6
MIPRO 2014, 26-30 May 2014, Opatija, Croatia An Approach using Siu!ation "# chni$u#s t o #stiat# %ua!ity o& S#r'ic# Para#t#rs in Counication (#t)or*s +r'oj# oa.ins*i/, P#tar n##'i / acu!ty o& !#ctrica! ngin##ring an Coputing5#part#nt o& una#nta!s o& !#ctrica! ngin##ring an M#asur##nts, 7agr#8, Croatia hr'oj#9*oacins*i:&#r9hr Quality of Service in communication networks is a request for minimal required performance needed for transferring IP packets in a network. The performance can be ensured with a set of mechanisms that guarantee required performance through shaping and policing of packet traffic. Their goal is to prevent congestion, packet delays and packet loss. The goal of this paper is to introduce some of the commonly used Quality of Service mechanisms and analye their impact on network traffic in order to improve its performance, mainly packet delay and buffer congestion. The mechanisms are analyed trough a simple packet transfer simulator that offers different combinations of mechanisms to be used in t he simulation, allowing different packet transfer scenario setups. !fter a simulation is complete, the simulator offers results along with relevant packet transfer graphs. The impact of Quality of Service mechanisms is analyed and interpreted trough those results. It is shown that combining Quality of Service mechanisms in packet transfer generally gives better results. r#gu!at# th# trans&#r, so that th# #sir# p#r&oranc# r#$uir##nts ar# #t ;1<9 "h#r# ar# i&&#r#nt approach#s to so!'ing ina#$uat# trans&#r p#r&oranc#, ain!y 8y introucing i&&#r#nt %oS #chaniss to counication n#t)or*s ;2, 3, 4<9 A!so, i&&#r#nt approach#s an r#co#nations ar# gi'#n through a nu8#r o& RCs, r#garing guarant## %oS ;=<, #&ining a in&oration o#! &or %oS n #t)or* anaging  po!ici#s ;6<, %oS M#chanis S#!#ction ;>< an oth#rs ;?, @<9 A!! o& th#s# approach#s h#!p r#uc# pac*#t #!a y an  pac*#t !oss9 "h# pro8!# this pap#r ana!y#s through a sip!# n#t)or* siu!ator is ho) #act!y %oS #chaniss ipact a counication n#t)or* p#r&oranc# in att#pt to ipro'# pac*#t trans&#r, priari!y 8y r#ucing pac*#t #!ays an pac*#t !oss associat# )ith th# 8u&&#r #chanis9 !lso, different combinations of mechanisms give different II9 %BAI"D O SREIC MC+A(ISMS results, allowing more fle"ibility in improving packet transfer in a desired way, but also making it more predictable and controllable. So# o& th# ain %oS #chaniss that ar# us# in counication n#t)or*s )i!! 8# #p!ain# in this  paragraph, a!ong )ith th#ir #sir# #&&#cts an shortcoings9 I9 I  ("ROBC"IO( #etransmission$ r#s#ns pac*#ts !ost in transission, Fasic #!##nts o& a coput#r n#t)or* inc!u# har)ar#, so&t)ar# an protoco!s9 "h# int#rr#!ationship o& th#s# #!##nts constitut#s th# in&rastructur# o& th# n#t)or*9 A n#t)or* in&rastructur# is th# topo!ogy in )hich th# no#s o& a n#t)or* ar# conn#ct# to #ach oth# r9 Conn#ctions in'o!'# rout#rs, s)itch#s, 8rig#s an hu8s conn#ct# 8y ca8!#s or )ir#!#ss t#chno!ogi#s9 F#caus# o& th# )ay th# counication n#t)or*s ar#  8ui!t, )ith any possi8!# trans&#r rout#s an )ith  possi8i!ity that so# rout#s can 8# cong#st# or s!o)#r than oth#rs, IP pac*#ts tra'#!!ing through th# n#t )or* can arri'# too !at#, )ith #rrors, in )rong or#r or 8# iscar# uring trans&#r9 "his caus#s trans&#r #rrors an #graation o& trans&#r p#r&oranc#9 %ua!ity o& S#r'ic# G%oSH is a s#t o & r#$uir##nts that guarant## c#rtain p #r&oranc# in pac*#t trans&#r9 A goo ip!##ntation o& %oS usua!!y inii#s pac*#t #!ay an r#uc#s th# nu8#r o & iscar# pac*#ts9 "o #nsur# such p#r&oranc#, %oS has a s#t o& #chaniss that h#!p thus pr#'#nting !oss o& in&oration9 It h#!ps )ith ac$uiring cop!#t# in&oration, 8ut )ith a cost - incr#as# trans&#r ti# an pac*#t #!ay9 %uffer$ contain#r us# to t#porari!y stor# pac*#ts that cannot 8# proc#ss# i#iat#!y upon arri'a!,  pr#'#nting r#transission to caus# 8igg#r ti# #!ay an unn### 8an)ith usag#9 +o)#'#r, incr#asing th#  8u&&#r si# a!so #t#ns pac*#t #!ay, un!#ss oth #r #chaniss ar# app!i# ;10<9 %uffer collecting strategies$ 'arious )ays o& co!!#cting pac*#ts &ro th# 8u&&#r, in'o!'ing 'arious crit#ria !i*# or#r o& arri'a!, pac*#t priority, #tc9 i&&#r#nt approach#s try to ipro'# on i&&#r#nt asp#cts o& %oS9 or instanc#, roun ro8in 8u&&#r co!!#cting strat#gy is us# to #'#n 8u&&#r #!ay ti#s o& pac *#ts in th# 8u&&#r9 Packet priority$ c!assi&i#s pac*#ts 8y th#ir iportanc# - iportant pac*#ts g#t high#r priority an )i!! 8#  proc#ss# &irst9 "his h#!ps r#ucing pac*#t !oss an #!ays o& pac*#ts )ith high#r priority, 8ut 8y incr#asing pac*#t #!ay o& !o)#r priority pac*#ts ;11<9 1642

