avalable onlne at http://sck.org J. Math. Comput. Sc. 3 (3), No., 6-3 ISSN: 97-537 COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS
More informationLecture 4. Instructor: Haipeng Luo
Lecture 4 Instructor: Hapeng Luo In the followng lectures, we focus on the expert problem and study more adaptve algorthms. Although Hedge s proven to be worst-case optmal, one may wonder how well t would
More informationEconomics 101. Lecture 4 - Equilibrium and Efficiency
Economcs 0 Lecture 4 - Equlbrum and Effcency Intro As dscussed n the prevous lecture, we wll now move from an envronment where we looed at consumers mang decsons n solaton to analyzng economes full of
More informationNP-Completeness : Proofs
NP-Completeness : Proofs Proof Methods A method to show a decson problem Π NP-complete s as follows. (1) Show Π NP. (2) Choose an NP-complete problem Π. (3) Show Π Π. A method to show an optmzaton problem
More informationmultiprogrammed, hard real-time environments Giuseppe Lipari John Carpenter Sanjoy Baruah particular server.
A framework for achevng nter-applcaton solaton n multprogrammed, hard real-tme envronments Guseppe Lpar John Carpenter Sanjoy Baruah Abstract A framework for schedulng a number of derent real-tme applcatons
More informationCalculation of time complexity (3%)
Problem 1. (30%) Calculaton of tme complexty (3%) Gven n ctes, usng exhaust search to see every result takes O(n!). Calculaton of tme needed to solve the problem (2%) 40 ctes:40! dfferent tours 40 add
More informationParametric fractional imputation for missing data analysis. Jae Kwang Kim Survey Working Group Seminar March 29, 2010
Parametrc fractonal mputaton for mssng data analyss Jae Kwang Km Survey Workng Group Semnar March 29, 2010 1 Outlne Introducton Proposed method Fractonal mputaton Approxmaton Varance estmaton Multple mputaton
More informationOverhead-Aware Compositional Analysis of Real-Time Systems
Overhead-Aware ompostonal Analyss of Real-Tme Systems Lnh T.X. Phan, Meng Xu, Jaewoo Lee, nsup Lee, Oleg Sokolsky PRESE enter Department of omputer and nformaton Scence Unversty of Pennsylvana ompostonal
More informationSingle-Facility Scheduling over Long Time Horizons by Logic-based Benders Decomposition
Sngle-Faclty Schedulng over Long Tme Horzons by Logc-based Benders Decomposton Elvn Coban and J. N. Hooker Tepper School of Busness, Carnege Mellon Unversty ecoban@andrew.cmu.edu, john@hooker.tepper.cmu.edu
More informationResource-Efficient Scheduling Of Multiprocessor Mixed-Criticality Real-Time Systems
Unversty of Pennsylvana ScholarlyCommons Publcly Accessble Penn Dssertatons 2017 Resource-Effcent Schedulng Of Multprocessor Mxed-Crtcalty Real-Tme Systems Jaewoo Lee Unversty of Pennsylvana, jaewoo@cs.upenn.edu
More informationQueueing Networks II Network Performance
Queueng Networks II Network Performance Davd Tpper Assocate Professor Graduate Telecommuncatons and Networkng Program Unversty of Pttsburgh Sldes 6 Networks of Queues Many communcaton systems must be modeled
More informationThe optimal delay of the second test is therefore approximately 210 hours earlier than =2.
