Ac as de las XV Jo nadas
de Ingenie ía Telemá ica
(JITEL 2021),
A Co uña (España),
27-29 de oc ub e de 2021.
Hyb id Au onomous Connec ed Vehicle
pla ooning wi h Fede a ed Lea ning: S a e o he
a and simula ion F amewo k
P ashan h Kannan, Yaqoob Al-Zuhai i, Mónica Aguila Iga ua
Depa men o Ne wo k Enginee ing
Uni e si a Poli ècnica de Ca alunya (UPC)
Ba celona, Spain
p [email protected], yaqoob.al-zuhai[email p o ec ed], monica.aguila @upc.edu
In his pape , we p esen a b ie o e iew abou echnologies
and s a e o a me hods being used o Hyb id Au onomous
Connec ed Vehicles (HACV). Mo eo e , Fede a ed Lea ning
(FL) help in a oiding ansmission o aw local da a in he design
o machine lea ning models o di e se pu poses, which help o
ensu e p i acy o sensible da a. On he o he hand, o educe
ailpipe emissions hyb id elec ic ehicles a e equi ed, as
comple e con e sion o engines o elec ic migh ake ime.
Communica ion and connec i i y o ehicles and in as uc u e
a e inc easing o connec ed au onomous ehicles (CAV) ha
play a i al ole in u u e anspo a ion. High speed, eliable and
e icien communica ion be ween ehicles and in as uc u e is
made possible wi h i h gene a ion (5G) wi eless echnology.
He e, Pla ooning and 5G helps in joining a clus e o ehicles ha
aids in educing uel consump ion and allowing exchange o
ene gy using wi eless cha ging.
Keywo ds: Pla ooning, FL, VANET, V2X, Au onomous ehicles
1. INTRODUCTION
The pu pose/goal o his a icle is o add ess he gap in
u ilizing di e se wi eless echnologies o ehicula ne wo ks
and c ea ing a new ede a ed lea ning a chi ec u e o
p o ec ion o da a. In his scena io pla ooning o ehicles is
in eg a ed wi h ehicula communica ion echnology using
ei he 5G/Ad ance long e m e olu ion (LTE) o IEEE
802.11p. This amewo k migh help o inc ease sa e y and
e iciency o ehicles and allows wi eless cha ging o ehicle
ba e ies. We also ocus on analyzing he impac o
communica ion ailu e o pla oon ehicles in case o mixed
a ic condi ions ha ing manned and unmanned ehicles
oge he . The es o he pape p o ides a b ie ou lay o cu en
echnology embedded wi h hyb id ehicles ( uel and elec ic),
connec ed ehicles (5G and/o IEEE 802.11p), ehicula
communica ion and ede a ed lea ning (FL), along wi h
di e en simula o s ha a e use ul in de eloping he scena ios
and es ou p oposals.
1.1. Sha ed Au onomous and Connec ed Hyb id Vehicles
One o he cu en majo di ec ions o esea ch is in elligen
anspo a ion sys ems (ITS) whe e ci ies a e connec ed
h ough au onomous ehicles (AV) and wi eless echnology
ha helps in a ic low managemen , ehicula
communica ion and dec easing CO2 emissions o a oid he
‘u ban oad low conges ion’, educe ‘en i onmen al
pollu ion’ and ensu e ‘sa e y and ide com o ’ o he ehicles.
The le el o au onomy has been di ided in o six le els based
on Socie y o Au omo i e Enginee s (SAE J3016) s anda ds,
anging om Le el 0 o Le el 5 [1]. Acco ding o he su ey
[2], 69% o esponden s s a ed ha ully au oma ed d i ing will
each a 50% ma ke sha e be o e 2050. Howe e ,
unde s anding p e e ences and public esponse abou AVs
could esul in Sha ed Au onomous Vehicles (SAV) and
Connec ed Au onomous Vehicles (CAV) ha a e closely
linked o ITS. SAV and CAV a e bene icial when compa ed o
ha o AV [3]. I seems clea ha SAVs a e an in e es ing
al e na e mode o anspo ha eplaces p i a e ehicles [4].
Howe e , unce ain ies o u ilizing SAV and CAV wi h espec
o eal wo ld a ic made i a challenging aspec when
conside ing mixed a ic scena ios. Also, he powe ain
aspec is o be conside ed as a majo a ea o esea ch ha needs
o be ocused, SAV and CAV wi h hyb id elec ic mode ha e
a g ea po en ial in he deple ion o ailpipe emissions and
ene gy u iliza ion [5]. A p esen , ze o emission om ehicles
is possible o a limi ed ex en as comple e con e sion o
con en ional in e nal combus ion ehicle o elec ic ehicle
migh ake a lo o ime due o lack o cha ging in as uc u e,
ange o elec ic ehicle, cha ging ime and also ini ial cos o
he ehicle [6]. Due o he s ingen emission no ms and also
wi h he s ic egula ions made a change o look o al e na e
277
powe sou ces. By 2050, Eu opean Union aims o educe
g eenhouse gas emissions o ze o ha helps in de elopmen o
esea ch ac i i ies [7]. Hence, o educe he pe cen age o CO2
emissions hyb id elec ic au onomous ehicles (HEAV) play a
i al ole. This educes uel consump ion by 40-50% by
inc easing he e iciency [8].