An Approach Using Simulation Techniques to Estimate Quality of Service Parameters in Communication Networks-2014

  • Upload
    qhid

  • View
    222

  • Download
    0

Embed Size (px)

Citation preview

8/9/2019 An Approach Using Simulation Techniques to Estimate Quality of Service Parameters in Communication Networks-2…

http://slidepdf.com/reader/full/an-approach-using-simulation-techniques-to-estimate-quality-of-service-parameters 1/6

MIPRO 2014, 26-30 May 2014, Opatija, Croatia

An Approach using Siu!ation "#chni$u#s to#stiat# %ua!ity o& S#r'ic# Para#t#rs in

Counication (#t)or*s

+r'oj# oa.ins*i/, P#tar n##'i /

acu!ty o& !#ctrica! ngin##ring an Coputing5#part#nt o& una#nta!s o& !#ctrica! ngin##ring anM#asur##nts, 7agr#8, Croatia

hr'oj#9*oacins*i:&#r9hr 

Quality of Service in communication networks is a request

for minimal required performance needed for transferring

IP packets in a network. The performance can be ensured

with a set of mechanisms that guarantee requiredperformance through shaping and policing of packet traffic.

Their goal is to prevent congestion, packet delays and

packet loss. The goal of this paper is to introduce some of 

the commonly used Quality of Service mechanisms and

analye their impact on network traffic in order to improve

its performance, mainly packet delay and buffer congestion.

The mechanisms are analyed trough a simple packet

transfer simulator that offers different combinations of 

mechanisms to be used in the simulation, allowing different

packet transfer scenario setups. !fter a simulation is

complete, the simulator offers results along with relevant

packet transfer graphs. The impact of Quality of Service

mechanisms is analyed and interpreted trough those

results. It is shown that combining Quality of Servicemechanisms in packet transfer generally gives better results.

r#gu!at# th# trans&#r, so that th# #sir# p#r&oranc#r#$uir##nts ar# #t ;1<9

"h#r# ar# i&&#r#nt approach#s to so!'ing ina#$uat#trans&#r p#r&oranc#, ain!y 8y introucing i&&#r#nt %oS#chaniss to counication n#t)or*s ;2, 3, 4<9 A!so,i&&#r#nt approach#s an r#co#nations ar# gi'#nthrough a nu8#r o& RCs, r#garing guarant## %oS ;=<,#&ining a in&oration o#! &or %oS n#t)or* anaging po!ici#s ;6<, %oS M#chanis S#!#ction ;>< an oth#rs ;?,@<9 A!! o& th#s# approach#s h#!p r#uc# pac*#t #!ay an pac*#t !oss9 "h# pro8!# this pap#r ana!y#s through asip!# n#t)or* siu!ator is ho) #act!y %oS #chanissipact a counication n#t)or* p#r&oranc# in att#ptto ipro'# pac*#t trans&#r, priari!y 8y r#ucing pac*#t#!ays an pac*#t !oss associat# )ith th# 8u&&#r #chanis9

!lso, different combinations of mechanisms give different II9 %BAI"D O SREIC MC+A(ISMS

results, allowing more fle"ibility in improving packet

transfer in a desired way, but also making it more

predictable and controllable.

So# o& th# ain %oS #chaniss that ar# us# incounication n#t)or*s )i!! 8# #p!ain# in this paragraph, a!ong )ith th#ir #sir# #&&#cts anshortcoings9

I9 I ("ROBC"IO( #etransmission$ r#s#ns pac*#ts !ost in transission,

Fasic #!##nts o& a coput#r n#t)or* inc!u#har)ar#, so&t)ar# an protoco!s9 "h# int#rr#!ationship o& th#s# #!##nts constitut#s th# in&rastructur# o& th#n#t)or*9 A n#t)or* in&rastructur# is th# topo!ogy in)hich th# no#s o& a n#t)or* ar# conn#ct# to #ach oth#r9

Conn#ctions in'o!'# rout#rs, s)itch#s, 8rig#s an hu8sconn#ct# 8y ca8!#s or )ir#!#ss t#chno!ogi#s9

F#caus# o& th# )ay th# counication n#t)or*s ar# 8ui!t, )ith any possi8!# trans&#r rout#s an )ith possi8i!ity that so# rout#s can 8# cong#st# or s!o)#r than oth#rs, IP pac*#ts tra'#!!ing through th# n#t)or* canarri'# too !at#, )ith #rrors, in )rong or#r or 8# iscar#uring trans&#r9 "his caus#s trans&#r #rrors an#graation o& trans&#r p#r&oranc#9

%ua!ity o& S#r'ic# G%oSH is a s#t o& r#$uir##nts thatguarant## c#rtain p#r&oranc# in pac*#t trans&#r9 A gooip!##ntation o& %oS usua!!y inii#s pac*#t #!ayan r#uc#s th# nu8#r o& iscar# pac*#ts9 "o #nsur#such p#r&oranc#, %oS has a s#t o& #chaniss that h#!p

thus pr#'#nting !oss o& in&oration9 It h#!ps )ithac$uiring cop!#t# in&oration, 8ut )ith a cost -incr#as# trans&#r ti# an pac*#t #!ay9