THE IEC 61508 FORMULAS 223 The optmal delay of the second test s therefore approxmately 210 hours earler than =2. 8.4 The IEC 61508 Formulas IEC 61508-6 provdes approxmaton formulas for the PF for smple
More informationISSN: ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT) Volume 3, Issue 1, July 2013
ISSN: 2277-375 Constructon of Trend Free Run Orders for Orthogonal rrays Usng Codes bstract: Sometmes when the expermental runs are carred out n a tme order sequence, the response can depend on the run
More informationComputer Control: Task Synchronisation in Dynamic Priority Scheduling
Computer Control: Task Synchronsaton n Dynamc Prorty Schedulng Sérgo Adrano Fernandes Lopes Department of Industral Electroncs Engneerng School Unversty of Mnho Campus de Azurém 4800 Gumarães - PORTUGAL
More information18.1 Introduction and Recap
CS787: Advanced Algorthms Scrbe: Pryananda Shenoy and Shjn Kong Lecturer: Shuch Chawla Topc: Streamng Algorthmscontnued) Date: 0/26/2007 We contnue talng about streamng algorthms n ths lecture, ncludng
More informationA Simple Inventory System
A Smple Inventory System Lawrence M. Leems and Stephen K. Park, Dscrete-Event Smulaton: A Frst Course, Prentce Hall, 2006 Hu Chen Computer Scence Vrgna State Unversty Petersburg, Vrgna February 8, 2017
More informationCHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE
CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE Analytcal soluton s usually not possble when exctaton vares arbtrarly wth tme or f the system s nonlnear. Such problems can be solved by numercal tmesteppng
More informationClock-Driven Scheduling (in-depth) Cyclic Schedules: General Structure
CPSC-663: Real-me Systems n-depth Precompute statc schedule o-lne e.g. at desgn tme: can aord expensve algorthms. Idle tmes can be used or aperodc jobs. Possble mplementaton: able-drven Schedulng table
More informationIntroduction to Vapor/Liquid Equilibrium, part 2. Raoult s Law:
CE304, Sprng 2004 Lecture 4 Introducton to Vapor/Lqud Equlbrum, part 2 Raoult s Law: The smplest model that allows us do VLE calculatons s obtaned when we assume that the vapor phase s an deal gas, and
More informationParallel Real-Time Scheduling of DAGs
Washngton Unversty n St. Lous Washngton Unversty Open Scholarshp All Computer Scence and Engneerng Research Computer Scence and Engneerng Report Number: WUCSE-013-5 013 Parallel Real-Tme Schedulng of DAGs
More informationStructure and Drive Paul A. Jensen Copyright July 20, 2003
Structure and Drve Paul A. Jensen Copyrght July 20, 2003 A system s made up of several operatons wth flow passng between them. The structure of the system descrbes the flow paths from nputs to outputs.
More informationAn Admission Control Algorithm in Cloud Computing Systems
An Admsson Control Algorthm n Cloud Computng Systems Authors: Frank Yeong-Sung Ln Department of Informaton Management Natonal Tawan Unversty Tape, Tawan, R.O.C. ysln@m.ntu.edu.tw Yngje Lan Management Scence
More informationPartitioned Scheduling of Multi-Modal Mixed-Criticality Real-Time Systems on Multiprocessor Platforms
Parttoned Schedulng of Mult-Modal Mxed-Crtcalty Real-Tme Systems on Multprocessor Platforms Donso de Nz SEI, Carnege Mellon Unversty Lnh T.X. Phan Unversty of Pennsylvana Abstract Real-tme systems are
More informationSimultaneous Optimization of Berth Allocation, Quay Crane Assignment and Quay Crane Scheduling Problems in Container Terminals
Smultaneous Optmzaton of Berth Allocaton, Quay Crane Assgnment and Quay Crane Schedulng Problems n Contaner Termnals Necat Aras, Yavuz Türkoğulları, Z. Caner Taşkın, Kuban Altınel Abstract In ths work,
More informationOn the correction of the h-index for career length
1 On the correcton of the h-ndex for career length by L. Egghe Unverstet Hasselt (UHasselt), Campus Depenbeek, Agoralaan, B-3590 Depenbeek, Belgum 1 and Unverstet Antwerpen (UA), IBW, Stadscampus, Venusstraat
More informationNUMERICAL DIFFERENTIATION
NUMERICAL DIFFERENTIATION 1 Introducton Dfferentaton s a method to compute the rate at whch a dependent output y changes wth respect to the change n the ndependent nput x. Ths rate of change s called the
