An Online Delay Efficient Multi-Class Packet Scheduler for Heterogeneous M2M Uplink Traffic
|
|
- Laureen Henderson
- 5 years ago
- Views:
Transcription
1 An Onlne Delay Effcent Mult-Class Packet Scheduler for Heterogeneous M2M Uplnk Traffc Akshay Kumar, Ahmed Abdelhad and Charles Clancy Hume Center, Vrgna Tech Emal:{akshay2, aabdelhad, Abstract The sensory traffc n Machne-to-Machne (M2M) communcatons exhbt heterogenety n several dmensons such as packet arrval rate, delay tolerance requrements etc. However, the desgn of most of the exstng M2M packet schedulers do not account for ths ntrnsc heterogenety n sensor traffc. Therefore, motvated by ths, we study the delay-performance of a heterogeneous M2M uplnk channel from the sensors to the Programmable Logc Controller (PLC) n an ndustral automaton scenaro. Besdes the heterogenety across the sensors, the data from each sensor s also heterogeneous and can be classfed as ether Perodc Update (PU) or Event Drven (ED) data. The PU arrvals are perodc, synchronzed by the PLC and need to be processed by a prespecfed frm latency deadlne. On the other hand, the ED arrvals are random and have low-arrval rate wthout any hard servce deadlne. We map these contrastng arrval, servce characterstcs of PU and ED packets from dfferent sensor usng step and sgmodal functons of servce latency respectvely. Due to the heterogenety, the aggregated traffc at PLC forms multple queues/classes whch are then served so as to maxmze a proportonally-far system utlty. Specfcally the proposed mult-class scheduler gves prorty to ED data from dfferent sensors as long as the latency deadlne s met for the PU data. We mnmze successve PU falures usng novel penalty functons and clear faled PU packets to reduce network congeston. Usng extensve smulatons, we compare the performance of our scheme wth varous state-of-the art packet schedulers and show that t outperforms other schedulers when network congeston ncreases. The performance gap further ncreases wth heterogenety n latency requrements and ncrease n penalty for successve PU falures. Index Terms M2M, Latency, Qualty-of-Servce, MultClass Scheduler, Industral Automaton I. INTRODUCTION Automaton for ndustral montorng and control processes has become commonplace these days. It can be modeled as a star-topology Machne-to-Machne (M2M) network of varous process sensors communcatng wrelessly to a Programmable Logc Controller (PLC) on the uplnk channel whch then feeds back the output to a control devce ( downlnk channel ) to ensure the desred process operaton. Typcally, the ndustral M2M traffc has very low latency requrements (of order of few mllseconds). It can be broadly categorzed as ether non real-tme (no deadlne for task completon) or soft realtme (decreased utlty f deadlne not met) or frm (zero utlty f deadlne not met) or hard (system falure f deadlne not Ths research s based upon work supported by the Natonal Scence Foundaton under Grant No met). For nstance, the messages from a nstrumented protectve system or safeguardng system have hard real-tme servce requrements whereas preventve mantenance applcatons are typcally served n non real-tme. Wth ncrease n network sze, the avalable computatonal and communcaton resources gets shared among a large number of sensors. Ths makes t harder to provde real-tme servce due to smultaneous access attempts from multple sensors [] and also wastage of wreless spectrum contnuously allocated to sensors wth very low transmsson duty cycle [2]. Therefore n ths paper, we develop an onlne delay-effcent packet scheduler for heterogeneous M2M uplnk traffc. To begn wth, we use source M2M traffc model n [3] to classfy the sensor data nto Perodc Update (PU) and Event Drven (ED) packets. The PU arrvals are perodc wth frm latency deadlne whle the ED arrvals are random wthout hard servce deadlnes. Paper Contrbutons: We map the contrastng delay requrements of PU and ED packets from dfferent sensors usng step and sgmodal utlty functons. We defne the overall system utlty as the product of (tme) averaged utltes of all PU and ED traffc-classes n order to ensure proportonal farness. Our goal s to determne the optmal schedulng polcy that maxmzes the proposed system utlty. However, due to the randomness n arrval and servce tmes, there exsts no delayoptmal onlne algorthms [4]. Therefore, we propose an onlne delay-effcent mult-class heurstc scheduler for the uplnk traffc at the PLC, that gves prorty to ED data as long as that does not result n a utlty loss for the PU data. Ths results n a hgher utlty for the ED data. However as the sze of networks ncreases, PU packets ncreasngly fal to meet ther deadlnes. We remove the faled PU packets from the system to reduces congeston and thus mprove overall system utlty. Furthermore, snce successve packet drops are worse than solated packet drops, we ntroduce a novel penalty functon as part of PU utlty functon to reduce successve PU falures. Usng extensve smulatons, we show that ths results n an overall ncrease n system utlty as compared to other popular schedulng We assume frm servce deadlne for PU nstead of hard deadlne because typcally the perodc data does not change much from one perod to another, especally f the perod s small. So f a PU packet fals to meet servce deadlne, the PLC can use an estmate for the PU based on the recent hstory, albet wth some degradaton n utlty.
2 polces such as Frst-Come-Frst-Serve (), Earlest Due Date (), prorty schedulng etc. Lastly, the proposed scheduler s agnostc to the wreless technology beng used for the M2M uplnk and s also ndependent of the hardwaresoftware archtecture used n practcal real-tme embedded systems. Also t can easly adapt to accommodate tme-varyng arrval and servce rates for the PU and ED packets. Paper Outlne: The rest of the paper s organzed as follows. Secton II detals the related work. Secton III ntroduces the system model and defnes the utlty functons for PU and ED data. Then n Secton IV, we formulate the utlty maxmzaton problem and n Secton V descrbe the proposed schedulng scheme at PLC. Secton VI presents smulaton results. Fnally Secton VII draws some conclusons. II. RELATED WORK A number of pror works have looked nto ntegratng the desgn of real-tme schedulng schemes n mddle-ware based archtectures so as to meet the hard deadlne of real-tme servces. Tommaso et. al. n [5] presented a real-tme servce orented archtecture for ndustral automaton. However, these works are tangental to our lne of work n the sense that our proposed scheduler s agnostc of the hardware-mddlewaresoftware archtecture and maps the real-tme QoS requrements of M2M traffc nto utlty functons that needs to be maxmzed. Another lne of work focuses on QoS-aware packet scheduler for M2M traffc n Long Term Evoluton (LTE) network (see [6] and references theren). Most of these works use some varants of Access Grant Tme Interval scheme for allocatng fxed or dynamc access grants over perodc tme ntervals to M2M devces. Nusrat et. al. n [7] desgned