%uffer$ contain#r us# to t#porari!y stor# pac*#tsthat cannot 8# proc#ss# i#iat#!y upon arri'a!, pr#'#nting r#transission to caus# 8igg#r ti# #!ay anunn### 8an)ith usag#9 +o)#'#r, incr#asing th# 8u&&#r si# a!so #t#ns pac*#t #!ay, un!#ss oth#r #chaniss ar# app!i# ;10<9

%uffer collecting strategies$ 'arious )ays o& co!!#cting pac*#ts &ro th# 8u&&#r, in'o!'ing 'ariouscrit#ria !i*# or#r o& arri'a!, pac*#t priority, #tc9 i&&#r#ntapproach#s try to ipro'# on i&&#r#nt asp#cts o& %oS9or instanc#, roun ro8in 8u&&#r co!!#cting strat#gy is us#to #'#n 8u&&#r #!ay ti#s o& pac*#ts in th# 8u&&#r9

Packet priority$ c!assi&i#s pac*#ts 8y th#ir iportanc#- iportant pac*#ts g#t high#r priority an )i!! 8#

 proc#ss# &irst9 "his h#!ps r#ucing pac*#t !oss an #!ayso& pac*#ts )ith high#r priority, 8ut 8y incr#asing pac*#t#!ay o& !o)#r priority pac*#ts ;11<9

1642

8/9/2019 An Approach Using Simulation Techniques to Estimate Quality of Service Parameters in Communication Networks-2…

http://slidepdf.com/reader/full/an-approach-using-simulation-techniques-to-estimate-quality-of-service-parameters 2/6

Policing mechanisms$ #chaniss us# to iscar pac*#ts &ro th# 8u&&#r to pr#'#nt its cong#stion, a!!o)ingn#)#r pac*#ts to #nt#r th# 8u&&#r in &a'our o& iscar# pac*#ts9 "h# ost coon po!icing #chaniss ar# RGRano ar!y #t#ction a!so iscar or ropH;12< anJR GJ#ight# R, us#s pac*#t priority in #ciing)hich pac*#ts to iscarH9 I& th# priority #chanis isus#, JR can iscar !#ss signi&icant pac*#ts to pr#'#nt cong#stion an a!!o) &utur# high#r priority pac*#ts to 8# r#c#i'# ;3, 13<9

Shaping mechanisms$ #chaniss us# to shap# th#s#ning &!o) o& pac*#ts to pr#'#nt 8u&&#r cong#stion, pac*#t rop an r#transission, a'oiing unn#c#ssary 8an)ith usag#, possi8!# cong#stion an pac*#t #!ays9"h# ost coon ar# th# "o*#n Fuc*#t an #a*yFuc*#t a!goriths ;14, 1=<9

"AF I9 SIMBA"IO( SC(ARIOS A( MC+A(ISMS BS

r#c#i'#r sp## - s!o)o)n in i!!is#consH an choos#th# #sir# %oS #chaniss, )ith th# #c#ption o& r#transission, as it cannot 8# isa8!#9

A&t#r th# siu!ation is cop!#t#, thr## graphs ar#g#n#rat#K t)o pac*#t trans&#r #!ay graphs G)ith an)ithout ar*# prioriti#sH that sho) th# ti# #!ay o& #ach pac*#t in th# 8u&&#r an a 8u&&#r capacity graph that

III9 "+ SIMBA"OR O %BAI"D O SREIC sho)s ho) th# 8u&&#r capacity chang# uring th#MC+A(ISMS

In or#r to 8#tt#r stuy th# ipact o& i&&#r#nt %oS#chaniss on a counication n#t)or*, a n#t)or* siu!ator )as #'#!op#9 Its &unction is to siu!at# asip!# "CP pac*#t trans&#r 8#t)##n t)o n#t)or* no#s,in a u88#!! topo!ogy, )ith i&&#r#nt %oS #chanissin'o!'# an trac* 8u&&#r capacity, pac*#t !oss an pac*#tti# #!ay9 "h# #chaniss ip!##nt# in th# siu!ator ar#K r#transission, 8u&&#r, 8u&&#r co!!#cting strat#gi#sG&irst pac*#t in th# 8u&&#r or roun ro8inH, pac*#t priority)ith 3 prioriti#s G0-2, 0 8#ing th# high#stH an th# "o*#nFuc*#t a!gorith9

"h# siu!ator in its #ss#nc# is a u!tithr#a# La'aapp!ication, in )hich th# s#n#r thr#a g#n#rat#s pac*#ts

an s#ns th# to th# 8u&&#r, )hi!# th# r#c#i'#r thr#aco!!#cts pac*#ts &ro th# 8u&&#r9 "h#r# is a!so a thirthr#a that co!!#cts r#!#'ant trans&#r ata9 Pac*#tg#n#ration is a sip!#, 8#st #&&ort so!ution, pac*#t !#ngthis &i# an pac*#t priority is g#n#rat# rano!y9

"h# siu!ator is contro!!# through th# ain scr##n9"h# us#r can s#t up th# #sir# siu!ation para#t#rsG8u&&#r si#, nu8#r o& pac*#ts g#n#rat#, s#n#r an

trans&#r9 "h# siu!ator a!so g#n#rat#s !og &i!#s an

isp!ays r#!#'ant trans&#r ataK tota! trans&#r ti#, nu8#r o& iscar# pac*#ts Gtota! an i'i# 8y prioriti#sH,nu8#r o& pac*#ts iscar# in r#transission ana'#rag# pac*#t ti# #!ay insi# th# 8u&&#r Gtota! ani'i# 8y prioriti#sH9 An #ap!# o& a &inish# siu!ationis sho)n in igur# 19