More informationParametric Utilization Bounds for Fixed-Priority Multiprocessor Scheduling
2012 IEEE 26th Internatonal Parallel and Dstrbuted Processng Symposum Parametrc Utlzaton Bounds for Fxed-Prorty Multprocessor Schedulng Nan Guan 1,2, Martn Stgge 1, Wang Y 1,2 and Ge Yu 2 1 Uppsala Unversty,
More informationNotes on Frequency Estimation in Data Streams
Notes on Frequency Estmaton n Data Streams In (one of) the data streamng model(s), the data s a sequence of arrvals a 1, a 2,..., a m of the form a j = (, v) where s the dentty of the tem and belongs to
More informationO-line Temporary Tasks Assignment. Abstract. In this paper we consider the temporary tasks assignment
O-lne Temporary Tasks Assgnment Yoss Azar and Oded Regev Dept. of Computer Scence, Tel-Avv Unversty, Tel-Avv, 69978, Israel. azar@math.tau.ac.l??? Dept. of Computer Scence, Tel-Avv Unversty, Tel-Avv, 69978,
More informationResource Sharing. CSCE 990: Real-Time Systems. Steve Goddard. Resources & Resource Access Control (Chapter 8 of Liu)
CSCE 990: Real-Tme Systems Resource Sharng Steve Goddard goddard@cse.unl.edu http://www.cse.unl.edu/~goddard/courses/realtmesystems Resources & Resource Access Control (Chapter 8 of Lu) Real-Tme Systems
More informationEfficient Feasibility Analysis for Real-Time Systems with EDF scheduling*
Effcent Feasblty Analyss for Real-Tme Systems wth EF schedulng* Karsten Albers, Frank Slomka epartment of omputer Scence Unversty of Oldenburg Ammerländer Heerstraße 114-118 26111 Oldenburg, Germany {albers,
More informationDurban Watson for Testing the Lack-of-Fit of Polynomial Regression Models without Replications
Durban Watson for Testng the Lack-of-Ft of Polynomal Regresson Models wthout Replcatons Ruba A. Alyaf, Maha A. Omar, Abdullah A. Al-Shha ralyaf@ksu.edu.sa, maomar@ksu.edu.sa, aalshha@ksu.edu.sa Department
More informationProcrastination Scheduling for Fixed-Priority Tasks with Preemption Thresholds
Procrastnaton Schedulng for Fxed-Prorty Tasks wth Preempton Thresholds XaoChuan He, Yan Ja Insttute of Network Technology and Informaton Securty School of Computer Scence Natonal Unversty of Defense Technology
More informationTwo-Phase Low-Energy N-Modular Redundancy for Hard Real-Time Multi-Core Systems
1 Two-Phase Low-Energy N-Modular Redundancy for Hard Real-Tme Mult-Core Systems Mohammad Saleh, Alreza Ejlal, and Bashr M. Al-Hashm, Fellow, IEEE Abstract Ths paper proposes an N-modular redundancy (NMR)
More informationLecture 14: Bandits with Budget Constraints
IEOR 8100-001: Learnng and Optmzaton for Sequental Decson Makng 03/07/16 Lecture 14: andts wth udget Constrants Instructor: Shpra Agrawal Scrbed by: Zhpeng Lu 1 Problem defnton In the regular Mult-armed
More informationKernel Methods and SVMs Extension
Kernel Methods and SVMs Extenson The purpose of ths document s to revew materal covered n Machne Learnng 1 Supervsed Learnng regardng support vector machnes (SVMs). Ths document also provdes a general
More informationQuantifying the Exact Sub-Optimality of Non-Preemptive Scheduling
Quantfyng the Exact Sub-Optmalty of Non-Preemptve Schedulng Robert I. Davs 1, Abhlash Thekklakattl 2, Olver Gettngs 1, Radu Dobrn 2, and Saskumar Punnekkat 2 1 Real-Tme Systems Research Group, Unversty
More informationPartitioned EDF Scheduling in Multicore systems with Quality of Service constraints
Parttoned EDF Schedulng n Multcore systems wth Qualty of Servce constrants Nadne Abdallah, Audrey Queudet, Marylne Chetto, Rafc Hage Chehade To cte ths verson: Nadne Abdallah, Audrey Queudet, Marylne Chetto,
More informationExperience with Automatic Generation Control (AGC) Dynamic Simulation in PSS E
Semens Industry, Inc. Power Technology Issue 113 Experence wth Automatc Generaton Control (AGC) Dynamc Smulaton n PSS E Lu Wang, Ph.D. Staff Software Engneer lu_wang@semens.com Dngguo Chen, Ph.D. Staff
More informationScheduling Motivation
76 eal-me & Embedded Systems 7 Uwe. Zmmer - he Australan Natonal Unversty 78 Motvaton n eal-me Systems Concurrency may lead to non-determnsm. Non-determnsm may make t harder to predct the tmng behavour.