a packet scheduler for M2M n LTE so as to maxmze the percentage of uplnk packets that satsfy ther ndvdual budget lmts. Unlke our work, these works desgn packet scheduler specfc to a wreless standard such as LTE and are thus heavly nfluenced by the Medum Access Control archtecture of LTE. Also unlke our work, they also don t focus on the heterogenety n packet rate, servce delay requrements of traffc from dfferent sensors. A number of schedulng algorthms have been proposed specfcally for real-tme embedded systems (see [8] and references theren) that are agnostc to the applcaton scenaro, wreless technology used or hardware-software archtecture. These schemes assume hybrd task sets comprsng of hard perodc requests and soft real-tme aperodc requests. The goal of all these schemes s to guarantee completon of servce (before deadlne) for all perodc request and smultaneously am to reduce the average response tmes of aperodc requests. They are broadly classfed nto Fxed-prorty and Dynamc prorty assgnments. Fxed prorty schemes schedule perodc tasks usng Rate-Monotonc algorthm but dffer n servce for aperodc tasks. Dynamc schedulng algorthms schedule perodc tasks usng scheme and allow better processor utlzaton and enhance aperodc responsveness as compared to the fxed prorty schemes. The drawbacks of these schemes Sensor 2 Sensor Sensor 3 Wreless PLC... Class - PU Class 2 - PU Class - ED Class 2 - ED λ pu λ 2 pu λ ed λ 2 ed Sensor N Fg.. System model llustratng queung process n the M2M uplnk. We consder only four dfferent classes for ease of llustraton. s that they assume that the servce tmes for each tasks are known apror. Hence they schemes are not truly onlne n ther present form and would requre consderable modfcatons. The work that s most closely related to our work s by Kumar et. al. n [9] wheren they proposed a delay-effcent heurstc packet scheduler for M2M uplnk. However ths work assumed the traffc orgnatng from dfferent sensors s dentcal havng same arrval rate, servce delay requrements etc. Also, unlke our work, ther proposed scheduler does not dfferentate between successve PU falures and solated PU falures. III. SYSTEM MODEL Fg shows the system model llustratng the queung process n the uplnk channel (.e., from sensors to the PLC) n an ndustral automaton settng. Each of the N sensors transmts two types of packets to the PLC: frequent perodc updates (PU) for some process related measurement or sporadc eventdrven (ED) packets trggered when the sensor measurement for a physcal quantty cross ts threshold. In the most general sense, the arrval rate and latency requrements of uplnk traffc s dfferent for each sensor. Therefore, we model the PU arrvals from the th sensor as a determnstc perodc process wth perod Tpu and thus rate λ pu = /Tpu whle the ED arrvals s modeled as a Posson process wth rate λ ed. Let s pu and s ed denote the sze for PU and ED packets from the th sensor. The servce tme for PU and ED packets from the th sensor are assumed to be exponental wth rate µ pu and µ ed respectvely, gven by µ pu = µ/s pu and µ ed = µ/s ed, () where µ s the servce rate at PLC per unt packet sze. In general, the total tme spent by a packet n the system, T, can
3 ) be wrtten as sum of followng components, T = T trans + T prop + T cong + T queue + T ser, (2) whch denote the followng component delays: T trans : Transmsson delay at the sensor. T prop : Propagaton delay from the sensor to PLC. T cong : Congeston delay due to shared wreless channel n large-scale sensor network. T queue : Queung delay at the PLC. T ser : Processng tme for a packet at the PLC. In ths work, we gnore all the terms except the queung delay at PLC T queue and the servce tme T ser. Ths s because the packet sze s usually qute small to gnore T trans and the sensors and PLC are usually close enough on the shop floor to permt the usage of low power wreless sgnalng and thus gnore T prop. Also, we assume that the avalable wreless spectrum s large enough to allocate a dedcated transmsson channel to each sensor; so there s no congeston delay T cong. In the most general case, the PU and ED packets from N sensors form 2N separate queues/traffc-classes. Now the queung delay for each class depends upon the schedulng polcy adopted at the PLC. The schedulng polcy at PLC should be chosen as to maxmally satsfy the latency constrants of PU and ED packets. IV. PROBLEM FORMULATION We frst map the latency requrements onto the utlty functons for the PU and ED classes as shown n Fg. IV-B and IV-B. A. Utlty Functons ) PU utlty: As stated prevously, the PU packets need to be processed wthn a prespecfed (usually by PLC) tme nterval at the end of whch there s no utlty n servng the packet. For a PU packet from th sensor wth latency lpu, we defne the utlty functon as, { Upu(l pu) f lpu < ld = f lpu ld, (3) where ld s the latency deadlne at whch utlty drops to. We now defne a penalty functon so as to lmt the number of consecutve PU packets that fal to meet the latency deadlne 2. Let r denote the run-length 3 of PU drops for th sensor, then the correspondng penalty s, P pu(r ) = r r γ, r (4) where γ s a penalty functon parameter denotng the degree of penalty for th sensor. γ = mples that we do not penalze consecutve PU drops. 2 Hereafter, we use the term PU drop to ndcate that PU packet fals to meet ts deadlne. 3 A run s a maxmal length sequence of consecutve dentcal outcomes wthn a larger sequence. For example, let [SSDDSDSDDDS] denote a sequence of PU success (S) and drops (D). Then we have three runs of PU drops of lengths, 2 and 3. (l pu Utlty of PU packet from th sensor, U pu Fg. 2. Latency of PU packet from th sensor, (l pu ) n ms l d = ms Utlty functon for PU packet. 2) ED utlty: Unlke, the PU packets, ED packets do not have a hard deadlne, rather ther utlty s a strctly decreasng functon of latency. We defne the ED utlty for th sensor as a sgmodal functon, as n [], [], gven by, ( ) Ued(l ed) = c + e d a(l ed b), (5) where, c = +ea b and d e a b = +e. We note that U a b ed () = and Ued ( ) =. The parameter a s the utlty roll-off factor whose value depends on the crtcalty of the applcaton. The nflecton pont of (5) occurs at led = b. B. System utlty functon For a gven schedulng polcy P, we frst defne a proportonally far system utlty functon as, V(P) = N U puβ pu (P) U edβ ed (P), (6) = where U pu(p) and U ed (P) are the average utlty of PU and ED packets n the steady state gven as, U pu(p) = lm T s U ed(p) = lm T s M pu (Ts) j= U pu(l,j pu (P)) R (T s) Med(T s) j= U ed (l,j M ed (T s ) j= Ppu(r,j (P)) Mpu(T, s ) (7) ed (P)), (8) where Mpu(T s ) and Med (T s) are the number of PU and ED packets from th sensor served n tme T s. R (T s ) denotes the number of runs of PU drops n tme T s and r,j denotes the run-length of j th PU drop for sensor. lpu,j and l,j ed denote the latency of the j th packet from th sensor of type PU and ED respectvely. The parameters βpu and βed denote the prorty gven to average PU and ED utlty for th sensor.