"h#r# ha'# 8##n a !ot o& i&&#r#nt stui#s o& %oS#chaniss )h#r# a siu!ation approach )as us#, 8utth#y )#r# ost!y us# in sp#ci&ic ar#as !i*# Coposit#J#8 S#r'ic#s ;16<, EoIP ;1><, Ei#o "ra&&ic ;1?<, Mo8i!#A +oc (#t)or*s ;1@<, BM"S "#rr#stria! Raio Acc#ss (#t)or*s ;20< or #'a!uating cop!# syst#s ;21<9 A!so,th#r# ar# iscr#t# #'#nt siu!ators targ#t# at n#t)or*ing

r#s#arch9 On# #ap!# is th# ns-2 siu!ator ;22<9 "h#siu!ator o& this pap#r is construct# to ip!##nt on!yth# n### copon#nts an its #!i8#rat# sip!i&icationa!!o)s 8#tt#r #phasis an c!#ar#r stuy on ho) th# 8u&&#r han!#s incoing pac*#ts )ith i&&#r#nt %oS#chaniss in'o!'# an ho) o thos# #chanissipact pac*#t #!ay an cong#stion in th# 8u&&#r9

"h#r# is a!so a att#r o& pac*#t g#n#ration9 Jhi!#using Poisson istri8ution &or pac*#t g#n#ration ;23<

 pro'i#s a or# r#a!istic siu!ation, it )as notn#c#ssary &or th# ana!ysis o& o8s#r'# %oS#chaniss an th#ir ipact on 8u&&#r capacityan pac*#t ti# #!ay9

"h# ipact on pac*#t trans&#r )i!! 8#o8s#r'# an #asur# in &our trans&#r sc#narios9 In #'#ry sc#nario, th# 8asic trans&#r  para#t#rs )i!! r#ain constantK

ξ

ξ

ξ

ξ

ξ

Fu&&#r si#K 60

Pac*#ts to s#nK 140

S#n#r sp##K 1

R#c#i'#r sp##K 3

"o*#nsK 60 Gaiu nu8#r o& 

to*#ns &or th# "o*#n Fuc*#t a!gorithH

igur# 19 Main siu!ator scr##n a&t#r a cop!#t# siu!ation

1643

Scenarios Priority%uffer

collecting

Token

bucket

Sc#nario 1 (o irst a'ai!a8!# (o

Sc#nario 2 D#s irst a'ai!a8!# D#s

Sc#nario 3 (o Roun ro8in (o

Sc#nario 4 D#s Roun ro8in D#s

8/9/2019 An Approach Using Simulation Techniques to Estimate Quality of Service Parameters in Communication Networks-2…

http://slidepdf.com/reader/full/an-approach-using-simulation-techniques-to-estimate-quality-of-service-parameters 3/6

igur# 29 Pac*#t #!ay uring trans&#r graph, sc#nario 1 Gno #chaniss, 8asic co!!#ctingH

igur# 39 Fu&&#r capacity uring trans&#r graph, sc#nario 1 Gno #chaniss, 8asic co!!#ctingH

ach sc#nario )as run 2= ti#s an a'#rag# 'a!u#s ar# pr#s#nt#9 "his )as n#c#ssary to r#uc# #asuring #rrors, 8#caus# th# siu!ation )as run on a p#rsona! coput#r )hich contain# oth#r running proc#ss#s9

sc#nario )i!! 8# #p!ain# in or# #tai!, a!ong )ith itsr#su!ts, through on# typica! siu!ation o& #ach sc#nario9

 A. Scenario 1 - no mechanisms, basic collecting 

"h# &irst sc#nario us#s on!y th# una'oia8!#IE9 SIMBA"IO( SC(ARIOS #chaniss Gr#transission an 8u&&#rH a!ong )ith th#

"h# siu!ation is i'i# in &our i&&#r#nt sc#narios,#p#ning on th# co8ination o& %oS #chaniss us#,as s##n in "a8!# 19 A!! sc#narios us# r#transission anth# 8u&&#r #chanis )ith on# o& th# 8u&&#r co!!#ctingstrat#gi#s9 Optiona! #chaniss ar# pac*#t priority, 8u&&#r co!!#cting strat#gy an th# to*#n 8uc*#t a!gorith9 ach

 8u&&#r co!!#cting strat#gy that co!!#cts th# &irst a'ai!a8!# pac*#t in th# 8u&&#r9

#!ays o& pac*#ts can 8# s##n on igur# 29 "h# graph

ar*s i&&#r#nt pac*#t prioriti#s )ith i&&#r#nt co!ours, 8ut 8#caus# th# priority #chanis isnt acti'at#, pac*#t priority has no #&&#ct on #!ay ti#s9 "h#r# ar# thr##

1644

8/9/2019 An Approach Using Simulation Techniques to Estimate Quality of Service Parameters in Communication Networks-2…

http://slidepdf.com/reader/full/an-approach-using-simulation-techniques-to-estimate-quality-of-service-parameters 4/6

)#r# ropp# in trans&#r G12 priority 0, 12 priority 1 an14 priority 2H9 A'#rag# pac*#t ti# #!ay is 161 s G164s priority 0, 1=4 s priority 1 an 16? s priority 2H9"h# r#su!ts c!#ar!y sho) that )ithout a priority#chanis a!! pac*#ts ar# tr#at# #$ua!!y9 "his is th# 8asic sc#nario on )hich a!! oth#rs )i!! try an ipro'#9