More informationSynchronization Protocols. Task Allocation Bin-Packing Heuristics: First-Fit Subtasks assigned in arbitrary order To allocate a new subtask T i,j
End-to-End Schedulng Framework 1. Tak allocaton: bnd tak to proceor 2. Synchronzaton protocol: enforce precedence contrant 3. Subdeadlne agnment 4. Schedulablty analy Tak Allocaton Bn-Packng eurtc: Frt-Ft
More informationjitter Abstract Output jitter the variation in the inter-completion times of successive jobs of the same
Schedulng perodc task systems to mnmze output jtter Sanjoy Baruah Gorgo Buttazzo y Sergey Gornsky z Guseppe Lpar y Abstract Output jtter the varaton n the nter-completon tmes of successve jobs of the same
More informationFinding Dense Subgraphs in G(n, 1/2)
Fndng Dense Subgraphs n Gn, 1/ Atsh Das Sarma 1, Amt Deshpande, and Rav Kannan 1 Georga Insttute of Technology,atsh@cc.gatech.edu Mcrosoft Research-Bangalore,amtdesh,annan@mcrosoft.com Abstract. Fndng
More informationReal-Time Scheduling of Parallel Tasks under a General DAG Model
Washngton Unversty n St. Lous Washngton Unversty Open Scholarshp All Computer Scence and Engneerng Research Computer Scence and Engneerng Report Number: WUCSE-01-14 01 Real-Tme Schedulng of Parallel Tasks
More informationMILP-based Deadline Assignment for End-to-End Flows in Distributed Real-Time Systems
MILP-based Deadlne Assgnment for End-to-End Flows n Dstrbuted Real-Tme Systems Bo Peng Department of Computer Scence Wayne State Unversty Detrot, Mchgan, USA et7889@wayne.edu Nathan Fsher Department of
More informationWinter 2008 CS567 Stochastic Linear/Integer Programming Guest Lecturer: Xu, Huan
Wnter 2008 CS567 Stochastc Lnear/Integer Programmng Guest Lecturer: Xu, Huan Class 2: More Modelng Examples 1 Capacty Expanson Capacty expanson models optmal choces of the tmng and levels of nvestments
More informationOn the Multicriteria Integer Network Flow Problem
BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 5, No 2 Sofa 2005 On the Multcrtera Integer Network Flow Problem Vassl Vasslev, Marana Nkolova, Maryana Vassleva Insttute of
More informationMin Cut, Fast Cut, Polynomial Identities
Randomzed Algorthms, Summer 016 Mn Cut, Fast Cut, Polynomal Identtes Instructor: Thomas Kesselhem and Kurt Mehlhorn 1 Mn Cuts n Graphs Lecture (5 pages) Throughout ths secton, G = (V, E) s a mult-graph.
More informationThe Minimum Universal Cost Flow in an Infeasible Flow Network
Journal of Scences, Islamc Republc of Iran 17(2): 175-180 (2006) Unversty of Tehran, ISSN 1016-1104 http://jscencesutacr The Mnmum Unversal Cost Flow n an Infeasble Flow Network H Saleh Fathabad * M Bagheran
More informationAppendix B: Resampling Algorithms
407 Appendx B: Resamplng Algorthms A common problem of all partcle flters s the degeneracy of weghts, whch conssts of the unbounded ncrease of the varance of the mportance weghts ω [ ] of the partcles
More informationDepartment of Quantitative Methods & Information Systems. Time Series and Their Components QMIS 320. Chapter 6
Department of Quanttatve Methods & Informaton Systems Tme Seres and Ther Components QMIS 30 Chapter 6 Fall 00 Dr. Mohammad Zanal These sldes were modfed from ther orgnal source for educatonal purpose only.