4 ) (l ed Utlty of ED packet from th sensor, U ed. a = ms - b = 2 ms Fg. 3. C. Optmal scheduler Latency of ED packet from th sensor (l ed ) n ms Utlty functon for ED packet. We now descrbe the utlty maxmzaton problem that needs to be solved to determne the optmal schedulng polcy at PLC. It s gven by, max P V(P) = N U puβ pu (P) U edβ ed (P). (9) = If the servce tmes of PU and ED were determnstc or known apror, then the optmal scheme s to schedule PU jobs from sensor, so that they are completed exactly at the deadlne ld. However, snce the arrval tme of ED packet, the servce tmes for PU and ED packets are random, t s not possble to determne an onlne optmal scheduler [4]. Therefore, we propose an onlne heurstc scheduler that ams to maxmze the utlty functon and s descrbed n the next secton. V. PROPOSED SCHEDULER A. Servce order between PU and ED classes The utlty of PU remans constant at untl the deadlne, at whch t drops to. Therefore, t would be best (from the perspectve of ED latency) to delay the servce of PU as much as possble but ensurng that we serve t before deadlne. However, the randomness n PU and ED servce tmes makes t hard to determne the optmal start of servce for each PU packet n real-tme. To overcome ths problem, our proposed scheduler gves prorty to ED packets as long as the current latency of PU packet from any sensor s less than ts preset threshold lt, ( < lt < ld ), =, 2,, N. If the latency of any PU packet exceed ts threshold whle PLC s processng any ED packet, then that ED packet gets mmedately preempted by the PU packet. Ths ensures that we reduce latency of ED packets whle ensurng that most of the PU packets meet ther latency deadlne. B. Servce order wthn PU classes We now dscuss how the proposed scheduler determnes the order of servce between packets of dfferent PU classes. It s more mportant to meet latency deadlne of PU class wth hgh arrval perod snce the sensor data may have changed substantally between two consecutve arrvals. Therefore, we prortze the servce of PU classes n decreasng order of arrval tme-perod,.e., PU class wth hghest arrval perod gets hghest (preemptve) prorty. However, smlar to prevous secton, the prorty of PU classes comes nto effect only when the packet latency exceeds ts correspondng latency threshold. C. Servce order wthn ED classes We now dscuss how the proposed scheduler determnes the order of servce between packets of dfferent ED classes. We assgn hgher prorty to ED classes that are more delay senstve (.e., have hgher a or lower b ). Unlke prevously, the prorty order among ED classes comes nto effect mmedately after the packet enters the system. However, f the latency of th ED class exceeds a certan threshold δ whle there are packet of lower prorty ED classes watng for servce, then ts gets preempted from server by the ED class wth next hghest prorty. By defnton, the threshold δ = for ED class wth least prorty. Ths preempton ensures that ED packets wth unusually large servce tme do not ncrease latency of subsequent packets. Thus, we reduce latency of delay-senstve ED class wthout ncurrng very large latency for delay-tolerant classes. Now, n order to maxmze the system utlty for gven set of system parameters, we need to determne jontly optmal value of PU and ED thresholds, lt and δ = {, 2,, N}. Usng smulatons, the optmal thresholds can be stored n a Look-Up-Table (LUT) pror to the deployment of proposed scheduler. Thus the proposed scheme can easly adapt to changng packet arrval rate, packet sze, number of sensor nodes etc. by lookng up and usng the optmal lt, δ for each case. Another novel feature of our scheduler s that we drop PU packets from servce or queue that exceed ther latency deadlne as there s no resultant utlty from servcng such packets. However, droppng them from servce wll reduce congeston and thus reduce latency for packets of all the classes. The proposed scheduler s descrbed n detal n Algorthm. VI. SIMULATION RESULTS In ths secton, we use Monte-Carlo smulatons to evaluate the system utlty performance of our scheduler aganst varous standard schedulng polces such as,, Preemptve prorty schedulng. The smulaton tme T s s set to a large value (4 s) to ensure a steady-state queung behavor. N s set to 5 sensors. Although the proposed scheduler s desgned for 2N PU and ED classes, the smulatons for proposed scheduler becomes computatonally ntractable as N ncreases. Ths s because t requres us to determne 2N jontly optmal latency thresholds 4. Therefore for computatonal tractablty, we assume that half of sensory traffc (.e., 25 sensors) can 4 The number of threshold varables s 2N rather than 2N because the optmal δ for least prorty ED class n.
5 System utlty, V V = System utlty, V V = 3 Prorty to PU [l d, l d 2 ] = [7.8, 8] [a, b, s ] = [5, 9, ] ] = [, 2, ] T pu = T 2 pu = 88 ms γ =. Prorty to PU [l d, l d 2 ] = [4, 8] [a, b, s ] = [,, ] ] = [, 2, ] [T pu ] = [5, 62.5] ms γ =. 5 ED arrval rate, λ ed. 5 ED arrval rate, λ ed Fg. 4. System utlty wth roughly smlar latency requrements of sensors. Fg. 5. System utlty wth dverse latency requrements of sensors. be classfed nto one set of PU and ED class (.e., PU and ED), whle the other half nto PU2 and ED2 class. So, we consder a total of 4 traffc classes. We assume the arrval perod per sensor for PU to be Tpu = 5 ms, whle for PU2 to be Tpu 2 = 62.5 ms. The arrval rate per sensor for ED2 s set to λ 2 ed =.4/ms and for ED t s vared from.4./ms. PLC servce rate s µ = bts/ms. The relatve mportance of all classes s set equal,.e. βpu = βpu 2 = βed = βed 2 =, unless mentoned otherwse. The sze of packets for all classes are set equal to. The latency deadlne for PU s, ld = 4 ms and for PU2 s l2 d = 8 ms. The utlty functon parameters for ED are set to, [a, b, s ] = [,, ] and for ED2 are ] = [, 2, ] unless mentoned otherwse. The latency threshold for ED s, δ = b + (4/a ) and for ED2 s δ 2 =. For each smulaton scenaro, the latency threshold for PU and PU2 class are determned by jontly maxmzng system utlty over lt ld and l2 t ld 2. A. Impact of latency-heterogenety across sensors In ths secton, we study the mpact of ncreasng heterogenety n latency requrements of varous sensors on the performance of dfferent schedulers. Fg. 4 shows plot of system utlty for a nearly homogeneous system wth Tpu = Tpu 2 = 88 ms, [ld, l2 d ] = [7.8, 8] ms, [a, a 2 ] = [5, 7]/ms and [b ] = [9, 2] ms. Then, we ncrease the heterogenety n system by usng the default latency parameters; the resultng system utlty s shown n Fg. 5. We note that the performance gap between the proposed scheduler and the other schemes, V, s much larger for heterogeneous system ( V = 3) compared to homogeneous system ( V = ), especally at hgher λ ed. Ths s because the mpact of prorty schedulng for dfferent PU and ED classes n proposed scheduler becomes more promnent when the traffc classes are dverse. B. Impact of choce of ED threshold, δ Fg. 6 shows the mpact of selectng optmal threshold δ for ED class on the system utlty. We note that the performance of proposed scheduler s sgnfcantly mproved ( V = ) compared to Fg. 5 ( V = 3) where δ was arbtrarly set to b + (4/a ), especally at hgher λ ed. Ths s because selectng optmal δ avods latency of ED2 to becomes very large when λ ed becomes very large relatve to λ ed2. System utlty, V. Prorty to PU [l d, l d 2 ] = [4, 8] [a, b, s ] = [,, ] ] = [, 2, ] [T pu ] = [5, 62.5] ms γ =. 5 ED arrval rate, λ ed V = Fg. 6. System utlty wth optmal selecton of threshold δ. C. Impact of penalzng consecutve PU drops So far, we have assumed there s no penalty for consecutve PU drops,.e., γ =. We now study the mpact of penalzng consecutve PU drops (γ > ) on the performance of dfferent schedulers. Fg. 7 shows the plot of system utlty when the PU drop penalty s small and equal for PU and PU2 (γ = ). We observe that whle the performance of proposed schedulers does not change apprecably, the performance of other schedulers drops down consderably and becomes close to at λ ed = 5. Agan ths s because the proposed scheduler easly adapts to the change n PU penalty by alterng the latency thresholds accordngly. But the other schedulers cannot do so and ther performance suffer. Ths degradaton n performance s more pronounced when the PU drop penalty for hgher prorty class PU s ncreased from γ =.2 to 2 as llustrated n Fg. 8.