 B. Scenario 2 - all mechanisms, basic collecting 

"his sc#nario as th# priority #chanis to sc#nario1, to o8s#r'# an #asur# its #&&#cts on pac*#t trans&#r,s##n on igur# 49 Pac*#ts ar# c!#ar!y istinguish# 8y th#ir  priorityK priority 0 has th# !o)#st ti# #!ay, &o!!o)# 8y priority 1 an priority 2 )ith th# 8igg#st ti# #!ay9

"his sc#nario a!so introuc#s th# "o*#n Fuc*#t

igur# 49 Pac*#t #!ay uring trans&#r graph, sc#nario 2 - a!!#chaniss, 8asic co!!#cting

istinct !in#s on th# graph, on# horionta! !in# an t)o!in#s )ith !in#ar gro)th9 "h# horionta! !in# is caus# 8y

th# 8u&&#r co!!#cting strat#gy, )hich co!!#cts th# &irst pac*#t in th# 8u&&#r it co#s across, r#su!ting in !o) ti##!ays9 "h# &irst !in#ar gro)th is caus# 8y th# #!ay pac*#ts ha'# )hi!# )aiting in th# 8u&&#r9 "h# s#con!in#ar gro)th is th# r#su!t o& r#transission that start#a&t#r approiat#!y ?= pac*#ts )#r# s#nt an th# 8u&&#r )as cong#st#9 "h# r#transission ti# #!ay app#arssa!!#r, 8ut it is in &act !arg#r9 "his graph sho)s on!y ti##!ays o& pac*#ts )hi!# )aiting in th# 8u&&#r9 Pac*#ts that)#nt into r#transission )#r# s#nt, r#j#ct#, r#transitt#again an th#n )ait# in th# 8u&&#r, so th#ir tota! ti##!ay is !arg#r than that o& th# pac*#ts that i#iat#!y)#nt into th# 8u&&#r9

Anoth#r r#!#'ant graph is th# 8u&&#r capacity graph,sho)n in igur# 39 "h# 8u&&#r capacity )as #pon#ntia!!y&i!!# to th# point o& cong#stion, at approiat#!y 13= s9A&t#r that, surp!us pac*#ts )#r# iscar# an s#nt intor#transission9 At that point, th# r#c#i'#r co!!#ct# pac*#ts, 8ut th# s#n#r &i!!# th# #pty 8u&&#r spot$uic*!y, as is sho)n on th# graph )ith th# horionta! !in#spanning to 200 s in trans&#r9 A&t#r that th# rop ins#ning rat# 8#caus# o& th# s!o) r#transission rat# thath#!p# !o)#r th# 8u&&#r capacity, unti! a!! pac*#ts )#r#co!!#ct# &ro th# 8u&&#r9

"h# tota! trans&#r ti# )as =0> s, tota! o& 3? pac*#ts

igur# =9 Fu&&#r capacity uring trans&#r graph, sc#nario 2 - a!!#chaniss, 8asic co!!#cting

a!gorith9 "h# i#a 8#hin th# a!gorith is to contro! th#rat# at )hich pac*#ts ar# s#nt9 "h# a!gorith p#rioica!!y&i!!s a contain#r G8uc*#tH )ith to*#ns, up to a s#t capacity9A pac*#t can on!y 8# s#nt i& th#r# ar# su&&ici#nt to*#ns inth# 8uc*#t to NpayN &or s#ning9 At &irst, )ith #nough

to*#ns at th# 8#ginning o& trans&#r, pac*#ts can 8# s#nt ina !arg# 8urst, unti! th# contain#r has no or# to*#ns9 A&t#r that, th# p#rioica! app#aranc# o& to*#ns contro!s th# s#nrat#9 "his h#!ps 8u&&#r cong#stion, r#uc#s th# n## &or r#transission an 8#caus# o& that a!so r#uc#s pac*#trop rat#s an #!ay ti#s ;14<9

"h# tota! trans&#r ti# )as 446 s an no pac*#ts)#r# ropp# in trans&#r9 A'#rag# pac*#t ti# #!ay is12> s G@ s priority 0, 11= s priority 1 an 2>2 s priority 2H9 In this cas#, co8ining 8oth #chanissr#a!!y pai o&&, #sp#cia!!y )h#n o8s#r'ing th# priority 0#!ay ti#s9 "h# ti# #!ay rop at th# #n o& th# graph isth# r#su!t o& th# "o*#n Fuc*#t a!gorith that r#uc# th#

s#n rat# onc# it ran out o& to*#ns9 "h# shaping#chanis h#!p# )ith pac*#t rops, 8ut this is ani#a!i# #ap!#, )h#r# th# a!goriths to*#ns ar#r#&i!!# at th# sa# rat# th# r#c#i'#r can r#c#i'# a pac*#t9In a r#a! !i&# situation, th# a!gorith )ou! r#$uir# acar#&u! para#t#r ca!i8ration to g#t as c!os# to th#i#a!i# #ap!# as possi8!#9 +igh#r s#ning rat# )ou!caus# pac*#t rops an !o)#r rat# )ou! caus# 8igg#r  pac*#t ti# #!ay9

"h# 8u&&#r capacity graph Gigur# =H !oo*s sii!ar toth# graph in sc#nario 1, on!y th# horionta! !in# )h#r# th#s#n#r an th# r#c#i'#r cop#t# )hi!# th# 8u&&#r is at &u!!