More informationChapter - 2. Distribution System Power Flow Analysis
Chapter - 2 Dstrbuton System Power Flow Analyss CHAPTER - 2 Radal Dstrbuton System Load Flow 2.1 Introducton Load flow s an mportant tool [66] for analyzng electrcal power system network performance. Load
More informationAnalysis of Queuing Delay in Multimedia Gateway Call Routing
Analyss of Queung Delay n Multmeda ateway Call Routng Qwe Huang UTtarcom Inc, 33 Wood Ave. outh Iseln, NJ 08830, U..A Errol Lloyd Computer Informaton cences Department, Unv. of Delaware, Newark, DE 976,
More informationKeynote: RTNS Getting ones priorities right
Keynote: RTNS 2012 Gettng ones prortes rght Robert Davs Real-Tme Systems Research Group, Unversty of York rob.davs@york.ac.uk What s ths talk about? Fxed Prorty schedulng n all ts guses Pre-emptve, non-pre-emptve,
More informationAmiri s Supply Chain Model. System Engineering b Department of Mathematics and Statistics c Odette School of Business
Amr s Supply Chan Model by S. Ashtab a,, R.J. Caron b E. Selvarajah c a Department of Industral Manufacturng System Engneerng b Department of Mathematcs Statstcs c Odette School of Busness Unversty of
More informationA Hybrid Variational Iteration Method for Blasius Equation
Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 1932-9466 Vol. 10, Issue 1 (June 2015), pp. 223-229 Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) A Hybrd Varatonal Iteraton Method
More informationNumerical Heat and Mass Transfer
Master degree n Mechancal Engneerng Numercal Heat and Mass Transfer 06-Fnte-Dfference Method (One-dmensonal, steady state heat conducton) Fausto Arpno f.arpno@uncas.t Introducton Why we use models and
More informationUsing non-preemptive regions and path modification to improve schedulability of real-time traffic over priority-based NoCs
Real-Tme Syst (2017) 53:886 915 DOI 10.1007/s11241-017-9276-5 Usng non-preemptve regons and path modfcaton to mprove schedulablty of real-tme traffc over prorty-based NoCs Meng Lu 1 Matthas Becker 1 Mors
More informationILP models for the allocation of recurrent workloads upon heterogeneous multiprocessors
Journal of Schedulng DOI 10.1007/s10951-018-0593-x ILP models for the allocaton of recurrent workloads upon heterogeneous multprocessors Sanjoy K. Baruah Vncenzo Bonfac Renato Brun Alberto Marchett-Spaccamela
More informationStatistics for Economics & Business
Statstcs for Economcs & Busness Smple Lnear Regresson Learnng Objectves In ths chapter, you learn: How to use regresson analyss to predct the value of a dependent varable based on an ndependent varable
More informationImprovements in the configuration of Posix b scheduling
Improvements n the confguraton of Posx 1003.1b schedulng Matheu Grener, Ncolas Navet To cte ths verson: Matheu Grener, Ncolas Navet. Improvements n the confguraton of Posx 1003.1b schedulng. Ncolas Navet
More informationEstimation: Part 2. Chapter GREG estimation
Chapter 9 Estmaton: Part 2 9. GREG estmaton In Chapter 8, we have seen that the regresson estmator s an effcent estmator when there s a lnear relatonshp between y and x. In ths chapter, we generalzed the
More informationCopyright 2017 by Taylor Enterprises, Inc., All Rights Reserved. Adjusted Control Limits for P Charts. Dr. Wayne A. Taylor
Taylor Enterprses, Inc. Control Lmts for P Charts Copyrght 2017 by Taylor Enterprses, Inc., All Rghts Reserved. Control Lmts for P Charts Dr. Wayne A. Taylor Abstract: P charts are used for count data
More informationCopyright 2017 by Taylor Enterprises, Inc., All Rights Reserved. Adjusted Control Limits for U Charts. Dr. Wayne A. Taylor
Taylor Enterprses, Inc. Adjusted Control Lmts for U Charts Copyrght 207 by Taylor Enterprses, Inc., All Rghts Reserved. Adjusted Control Lmts for U Charts Dr. Wayne A. Taylor Abstract: U charts are used
More informationOptimal Static Partition Configuration in ARINC653 System
JOURNAL OF ELECTRONIC SCIENCE AND TECHNOLOGY, VOL. 9, NO. 4, DECEMBER 7 Optmal Statc rtton Confguraton n ARINC6 System Sheng-Ln Gu, Le Luo, Sen-Sen Tang, and Yang Meng Abstract ARINC6 systems, whch have
More informationChapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems
Numercal Analyss by Dr. Anta Pal Assstant Professor Department of Mathematcs Natonal Insttute of Technology Durgapur Durgapur-713209 emal: anta.bue@gmal.com 1 . Chapter 5 Soluton of System of Lnear Equatons
More information