6 System utlty, V. Prorty to PU [l d, l d 2 ] = [4, 8] [a, b, s ] = [,, ] ] = [, 2, ] [T pu ] = [5, 62.5] ms γ =.2. 5 ED arrval rate, λ ed Fg. 7. System utlty wth small, equal PU drop penalty (γ =.2). System utlty, V. Prorty to PU [l d, l d 2 ] = [4, 8] [a, b, s ] = [,, ] ] = [, 2, ] [T pu ] = [5, 62.5] ms [γ, γ 2 ] = [2,.2]. 2 4 ED arrval rate, λ ed REFERENCES [] 3GPP, Study on RAN Improvements for Machne-type communcatons, tech. rep., Techncal report, TR , 22. [2] D. Drajc, S. Krco, I. Tomc, M. Popovc, N. Zeljkovc, N. Nkaen, and P. Svoboda, Impact of onlne games and m2m applcatons traffc on performance of hspa rado access networks, n Internatonal Conference on Innovatve Moble and Internet Servces n Ubqutous Computng, pp , July 22. [3] N. Nkaen, M. Laner, K. Zhou, P. Svoboda, D. Drajc, M. Popovc, and S. Krco, Smple traffc modelng framework for machne type communcaton, n Internatonal Symposum on Wreless Communcaton Systems, pp. 5, Aug 23. [4] T. Ta, J. Lu, and M. Shankar, Algorthms and optmalty of schedulng soft aperodc requests n fxed-prorty preemptve systems, Real-Tme Systems, vol., pp , Jan 996. [5] T. Cucnotta, A. Mancna, G. Anastas, G. Lpar, L. Mangeruca, R. Checcozzo, and F. Rusna, A real-tme servce-orented archtecture for ndustral automaton, IEEE Transactons on Industral Informatcs, vol. 5, pp , Aug 29. [6] A. Gotss, A. Loumpas, and A. Alexou, Evoluton of packet schedulng for machne-type communcatons over lte: Algorthmc desgn and performance analyss, n IEEE Globecom Workshop, pp , Dec 22. [7] N. Afrn, J. Brown, and J. Khan, A delay senstve lte uplnk packet scheduler for m2m traffc, n IEEE Globecom Workshop, pp , Dec 23. [8] G. Buttazzo, Hard Real-Tme Computng Systems: Predctable Schedulng Algorthms and Applcatons. New York: Sprnger, Thrd ed., 2. [9] A. Kumar, A. Abdelhad, and T. Clancy, An Onlne Delay Effcent Packet Scheduler for M2M Traffc n Industral Automaton, n Submtted to IEEE Internatonal Systems Conference (SysCon), 26. [] A. Abdelhad and T. Clancy, A Utlty Proportonal Farness Approach for Resource Allocaton n 4G-LTE, n IEEE Internatonal Conference on Computng, Networkng, and Communcatons (ICNC), CNC Workshop, 24. [] A. Abdelhad and T. Clancy, A Robust Optmal Rate Allocaton Algorthm and Prcng Polcy for Hybrd Traffc n 4G-LTE, n IEEE Internatonal Symposum on Personal, Indoor, and Moble Rado Communcatons (PIMRC), 23. Fg. 8. System utlty wth unequal PU drop penalty ([γ, γ 2 ] = [2,.2]). VII. CONCLUSIONS In ths paper, we presented a onlne delay-effcent packet scheduler for uplnk M2M traffc wth heterogenety n latency requrements both wthn and across sensors. The data from each sensor s classfed as ether PU or ED packets. The varyng latency requrements of PU and ED class for dfferent sensors s represented usng dfferent utlty functons. We then proposed a novel proportonally-far mult-class packet scheduler that prmarly gves prorty to ED data as long as the latency deadlne s met for the PU data. However as the sze of networks ncreases, PU packets ncreasngly fal to meet ther deadlnes. We remove the faled PU packets from the system to reduces congeston and thus mprove overall system utlty. We also ntroduce a novel penalty functon as part of PU utlty functon so as to reduce burst PU falures. Usng extensve smulatons, we dd a comparatve analyss of the proposed scheme wth varous state-of-the art schedulng polces such as,, Preemptve prorty etc. We note that as the uplnk traffc ncreases, the proposed scheduler outperforms other schedulers. The performance gap further ncreases wth ncreasng heterogenety n latency requrements and ncrease n penalty for successve PU drops.
7 Algorthm Proposed Onlne Packet Scheduler Inputs ld, l t = {, 2,, N} are latency deadlnes and thresholds for N PU classes. δ = {, 2,, N} are the latency threshold for N ED classes. T sm s the duraton of smulaton and T s the tme resoluton at whch the algorthm operates. Local T c stores current system tme. A and B denotes the set of all PU and ED classes. A e and B e denotes the set of all PU and ED classes that exceed ther correspondng threshold lt and δ respectvely. B denotes the set B B e. A d denotes the set of all PU classes that fal to meet ther deadlnes ld. A h, A e h, B h, B h denote the class wth hghest prorty n sets A, A e, B, B respectvely. z denote class of current packet beng served. Intalzaton Advance current tme T c to be the arrval tme of next packet n system. Set A d =. Set A h, B h from gven set of PU and ED classes. Begn Algorthm: : whle T c T sm do 2: Determne z and the set A d. 3: f A d then 4: Remove the Head-of-Lne (HoL) packet from all PU queues n set A d. go to 5. 5: end f 6: Determne the set A e, B e, B and the classes A e h, B h. 7: 8: f A e & (z / A e (z A e & z A e h )) then Preempt the current packet wth HoL of class A,h. 9: else f A e = & B e = & z B h then : Preempt the current packet wth HoL packet of class B h. : else f A e = & B e & z B h then 2: Preempt the current packet wth HoL of class B h. 3: else No preempton occurs; wat tll current packet leaves server. 4: Advance current tme T c to the servce completon event. 5: Recalculate the sets A e, B e, B and the classes A e h, B h. 6: f A e = & B e = then None of the PU or ED class exceed latency threshold 7: Serve the HoL packet from class B h. 8: else f A e = & B e then None of the PU class exceed latency threshold 9: Serve the HoL packet from class B h. 2: else Some of the PU class exceed latency threshold 2: Serve the HoL packet from class A e h. 22: end f 23: go to. 24: end f 25: T c = T c + T. 26: end whle
Real-Time Guarantees. Traffic Characteristics. Flow Control
Real-Tme Guarantees Requrements on RT communcaton protocols: delay (response s) small jtter small throughput hgh error detecton at recever (and sender) small error detecton latency no thrashng under peak
More informationAADL : about scheduling analysis
AADL : about schedulng analyss Schedulng analyss, what s t? Embedded real-tme crtcal systems have temporal constrants to meet (e.g. deadlne). Many systems are bult wth operatng systems provdng multtaskng
More informationWishing you all a Total Quality New Year!