igur# 69 Pac*#t #!ay uring trans&#r graph, sc#nario 3 - no#chaniss, roun ro8in

164=

8/9/2019 An Approach Using Simulation Techniques to Estimate Quality of Service Parameters in Communication Networks-2…

http://slidepdf.com/reader/full/an-approach-using-simulation-techniques-to-estimate-quality-of-service-parameters 5/6

igur# ?9 A'#rag# pac*#t ti# #!ay #p#ning on pac*#t priority

igur# >9 Pac*#t #!ay uring trans&#r graph, sc#nario 4 - a!!#chaniss, roun ro8in

E9 SIMBA"IO( RSB"S

capacity !asts !ong#r an #ns )ith a sooth !in#ar rop9

C. Scenario 3 - no mechanisms, round robin

In this an th# &o!!o)ing sc#nario, th# 8u&&#r co!!#ctingstrat#gy is chang#9 Pac*#ts )i!! 8# co!!#ct# using th#roun ro8in strat#gy9 "his sc#nario has no aitiona!#chaniss9 "h# #&&#ct this co!!#cting strat#gy has on pac*#t trans&#r can 8# s##n in igur# 69

"h# i#a o& th# roun ro8in co!!#cting strat#gy is toipro'# th# 8asic co!!#cting strat#gy9 F#caus# o& th# )ay pac*#ts ar# co!!#ct#, this strat#gy #'#ns ti# #!aysacross a!! pac*#ts that #nt#r th# 8u&&#r9 In th# graph this issho)n as a horionta! !in# in th# i!#9 ")o !in#ar gro)ths app#ar 8#caus# th# s#n#r is &i!!ing th# 8u&&#r &ast#r than th# r#c#i'#r can co!!#ct pac*#ts an 8#caus# o& 

in#'ita8!# r#transission onc# th# 8u&&#r is cong#st#9

"h# tota! trans&#r ti# )as 4?6 s, tota! o& 41 pac*#ts)#r# ropp# in trans&#r G1@ priority 0, 12 priority 1 an10 priority 2H9 A'#rag# pac*#t ti# #!ay is 166 s G1?2s priority 0, 1=> s priority 1 an 1=? s priority 2H9

"h# 8u&&#r capacity graph is sii!ar to th# sa# graphin sc#nario 19

 D. Scenario 4 - all mechanisms, round robin

In this sc#nario 8oth th# priority #chanis an th#shaping #chanis ar# co8in# )ith th# roun ro8in 8u&&#r co!!#ction strat#gy9 "h# #&&#ct this has on pac*#t

A&t#r a!! th# sc#narios )#r# run 2= ti#s, a'#rag#'a!u#s o& o8s#r'# para#t#rs )#r# ca!cu!at#9 "h# ost

int#r#sting para#t#rs to ana!y# ar# a'#rag# pac*#t ti##!ays #p#ning on pac*#t priority, gi'#n in igur# ?,an th# a'#rag# trans&#r ti#s, gi'#n in igur# @9

I& a'#rag# trans&#r ti#s ar# o8s#r'#, th# conc!usionis that sc#nario 3 Gno priority, roun ro8in, no "o*#nFuc*#tH has th# !o)#st trans&#r ti# o& 46@9>2 s9 "h#high#st trans&#r ti# can 8# s##n in sc#nario 4 Grounro8in, priority, "o*#n Fuc*#tH )ith =>=9? s, &o!!o)#c!os#!y 8y sc#nario 2 G=>=924 sH9 "h# conc!usion that can 8# ra)n &ro this &igur# is that i& !o)#st possi8!# trans&#r ti# is #sir# in a counication n#t)or*, in this cas#,on!y a 8u&&#r an th# roun ro8in co!!#cting strat#gy )i!!su&&ic#9

On th# oth#r han, i& a'#rag# pac*#t ti# #!ay#p#ning on pac*#t priority is o8s#r'# &or #achsc#nario, th#r# ar# t)o istinct patt#rns9 "h# #'#n co!unso& sc#narios 1 an 3 ar# th# r#su!t o& th# sc#narios notha'ing th# priority #chanis, an ar# not r#!#'ant &or this o8s#r'ation9 Oth#r t)o sc#narios ha'# or#int#r#sting r#su!ts9 "h# !o)#st ti# #!ay &or priority 0 can 8# s##n in sc#nario 2 G=904 sH, )ith sc#nario 4 c!os# 8yG6932 sH9 "h#y a!so ha'# th# !o)#st ti# #!ay o& priority1 pac*#ts G110916 s an 10@9>2 s r#sp#cti'#!yH9+o)#'#r, th#ir #!ay ti#s o& priority 2 pac*#ts issu8stantia!!y Gor# than t)o ti#sH high#r than insc#narios 1 an 39

trans&#r is 'isi8!# on igur# >9

"h# tota! trans&#r ti# )as 4=0 s an no pac*#ts )#r# ropp# in trans&#r9 A'#rag# pac*#t ti# #!ay is 133 s G? s priority 0,120 s priority 1 an 2>0 s priority 2H9 Inthis cas#, co8ining 8oth #chaniss a!so pai o&&, !i*# in sc#nario 2, #sp#cia!!y )h#no8s#r'ing th# priority 0 #!ay ti#s9 "h# ti##!ay rop at th# #n o& th# graph is th# r#su!to& th# "o*#n Fuc*#t a!gorith that r#uc#th# s#n rat# onc# it ran out o& to*#ns9