Total Qualty Management and Sx Sgma Post Graduate Program 214-15 Sesson 4 Vnay Kumar Kalakband Assstant Professor Operatons & Systems Area 1 Wshng you all a Total Qualty New Year! Hope you acheve Sx sgma
More informationReal-time Scheduling
Real-tme Schedulng COE718: Embedded System Desgn http://www.ee.ryerson.ca/~courses/coe718/ Dr. Gul N. Khan http://www.ee.ryerson.ca/~gnkhan Electrcal and Computer Engneerng Ryerson Unversty Overvew RTX
More informationRouting in Degree-constrained FSO Mesh Networks
Internatonal Journal of Hybrd Informaton Technology Vol., No., Aprl, 009 Routng n Degree-constraned FSO Mesh Networks Zpng Hu, Pramode Verma, and James Sluss Jr. School of Electrcal & Computer Engneerng
More informationChannel-Quality Dependent Earliest Deadline Due Fair Scheduling Schemes for Wireless Multimedia Networks
Channel-Qualty Dependent Earlest Deadlne Due Far Schedulng Schemes for Wreless Multmeda Networks Ahmed K. F. Khattab Khaled M. F. Elsayed ahmedkhattab@eng.cu.edu.eg khaled@eee.org Department of Electroncs
More informationCompiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz
Compler Desgn Sprng 2014 Regster Allocaton Sample Exercses and Solutons Prof. Pedro C. Dnz USC / Informaton Scences Insttute 4676 Admralty Way, Sute 1001 Marna del Rey, Calforna 90292 pedro@s.edu Regster
More informationFibre-Optic AWG-based Real-Time Networks
Fbre-Optc AWG-based Real-Tme Networks Krstna Kunert, Annette Böhm, Magnus Jonsson, School of Informaton Scence, Computer and Electrcal Engneerng, Halmstad Unversty {Magnus.Jonsson, Krstna.Kunert}@de.hh.se
More informationEnhanced Signaling Scheme with Admission Control in the Hybrid Optical Wireless (HOW) Networks
Enhanced Sgnalng Scheme wth Admsson Control n the Hybrd Optcal Wreless (HOW) Networks Yng Yan, Hao Yu, Henrk Wessng, and Lars Dttmann Department of Photoncs Techncal Unversty of Denmark Lyngby, Denmark
More informationQoS-aware composite scheduling using fuzzy proactive and reactive controllers
Khan et al. EURASIP Journal on Wreless Communcatons and Networkng 2014, 2014:138 http://jwcn.euraspjournals.com/content/2014/1/138 RESEARCH Open Access QoS-aware composte schedulng usng fuzzy proactve
More informationParallelism for Nested Loops with Non-uniform and Flow Dependences
Parallelsm for Nested Loops wth Non-unform and Flow Dependences Sam-Jn Jeong Dept. of Informaton & Communcaton Engneerng, Cheonan Unversty, 5, Anseo-dong, Cheonan, Chungnam, 330-80, Korea. seong@cheonan.ac.kr
More informationPriority-Based Scheduling Algorithm for Downlink Traffics in IEEE Networks
Prorty-Based Schedulng Algorthm for Downlnk Traffcs n IEEE 80.6 Networks Ja-Mng Lang, Jen-Jee Chen, You-Chun Wang, Yu-Chee Tseng, and Bao-Shuh P. Ln Department of Computer Scence Natonal Chao-Tung Unversty,
More informationLecture 7 Real Time Task Scheduling. Forrest Brewer
Lecture 7 Real Tme Task Schedulng Forrest Brewer Real Tme ANSI defnes real tme as A Real tme process s a process whch delvers the results of processng n a gven tme span A data may requre processng at a
More informationNetwork Coding as a Dynamical System
Network Codng as a Dynamcal System Narayan B. Mandayam IEEE Dstngushed Lecture (jont work wth Dan Zhang and a Su) Department of Electrcal and Computer Engneerng Rutgers Unversty Outlne. Introducton 2.
More informationSimulation Based Analysis of FAST TCP using OMNET++
Smulaton Based Analyss of FAST TCP usng OMNET++ Umar ul Hassan 04030038@lums.edu.pk Md Term Report CS678 Topcs n Internet Research Sprng, 2006 Introducton Internet traffc s doublng roughly every 3 months
More informationMultitasking and Real-time Scheduling
Multtaskng and Real-tme Schedulng EE8205: Embedded Computer Systems http://www.ee.ryerson.ca/~courses/ee8205/ Dr. Gul N. Khan http://www.ee.ryerson.ca/~gnkhan Electrcal and Computer Engneerng Ryerson Unversty
More informationVirtual Memory. Background. No. 10. Virtual Memory: concept. Logical Memory Space (review) Demand Paging(1) Virtual Memory
Background EECS. Operatng System Fundamentals No. Vrtual Memory Prof. Hu Jang Department of Electrcal Engneerng and Computer Scence, York Unversty Memory-management methods normally requres the entre process
More informationAnalysis of Collaborative Distributed Admission Control in x Networks
1 Analyss of Collaboratve Dstrbuted Admsson Control n 82.11x Networks Thnh Nguyen, Member, IEEE, Ken Nguyen, Member, IEEE, Lnha He, Member, IEEE, Abstract Wth the recent surge of wreless home networks,
More informationRAP. Speed/RAP/CODA. Real-time Systems. Modeling the sensor networks. Real-time Systems. Modeling the sensor networks. Real-time systems:
Speed/RAP/CODA Presented by Octav Chpara Real-tme Systems Many wreless sensor network applcatons requre real-tme support Survellance and trackng Border patrol Fre fghtng Real-tme systems: Hard real-tme:
More informationVideo Proxy System for a Large-scale VOD System (DINA)
Vdeo Proxy System for a Large-scale VOD System (DINA) KWUN-CHUNG CHAN #, KWOK-WAI CHEUNG *# #Department of Informaton Engneerng *Centre of Innovaton and Technology The Chnese Unversty of Hong Kong SHATIN,
More informationEfficient Distributed File System (EDFS)
Effcent Dstrbuted Fle System (EDFS) (Sem-Centralzed) Debessay(Debsh) Fesehaye, Rahul Malk & Klara Naherstedt Unversty of Illnos-Urbana Champagn Contents Problem Statement, Related Work, EDFS Desgn Rate
More informationAdaptive Resource Allocation Control with On-Line Search for Fair QoS Level
Adaptve Resource Allocaton Control wth On-Lne Search for Far QoS Level Fumko Harada, Toshmtsu Usho, Graduate School of Engneerng Scence Osaka Unversty {harada@hopf, usho@}sysesosaka-uacjp Yukkazu akamoto
More informationDistributed Resource Scheduling in Grid Computing Using Fuzzy Approach
Dstrbuted Resource Schedulng n Grd Computng Usng Fuzzy Approach Shahram Amn, Mohammad Ahmad Computer Engneerng Department Islamc Azad Unversty branch Mahallat, Iran Islamc Azad Unversty branch khomen,
More informationScheduling and queue management. DigiComm II
Schedulng and queue management Tradtonal queung behavour n routers Data transfer: datagrams: ndvdual packets no recognton of flows connectonless: no sgnallng Forwardng: based on per-datagram forwardng
More informationDESIGNING TRANSMISSION SCHEDULES FOR WIRELESS AD HOC NETWORKS TO MAXIMIZE NETWORK THROUGHPUT
DESIGNING TRANSMISSION SCHEDULES FOR WIRELESS AD HOC NETWORKS TO MAXIMIZE NETWORK THROUGHPUT Bran J. Wolf, Joseph L. Hammond, and Harlan B. Russell Dept. of Electrcal and Computer Engneerng, Clemson Unversty,
More informationComparison of Heuristics for Scheduling Independent Tasks on Heterogeneous Distributed Environments
Comparson of Heurstcs for Schedulng Independent Tasks on Heterogeneous Dstrbuted Envronments Hesam Izakan¹, Ath Abraham², Senor Member, IEEE, Václav Snášel³ ¹ Islamc Azad Unversty, Ramsar Branch, Ramsar,
More informationSolutions for Real-Time Communication over Best-Effort Networks
Solutons for Real-Tme Communcaton over Best-Effort Networks Anca Hangan, Ramona Marfevc, Gheorghe Sebestyen Techncal Unversty of Cluj-Napoca, Computer Scence Department {Anca.Hangan, Ramona.Marfevc, Gheorghe.Sebestyen}@cs.utcluj.ro
More informationA Frame Packing Mechanism Using PDO Communication Service within CANopen
28 A Frame Packng Mechansm Usng PDO Communcaton Servce wthn CANopen Mnkoo Kang and Kejn Park Dvson of Industral & Informaton Systems Engneerng, Ajou Unversty, Suwon, Gyeongg-do, South Korea Summary The
More informationA mathematical programming approach to the analysis, design and scheduling of offshore oilfields
17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 A mathematcal programmng approach to the analyss, desgn and
More informationS1 Note. Basis functions.