"h# 8u&&#r capacity graph is sii!ar to th#

on# in sc#nario 29

igur# @9 A'#rag# trans&#r ti#s &or a!! sc#narios

1646

or 8#st priority 0 an 1 pac*#t ti# #!ay, th# priority

8/9/2019 An Approach Using Simulation Techniques to Estimate Quality of Service Parameters in Communication Networks-2…

http://slidepdf.com/reader/full/an-approach-using-simulation-techniques-to-estimate-quality-of-service-parameters 6/6

#chanis shou! 8# us# a!ong )ith th# "o*#n Fuc*#ta!gorith, r#gar!#ss o& th# 8u&&#r co!!#cting strat#gy9Roun ro8in strat#gy o#s gi'# a 8it high#r priority 0#!ay, 8ut has 8#tt#r #!ays &or oth#r prioriti#s, )hich cana!so 8# r#!#'ant, #p#ning on th# situation9

EI9 CO(CBSIO(

"his pap#r #a!s )ith IP pac*#t #!ay, th# ipact it hason counication n#t)or*s an ho) it can 8# r#uc# toipro'# %ua!ity o& S#r'ic#, )ith th# 8u&&#r as th# aino8s#r'ation point9 A siu!ator )as #'#!op# to &urth#r stuy th# ipact #ach o& th# #chaniss has on 8u&&#r 

 pac*#t #!ays an 8u&&#r cong#stion, a!on# or co8in#)ith oth#r #chaniss9 "h# r#su!ts sho) that a c!#ar so!ution to %oS in counication n#t)or*s o#s not#ist9 Inst#a, th# so!ution gr#at!y #p#ns on th# n#t)or* its#!& an th# ini'iua! n##s &or that particu!ar n#t)or*9"h# r#su!ts ha'# a!so sho)n that i&&#r#nt co8inations o& %oS #chaniss ipro'# n#t)or* tra&&ic in i&&#r#nt)ays9 I& th#r# is a n## &or &ast pac*#t trans&#r, r#gar!#sso& pac*#t priority, th# roun ro8in 8u&&#r co!!#ctionstrat#gy can pro'# su&&ici#nt9 I&, on th# oth#r han, pac*#tsha'# i&&#r#nt prioriti#s or so# ar# or# s#nsiti'# to#!ays than oth#rs, priority #chanis shou! 8# us#co8in# )ith th# "o*#n Fuc*#t shaping #chanis9

Oth#r pr#'ious!y #ntion# stui#s a!so sho) thatn#t)or* tra&&ic ipro'#s )ith th# introuction o& %oS#chaniss9 Again, th#ir conc!usions ar# ra)n &rosiu!ations o& sp#ci&ic, cop!# syst#s, )ith r#a! !i&#app!ication, )hi!# siu!ation in this pap#r )as#!i8#rat#!y sip!i&i# to #phasi# ho) %oS#chaniss a&&#ct pac*#t #!ay an 8u&&#r cong#stion9

A!though an u!tiat# so!ution o#s not #ist, a car#&u!s#!#ction o& #chaniss can pro'i# a goo so!ution, 8#caus# using th# %oS #chaniss a*#s pac*#t trans&#r or# c!#ar, or# pr#icta8!# an or# anag#a8!#,#!iinating or !#ss#ning th# un#sira8!# #&&#cts in n#t)or* trans&#r9 "his a!so a!!o)s #c!aring inia! p#r&oranc#guarant##s n#c#ssary )h#n #sta8!ishing %oS in acounication n#t)or*;1, 24, 2=<9 As a n#t st#p, th#siu!ator )i!! ip!##nt or# a'anc# %oS#chaniss an )i!! 8# ipro'# to support !arg#r tra&&ic&!o)s an a!!o) &!o) custoiation to &urth#r stuy th#ipact o& %oS #chaniss on i&&#r#nt tra&&ic &!o)s9

R R(CS

;1< Sig#ti, "9 +attingh, C9 G2004H9 End-to-End QoS et!or" 

 Design# Qualit$ o% Ser&ice in 'As, (As, and )*s, Cisco Pr#ss,

Inianapo!is;2< S$stem i et!or"ing Qualit$ o% Ser&ice +)ersion elease 1 , IFM

Corp9, 200@9, A&ailable %rom#

httpK55pu8!i898ou!#r9i89co5in&oc#nt#r5is#ri#s5'6r105topic5ra* 

?5ra*?9p&, Accessed# 201491=919;3< Qualit$ o% Ser&ice et!or"ing , Cisco Pr#ss, A&ailable %rom#

httpK55oc)i*i9cisco9co5)i*i5%ua!ityo&S#r'ic#(#t)or*ing, Accessed# 201491>919

;4< Enter/rise QoS Solution e%erence et!or" Design 0uide, CiscoPr#ss, San Los#

;=< Sh#n*#r, S9, Partrig#, C9, Qu#rin, R9 G1@@>H9 S/eci%ication o% 

0uaranteed Qualit$ o% Ser&ice, RC 2212, S#pt#8#r 1@@>9

;6< Snir, D9, Ra8#rg, D9, Strassn#r, L9, Coh#n, R9, Moor#, F9 G2003H9 *olic$ Qualit$ o% Ser&ice +QoS n%ormation odel , RC 3644, (o'#8#r 20039

;>< Po!*, L9, h#si*an, S9, Caari!!o, Q9 G200@H9 Qualit$ o% Ser&ice

+QoS echanism Selection in the Session Descri/tion *rotocol +SD*, RC =432, March 200@9

;?< Chan, 9, Sahita, R9, +ahn, S9, McC!oghri#, 9 G2003H9 Di%%erentiated Ser&ices Qualit$ o% Ser&ice *olic$ n%ormation Base,RC 331>, March 20039