S1 Note. Bass functons. Contents Types of bass functons...1 The Fourer bass...2 B-splne bass...3 Power and type I error rates wth dfferent numbers of bass functons...4 Table S1. Smulaton results of type
More informationWITH rapid improvements of wireless technologies,
JOURNAL OF SYSTEMS ARCHITECTURE, SPECIAL ISSUE: HIGHLY-RELIABLE CPS, VOL. 00, NO. 0, MONTH 013 1 Adaptve GTS Allocaton n IEEE 80.15.4 for Real-Tme Wreless Sensor Networks Feng Xa, Ruonan Hao, Je L, Naxue
More informationFast Retransmission of Real-Time Traffic in HIPERLAN/2 Systems
Fast Retransmsson of Real-Tme Traffc n HIPERLAN/ Systems José A Afonso and Joaqum E Neves Department of Industral Electroncs Unversty of Mnho, Campus de Azurém 4800-058 Gumarães, Portugal {joseafonso,
More informationDynamic Bandwidth Provisioning with Fairness and Revenue Considerations for Broadband Wireless Communication
Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the ICC 008 proceedngs. Dynamc Bandwdth Provsonng wth Farness and Revenue Consderatons
More informationX- Chart Using ANOM Approach
ISSN 1684-8403 Journal of Statstcs Volume 17, 010, pp. 3-3 Abstract X- Chart Usng ANOM Approach Gullapall Chakravarth 1 and Chaluvad Venkateswara Rao Control lmts for ndvdual measurements (X) chart are
More informationUse of Genetic Algorithms in Efficient Scheduling for Multi Service Classes
Use of Genetc Algorthms n Effcent Schedulng for Mult Servce Classes Shyamale Thlakawardana and Rahm Tafazoll Centre for Communcatons Systems Research (CCSR), Unversty of Surrey, Guldford, GU27XH, UK Abstract
More informationReducing Frame Rate for Object Tracking
Reducng Frame Rate for Object Trackng Pavel Korshunov 1 and We Tsang Oo 2 1 Natonal Unversty of Sngapore, Sngapore 11977, pavelkor@comp.nus.edu.sg 2 Natonal Unversty of Sngapore, Sngapore 11977, oowt@comp.nus.edu.sg
More informationTHere are increasing interests and use of mobile ad hoc
1 Adaptve Schedulng n MIMO-based Heterogeneous Ad hoc Networks Shan Chu, Xn Wang Member, IEEE, and Yuanyuan Yang Fellow, IEEE. Abstract The demands for data rate and transmsson relablty constantly ncrease
More informationARTICLE IN PRESS. Signal Processing: Image Communication
Sgnal Processng: Image Communcaton 23 (2008) 754 768 Contents lsts avalable at ScenceDrect Sgnal Processng: Image Communcaton journal homepage: www.elsever.com/locate/mage Dstrbuted meda rate allocaton
More informationReal-time Fault-tolerant Scheduling Algorithm for Distributed Computing Systems
Real-tme Fault-tolerant Schedulng Algorthm for Dstrbuted Computng Systems Yun Lng, Y Ouyang College of Computer Scence and Informaton Engneerng Zheang Gongshang Unversty Postal code: 310018 P.R.CHINA {ylng,
More informationPricing Network Resources for Adaptive Applications in a Differentiated Services Network
IEEE INFOCOM Prcng Network Resources for Adaptve Applcatons n a Dfferentated Servces Network Xn Wang and Hennng Schulzrnne Columba Unversty Emal: {xnwang, schulzrnne}@cs.columba.edu Abstract The Dfferentated
More information6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour
6.854 Advanced Algorthms Petar Maymounkov Problem Set 11 (November 23, 2005) Wth: Benjamn Rossman, Oren Wemann, and Pouya Kheradpour Problem 1. We reduce vertex cover to MAX-SAT wth weghts, such that the
More informationMaintaining temporal validity of real-time data on non-continuously executing resources
Mantanng temporal valdty of real-tme data on non-contnuously executng resources Tan Ba, Hong Lu and Juan Yang Hunan Insttute of Scence and Technology, College of Computer Scence, 44, Yueyang, Chna Wuhan
More informationCHAPTER 2 PROPOSED IMPROVED PARTICLE SWARM OPTIMIZATION
24 CHAPTER 2 PROPOSED IMPROVED PARTICLE SWARM OPTIMIZATION The present chapter proposes an IPSO approach for multprocessor task schedulng problem wth two classfcatons, namely, statc ndependent tasks and
More informationHIERARCHICAL SCHEDULING WITH ADAPTIVE WEIGHTS FOR W-ATM *
Copyrght Notce c 1999 IEEE. Personal use of ths materal s permtted. However, permsson to reprnt/republsh ths materal for advertsng or promotonal purposes or for creatng new collectve wors for resale or
More informationTolerating Transient Faults in Statically Scheduled Safety-Critical Embedded Systems
Toleratng Transent Faults n Statcally Scheduled Safety-Crtcal Embedded Systems Nagaraan Kandasamy *, John P. Hayes *, and Bran T. Murray ** * Department of Electrcal Engneerng ** Delph Automotve Systems
More informationSubspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;
Subspace clusterng Clusterng Fundamental to all clusterng technques s the choce of dstance measure between data ponts; D q ( ) ( ) 2 x x = x x, j k = 1 k jk Squared Eucldean dstance Assumpton: All features
More informationQuantifying Responsiveness of TCP Aggregates by Using Direct Sequence Spread Spectrum CDMA and Its Application in Congestion Control
Quantfyng Responsveness of TCP Aggregates by Usng Drect Sequence Spread Spectrum CDMA and Its Applcaton n Congeston Control Mehd Kalantar Department of Electrcal and Computer Engneerng Unversty of Maryland,
More informationAn Optimal Algorithm for Prufer Codes *
J. Software Engneerng & Applcatons, 2009, 2: 111-115 do:10.4236/jsea.2009.22016 Publshed Onlne July 2009 (www.scrp.org/journal/jsea) An Optmal Algorthm for Prufer Codes * Xaodong Wang 1, 2, Le Wang 3,
More informationEvaluation of Parallel Processing Systems through Queuing Model
ISSN 2278-309 Vkas Shnde, Internatonal Journal of Advanced Volume Trends 4, n Computer No.2, March Scence - and Aprl Engneerng, 205 4(2), March - Aprl 205, 36-43 Internatonal Journal of Advanced Trends
More informationEfficient Broadcast Disks Program Construction in Asymmetric Communication Environments
Effcent Broadcast Dsks Program Constructon n Asymmetrc Communcaton Envronments Eleftheros Takas, Stefanos Ougaroglou, Petros copoltds Department of Informatcs, Arstotle Unversty of Thessalonk Box 888,
More informationFAHP and Modified GRA Based Network Selection in Heterogeneous Wireless Networks
2017 2nd Internatonal Semnar on Appled Physcs, Optoelectroncs and Photoncs (APOP 2017) ISBN: 978-1-60595-522-3 FAHP and Modfed GRA Based Network Selecton n Heterogeneous Wreless Networks Xaohan DU, Zhqng
More informationA New Token Allocation Algorithm for TCP Traffic in Diffserv Network
A New Token Allocaton Algorthm for TCP Traffc n Dffserv Network A New Token Allocaton Algorthm for TCP Traffc n Dffserv Network S. Sudha and N. Ammasagounden Natonal Insttute of Technology, Truchrappall,
More informationDelay Variation Optimized Traffic Allocation Based on Network Calculus for Multi-path Routing in Wireless Mesh Networks
Appl. Math. Inf. Sc. 7, No. 2L, 467-474 2013) 467 Appled Mathematcs & Informaton Scences An Internatonal Journal http://dx.do.org/10.12785/ams/072l13 Delay Varaton Optmzed Traffc Allocaton Based on Network
More informationNew Method Based on Priority of Heterogeneous Traffic for Scheduling Techniques in M2M Communications over LTE Networks
Receved: July 28, 2018 209 New Method Based on Prorty of Heterogeneous Traffc for Schedulng Technques n M2M Communcatons over LTE Networks Maryam Ouassa 1 * Abdallah Rhattoy 2 1 Informaton and Communcaton
More informationState of the Art in Differentiated
Outlne Dfferentated Servces on the Internet Explct Allocaton of Best Effort Packet Delvery Servce, D. Clark and W. Fang A Two bt Dfferentated Servces Archtecture for the Internet, K. Nchols, V. Jacobson,
More informationOn the Fairness-Efficiency Tradeoff for Packet Processing with Multiple Resources
On the Farness-Effcency Tradeoff for Packet Processng wth Multple Resources We Wang, Chen Feng, Baochun L, and Ben Lang Department of Electrcal and Computer Engneerng, Unversty of Toronto {wewang, cfeng,
More informationThe Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique
//00 :0 AM Outlne and Readng The Greedy Method The Greedy Method Technque (secton.) Fractonal Knapsack Problem (secton..) Task Schedulng (secton..) Mnmum Spannng Trees (secton.) Change Money Problem Greedy
More informationCluster Analysis of Electrical Behavior
Journal of Computer and Communcatons, 205, 3, 88-93 Publshed Onlne May 205 n ScRes. http://www.scrp.org/ournal/cc http://dx.do.org/0.4236/cc.205.350 Cluster Analyss of Electrcal Behavor Ln Lu Ln Lu, School
More informationA Genetic Algorithm Based Dynamic Load Balancing Scheme for Heterogeneous Distributed Systems
Proceedngs of the Internatonal Conference on Parallel and Dstrbuted Processng Technques and Applcatons, PDPTA 2008, Las Vegas, Nevada, USA, July 14-17, 2008, 2 Volumes. CSREA Press 2008, ISBN 1-60132-084-1
More informationA Semi-Distributed Load Balancing Architecture and Algorithm for Heterogeneous Wireless Networks
A Sem-Dstrbuted oad Balancng Archtecture and Algorthm for Heterogeneous reless Networks Md. Golam Rabul Ala Choong Seon Hong * Kyung Hee Unversty, Korea rob@networkng.khu.ac.kr, cshong@khu.ac.kr Abstract
More informationEfficient QoS Provisioning at the MAC Layer in Heterogeneous Wireless Sensor Networks
Effcent QoS Provsonng at the MAC Layer n Heterogeneous Wreless Sensor Networks M.Soul a,, A.Bouabdallah a, A.E.Kamal b a UMR CNRS 7253 HeuDaSyC, Unversté de Technologe de Compègne, Compègne Cedex F-625,
More informationRelated-Mode Attacks on CTR Encryption Mode
Internatonal Journal of Network Securty, Vol.4, No.3, PP.282 287, May 2007 282 Related-Mode Attacks on CTR Encrypton Mode Dayn Wang, Dongda Ln, and Wenlng Wu (Correspondng author: Dayn Wang) Key Laboratory
More informationCS 268: Lecture 8 Router Support for Congestion Control
CS 268: Lecture 8 Router Support for Congeston Control Ion Stoca Computer Scence Dvson Department of Electrcal Engneerng and Computer Scences Unversty of Calforna, Berkeley Berkeley, CA 9472-1776 Router
More informationAn Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation
17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 An Iteratve Soluton Approach to Process Plant Layout usng Mxed
More informationTechnical Report. i-game: An Implicit GTS Allocation Mechanism in IEEE for Time- Sensitive Wireless Sensor Networks
www.hurray.sep.pp.pt Techncal Report -GAME: An Implct GTS Allocaton Mechansm n IEEE 802.15.4 for Tme- Senstve Wreless Sensor etworks Ans Koubaa Máro Alves Eduardo Tovar TR-060706 Verson: 1.0 Date: Jul
More informationUSER CLASS BASED QoS DIFFERENTIATION IN e WLAN
USER CLASS BASED QoS DIFFERENTIATION IN 802.11e WLAN Amt Kejrwal, Ratan Guha School of Computer Scence Unversty of Central Florda Orlando, FL-32816 USA E-mal: {kejrwal, guha}@cs.ucf.edu Manak Chatterjee
More informationA Statistical Priority-based Scheduling Metric for M2M Communications in LTE Networks
1 A Statstcal Prorty-based Schedulng Metrc for M2M Communcatons n LTE Networks Ahmed Elhamy Mostafa, Student Member, IEEE and Yasser Gadallah, Senor Member, IEEE Abstract Resource allocaton, or schedulng,
More informationDelay Analysis and Time-Critical Protocol Design for In-Vehicle Power Line Communication Systems
Delay Analyss and Tme-Crtcal Protocol Desgn for In-Vehcle Power Lne Communcaton Systems Zhengguo Sheng, Daxn Tan, Vctor C.M. Leung and Gaurav Bansal Abstract Wth the emergng automated tasks n vehcle doman,
More informationFORESIGHTED JOINT RESOURCE RECIPROCATION AND SCHEDULING STRATEGIES FOR REAL-TIME VIDEO STREAMING OVER PEER-TO-PEER NETWORKS
FORESIGHTED JOINT RESOURCE RECIPROCATION AND SCHEDULING STRATEGIES FOR REAL-TIME VIDEO STREAMING OVER PEER-TO-PEER NETWORKS Sunghoon Ivan Lee, Hyunggon Park, and Mhaela van der Schaar Electrcal Engneerng
More informationMixed-Criticality Scheduling on Multiprocessors using Task Grouping
Mxed-Crtcalty Schedulng on Multprocessors usng Task Groupng Jankang Ren Lnh Th Xuan Phan School of Software Technology, Dalan Unversty of Technology, Chna Computer and Informaton Scence Department, Unversty
More informationSome Advanced SPC Tools 1. Cumulative Sum Control (Cusum) Chart For the data shown in Table 9-1, the x chart can be generated.