;@< Cra)!#y, 9, (air, R9, Rajagopa!an, F9, Sanic*, +9 G1@@?H9 A rame!or" %or QoS-based outing in the nternet , RC 23?6,August 1@@?9

;10< Q#ttys, L9 (icho!s, 9 G2011H9 Bu%%erbloat# Dar" Bu%%ers in the

 nternet , %u#u# - Eirtua!iation, Eo!u# @, Issu# 119;11< #, P9, i, D9, (i, 9 G2013H9 ulti/le Ser&ices Scheduling !ith

 *riorit$ Queuing odel , ILCI", Eo!u# 3, Issu# 1, A&ailable %rom#

httpK55ijcit9org5ijcitpap#rs5'o!3no15ILCI"-120>029p&, Accessed#201492=919

;12< Pati!, Q9, McC!#an, S9, Qaura', R9 G2011H9 Dro/ ail and ed 

Queue anagement !ith Small Bu%%ers# Stabilit$ and 56*  Bi%urcation, IC"AC", Eo!u# 2, Issu# 2, A&ailable %rom#

httpK55#prints9u!st#r9ac9u*521@?>515jctSp!Pap#r33@3449p&, Accessed# 201492=919

;13< rancis-Co8!#y, P9 G2004H9 Congestion Control , unpu8!ish#, A&ailable %rom#

httpK55&tp9utc!uj9ro5pu85us#rs5c#i!5prc5CO(QS"IO(20CO("R O9ppt, Accessed# 201491>919

;14< A!i Mantar, +9 G2000H9 he o"en Buc"et +'ea"$ Buc"et odel ,Syracus# Bni'#rsity, unpu8!ish#, A&ailable %rom#httpK55$8on#9int#rn#t29#u5885Fuc*#t9oc, Accessed# 201491=919

;1=< oo*os Foton, F9 G2001H9 ra%%ic Sha/ing , unpu8!ish#, A&ailable %rom# httpK55&tp9utc!uj9ro5pu85us#rs5tarc58iro9ppt, Accessed#2714.18.1.

;16< Si!'#r, Q9, Mau*o, A9, La&ri, R9, Mi!!#r, L9A9, Sh#th, A9P9 G2003H9 odeling and Simulation o% Qualit$ o% Ser&ice %or Com/osite (ebSer&ices in Proc##ings o& th# >th Jor! Mu!ticon&#r#nc# onSyst#ics, Cy8#rn#tics, an In&oratics GSCI03H, Or!ano, ,Lu!y 2003, (agi8 Ca!!aos, Anna M9 i Sciu!!o, "oshiui Ohta,an "#-ai iu Gs9H, Int#rnationa! Institut# o& In&oratics an

Syst#ics, 2003, pp9 420-42=9;1>< A!-(aaany, A9, Fourouc#n +9, A!-M#nthari J9 G200?H9 odeling and Simulation o% Qualit$ o% Ser&ice in )o* (ireless

 'A , Lourna! o& Coputing an In&oration "#chno!ogy, 'o! 16,no9 2, pp9 131142

;1?< Ch#n, F9, G2002H9 Simulation and Anal$sis o% Qualit$ o% Ser&ice *arameters in * et!or"s !ith )ideo ra%%ic

;1@< O!aosu, L9 F9, O A!ao, A9, Aigun M9, uoyi8o&arh#, L9O9 G2006H9 Design and Simulation o% a Qualit$ o% Ser&ice +QoS0uaranteed odel %or obile Ad 5oc et!or"s +Aets,SA"(AC 2006, pap#r no9 2=0

;20< QarcTa, A9 F9, U!'ar#-Capana, M9, EV$u#, 9, F#rroca!, L9G2002H9 Simulation o% Qualit$ o% Ser&ice echanisms in the 9S errestrial adio Access et!or" 9 Proc9 o& th# 4th I Con&#r#nc#on Mo8i!# an Jir#!#ss Counications (#t)or*s GMJC(2002H9 Stoc*ho!, S)##n, @-11 S#pt#8#r 20029 pp9 1>4-1>?9

;21< ant#rs, O9 G2010H9 %oS ana!ysis 8y siu!ation in R#o;22< he ns anual , unpu8!ish#, A&ailable %rom#httpK55)))9isi9#u5nsna5ns5ns-ocu#ntation9ht!, Accessed#20149029049

;23< !#inroc*, 9 G1@>6H9 Queueing S$stems, )ol. 2# Com/uter  A//lications, Ji!#y Int#rsci#nc#, (#) Dor* 

;24< oacins*i, +9 n##'ic, P9 G2011H9  Con%iguration o% Qualit$ o% Ser&ice *arameters in Communication et!or"s, Anna!s o& AAAM &or 2011 Proc##ings o& th# 22n Int#rnationa!AAAM Syposiu, 23-26th (o'#8#r 2011, Ei#nna, Austria,Eo!u# 22, (o9 1, ISS( 1>26-@6>@, ISF( @>?-3-@01=0@-?3-4,ata!inic, F9 G9H, pp9 136@-13>0, Pu8!ish# 8y AAAMInt#rnationa! Ei#nna, Ei#nna

;2=< oacins*i, +9 n##'ic, P9 G2013H9 Con%iguration o% Qualit$ o% 

Ser&ice *arameters in Communication et!or"s, 24nInt#rnationa! AAAM Syposiu, 23-26th (o'#8#r 2013,7aar, Croatia, Proc#ia ngin##ring, 'o! 6@, pp9 6==-6649

164>