Some Advanced SP Tools 1. umulatve Sum ontrol (usum) hart For the data shown n Table 9-1, the x chart can be generated. However, the shft taken place at sample #21 s not apparent. 92 For ths set samples,
More informationOn Achieving Fairness in the Joint Allocation of Buffer and Bandwidth Resources: Principles and Algorithms
On Achevng Farness n the Jont Allocaton of Buffer and Bandwdth Resources: Prncples and Algorthms Yunka Zhou and Harsh Sethu (correspondng author) Abstract Farness n network traffc management can mprove
More informationVirtual Machine Migration based on Trust Measurement of Computer Node
Appled Mechancs and Materals Onlne: 2014-04-04 ISSN: 1662-7482, Vols. 536-537, pp 678-682 do:10.4028/www.scentfc.net/amm.536-537.678 2014 Trans Tech Publcatons, Swtzerland Vrtual Machne Mgraton based on
More informationTOWARDS OPTIMAL RESOURCE ALLOCATION FOR DIFFERENTIATED MULTIMEDIA SERVICES IN CLOUD COMPUTING ENVIRONMENT. Xiaoming Nan, Yifeng He, and Ling Guan
TOWARDS OTIAL RESOURCE ALLOCATIO OR DIERETIATED ULTIEDIA SERVICES I CLOUD COUTIG EVIROET Xaomng an, Yfeng He, and Lng Guan Ryerson Unversty, Toronto, Canada ABSTRACT Cloud-based multmeda servces have been
More informationProblem Definitions and Evaluation Criteria for Computational Expensive Optimization
Problem efntons and Evaluaton Crtera for Computatonal Expensve Optmzaton B. Lu 1, Q. Chen and Q. Zhang 3, J. J. Lang 4, P. N. Suganthan, B. Y. Qu 6 1 epartment of Computng, Glyndwr Unversty, UK Faclty
More informationTN348: Openlab Module - Colocalization
TN348: Openlab Module - Colocalzaton Topc The Colocalzaton module provdes the faclty to vsualze and quantfy colocalzaton between pars of mages. The Colocalzaton wndow contans a prevew of the two mages
More informationOn the Exact Analysis of Bluetooth Scheduling Algorithms
On the Exact Analyss of Bluetooth Schedulng Algorth Gl Zussman Dept. of Electrcal Engneerng Technon IIT Hafa 3000, Israel glz@tx.technon.ac.l Ur Yechal Dept. of Statstcs and Operatons Research School of
More informationLoad-Balanced Anycast Routing
Load-Balanced Anycast Routng Chng-Yu Ln, Jung-Hua Lo, and Sy-Yen Kuo Department of Electrcal Engneerng atonal Tawan Unversty, Tape, Tawan sykuo@cc.ee.ntu.edu.tw Abstract For fault-tolerance and load-balance
More informationHermite Splines in Lie Groups as Products of Geodesics
Hermte Splnes n Le Groups as Products of Geodescs Ethan Eade Updated May 28, 2017 1 Introducton 1.1 Goal Ths document defnes a curve n the Le group G parametrzed by tme and by structural parameters n the
More informationScheduling Remote Access to Scientific Instruments in Cyberinfrastructure for Education and Research
Schedulng Remote Access to Scentfc Instruments n Cybernfrastructure for Educaton and Research Je Yn 1, Junwe Cao 2,3,*, Yuexuan Wang 4, Lanchen Lu 1,3 and Cheng Wu 1,3 1 Natonal CIMS Engneerng and Research
More informationResource Control for Loss-Sensitive Traffic in CDMA Networks
Resource Control for Loss-Senstve Traffc n CDMA Networks Vaslos A. Srs Insttute of Computer Scence, FORTH and Department of Computer Scence, Unversty of Crete P. O. Box 1385, GR 711 1, Heraklon, Crete,
More informationImproving Low Density Parity Check Codes Over the Erasure Channel. The Nelder Mead Downhill Simplex Method. Scott Stransky
Improvng Low Densty Party Check Codes Over the Erasure Channel The Nelder Mead Downhll Smplex Method Scott Stransky Programmng n conjuncton wth: Bors Cukalovc 18.413 Fnal Project Sprng 2004 Page 1 Abstract
More informationPerformance Comparison of a QoS Aware Routing Protocol for Wireless Sensor Networks
Communcatons and Network, 2016, 8, 45-55 Publshed Onlne February 2016 n ScRes. http://www.scrp.org/journal/cn http://dx.do.org/10.4236/cn.2016.81006 Performance Comparson of a QoS Aware Routng Protocol
More informationHigh Performance DiffServ Mechanism for Routers and Switches: Packet Arrival Rate Based Queue Management for Class Based Scheduling
Hgh Performance DffServ Mechansm for Routers and Swtches: Packet Arrval Rate Based Queue Management for Class Based Schedulng Bartek Wydrowsk and Moshe Zukerman ARC Specal Research Centre for Ultra-Broadband
More informationVerification by testing
Real-Tme Systems Specfcaton Implementaton System models Executon-tme analyss Verfcaton Verfcaton by testng Dad? How do they know how much weght a brdge can handle? They drve bgger and bgger trucks over
More informationPerformance Analysis of Markov Modulated 1-Persistent CSMA/CA Protocols with Exponential Backoff Scheduling
Performance Analyss of Markov Modulated -Persstent CSMA/CA Protocols wth ponental Backoff Schedulng Pu ng Wong, Dongje Yn, and Tony T. Lee, Abstract. Ths paper proposes a Markovan model of -persstent CSMA/CA
More informationA Free-Collision MAC Proposal for Networks
12th Brazlan Workshop on Real-Tme and Embedded Systems 89 A Free-Collson MAC Proposal for 802.11 Networks Omar Alment 1,2, Gullermo Fredrch 1, Gullermo Reggan 1 1 SITIC Group Unversdad Tecnológca Naconal
More informationTECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z.
TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS Muradalyev AZ Azerbajan Scentfc-Research and Desgn-Prospectng Insttute of Energetc AZ1012, Ave HZardab-94 E-mal:aydn_murad@yahoocom Importance of
More informationThe Data Warehouse in a Distributed Utility Environment
The Data Warehouse n a Dstrbuted Utlty Envronment Charles A. Mllgan Dstngushed Engneer, Sun Mcrosystems Charles.mllgan@sun.com Abstract Utlty provsonng, Grd resource management, nstant copy kosks, and
More informationConcurrent Apriori Data Mining Algorithms
Concurrent Apror Data Mnng Algorthms Vassl Halatchev Department of Electrcal Engneerng and Computer Scence York Unversty, Toronto October 8, 2015 Outlne Why t s mportant Introducton to Assocaton Rule Mnng
More informationImproved Resource Allocation Algorithms for Practical Image Encoding in a Ubiquitous Computing Environment
JOURNAL OF COMPUTERS, VOL. 4, NO. 9, SEPTEMBER 2009 873 Improved Resource Allocaton Algorthms for Practcal Image Encodng n a Ubqutous Computng Envronment Manxong Dong, Long Zheng, Kaoru Ota, Song Guo School
More informationParallel matrix-vector multiplication
Appendx A Parallel matrx-vector multplcaton The reduced transton matrx of the three-dmensonal cage model for gel electrophoress, descrbed n secton 3.2, becomes excessvely large for polymer lengths more
More informationComparisons of Packet Scheduling Algorithms for Fair Service among Connections on the Internet
Comparsons of Packet Schedulng Algorthms for Far Servce among Connectons on the Internet Go Hasegawa, Takahro Matsuo, Masayuk Murata and Hdeo Myahara Department of Infomatcs and Mathematcal Scence Graduate
More informationDynamic Bandwidth Allocation Schemes in Hybrid TDM/WDM Passive Optical Networks
Dynamc Bandwdth Allocaton Schemes n Hybrd TDM/WDM Passve Optcal Networks Ahmad R. Dhan, Chad M. Ass, and Abdallah Sham Concorda Insttue for Informaton Systems Engneerng Concorda Unversty, Montreal, Quebec,
More informationMobileGrid: Capacity-aware Topology Control in Mobile Ad Hoc Networks
MobleGrd: Capacty-aware Topology Control n Moble Ad Hoc Networks Jle Lu, Baochun L Department of Electrcal and Computer Engneerng Unversty of Toronto {jenne,bl}@eecg.toronto.edu Abstract Snce wreless moble
More informationAn Optimal Bandwidth Allocation and Data Droppage Scheme for Differentiated Services in a Wireless Network
Purdue Unversty Purdue e-pubs ECE Techncal Reports Electrcal and Computer Engneerng 3--7 An Optmal Bandwdth Allocaton and Data Droppage Scheme for Dfferentated Servces n a Wreless Network Waseem Shekh
More informationHalmstad University Post-Print
Halmstad Unversty Post-Prnt Admsson Control for Swtched Realtme Ethernet Schedulng Analyss versus etwor Calculus Xng Fan and Magnus Jonsson.B.: When ctng ths wor cte the orgnal artcle. Orgnal Publcaton:
More informationA QoS-aware Scheduling Scheme for Software-Defined Storage Oriented iscsi Target
A QoS-aware Schedulng Scheme for Software-Defned Storage Orented SCSI Target Xanghu Meng 1,2, Xuewen Zeng 1, Xao Chen 1, Xaozhou Ye 1,* 1 Natonal Network New Meda Engneerng Research Center, Insttute of
More informationInternet Traffic Managers
Internet Traffc Managers Ibrahm Matta matta@cs.bu.edu www.cs.bu.edu/faculty/matta Computer Scence Department Boston Unversty Boston, MA 225 Jont work wth members of the WING group: Azer Bestavros, John
More informationComputer Communications
Computer Communcatons 3 (22) 3 48 Contents lsts avalable at ScVerse ScenceDrect Computer Communcatons journal homepage: www.elsever.com/locate/comcom On the queueng behavor of nter-flow asynchronous network
More information