1.2. Pla oon ehicles
Communica ion and collabo a ion o ehicles while in
a el is possible using he pla oon sys em [9, 10] which
imp o es ehicle sa e y by educing acciden s, imp o ing uel
e iciency and inc easing d i e com o . He e a g oup o
ehicles pla oon by ha ing a pla oon leade (PL) ehicle ha
collec s he in o ma ion o loca ion and oad da a ahead and
leads he pla oon membe s (PM) ehicles ha a e connec ed
behind o na iga e on oad. As a esul , la e al and longi udinal
mo ion con ol o PM ehicles could be main ained [11].
Howe e , in c i ical si ua ions d i e s should s ay ale o a oid
he isk o a ali ies while ehicles a e connec ed in pla oon
[12]. Hence, au onomous pla oon ehicles (APV) a e playing
a i al ole o elimina e he human e o s when a elling in a
mixed a ic low scena io. This is ope a ed wi h he help o
coope a i e adap i e c uise con ol (CACC) ha combines
adap i e c uise con ol (ACC) and ehicle o ehicle
communica ion [13]. APV helps no only o connec
comme cial ehicles, bu also helps o connec SAV by
exchanging he ehicle in o ma ion such as ajec o y
p edic ions, kinema ic s a us, senso s and came a da a. The
leng h o pla oon and numbe o pla oon ehicles a e e y
much c ucial in managing sa e a el in u ban and highways
scena ios.
1.3. V2V and V2X communica ion
To a oid a ali y, main ain e iciency and inc ease he
sa e y o SAV and CAV, ehicle o ehicle (V2V) and ehicle
o in as uc u e (V2X) communica ions a e necessa y. He e,
oad-side uni s (RSUs) a e deployed o e a ic sign boa ds,
a ic signals, gas s a ions, e c., which es ablishes a wi eless
communica ion sc ip [14]. Vehicula ne wo ks allow ehicles
moun ed wi h communica ion de ices such as on-boa d uni s
(OBUs) and o he senso s ha communica e using a dedica ed
sho - ange communica ion (DSRC) based on IEEE 802.11p
s anda d o wi h cellula V2X (CV2X) [15]. P ope ou ing
p o ocols de eloped o ehicula communica ion mus cope
wi h po en ially high speeds o he nodes (i.e., ehicles) as well
as wi h a iable ne wo k connec i i y ha ing ei he spa se and
dense scena ios h oughou ime. Also, hey should conside
di e en da a p io i ies, e.g., sending high p io i y da a o a oid
collisions, low p io i y da a abou a ic conges ion
in o ma ion o a ic managemen pu poses [16].
1.4. LTE/5G in Vehicles
Ad anced wi eless echnologies such as LTE and i h
gene a ion (5G) communica ion plays a i al ole in
ans o ming V2V and V2X communica ion sys ems. [17]
sugges s ha o ensu e he sa e y o he ehicle, decision
making and con olling need o be main ained. Implemen ing
5G echnology along wi h V2X could sus ain e o - ee
na iga ion and accu a e ajec o y ange han ha o 4G o
DSRC wi eless echnologies. Fo ins ance, applying he b akes
in a panic si ua ion using 4G migh be 1.5 m o a ehicle,
whe eas o 5G i would be a ound jus 2.5 cm. This p o ides
p omising pe o mance imp o emen o b aking wi h 5G
echnology compa ed o ha o 4G [18]. Also, one o he
impo an a eas o esea ch is hyb id communica ion using
LTE/5G combined wi h DSRC/802.11p. The ad an ages and
ea u es o a hyb id cellula and DSRC echnology is
conside ed as an al e na i e o V2X communica ion o ITS.
This has an abili y o la ge co e age, be e pe o mance, and
uni e sal deploymen when compa ed wi h only using LTE/5G
o IEEE 802.11p [19, 20]. One o ecen s udies es s o
ad anced LTE ha e been conduc ed o 50 kmph a di e en
load condi ions whe e ca ie s implemen ed 10MHz and
15MHz o analyse he channel quali y. The esul s e lec ed
ha ca ie a ained an a e age o 17 Mb/s and 30 Mb/s
espec i ely [21].
Fig.1. Typical a chi ec u e o HACV Pla ooning wi h Fede a ed
Lea ning.
1.5. Fede a ed Lea ning in AV
Wi h he inc ease in connec i i y and communica ion o
ehicles, huge amoun s o da a is being mig a ed om one
poin o ano he . Da a p i acy and secu i y a e majo a eas ha
need o be ocused o a oid he leakage o sensible in o ma ion
ha a e being sha ed. Mi iga ion o da a leakage is due o
inadequa e s o age and compu ing o single node poin s such
as ehicle o oad side uni s. Hence, ede a ed lea ning (FL)
plays a i al ole in managing p i acy o da a [22]. FL is one
o he machine lea ning (ML) app oaches whe e models a e
ained locally wi h da a a ailable in e nally. Hence, only
upda es om he model a e being sha ed and di ec
ansmission o da a will no be p esen , hus ensu ing p i acy
and educing la ency. The p ocess o FL can be execu ed in he
ollowing p ocess: (i) Selec ion o clien s; (ii) Dissemina ion
model; (iii) Dis ibu ed lea ning; (i ) Feedback om clien s;
( ) Agg ega ion; ( i) Tes ing model; ( ii) Model upda e.
A possible implemen a ion o FL in a ehicula ne wo k is
using ee-based lea ning me hod and con olu ion neu al
278
ne wo king (CNN), which elimina es he p esence o malicious
da a and a oids secu i y ailu es o maximum ex en [23].
Howe e , as he ehicula en i onmen is dynamic and
complex, collabo a ion be ween ehicles and RSU is a
challenging ask ha c ea es delay in esponse.
He e, we p esen he opics ha will be s udied in he i s
au ho s’ doc o al hesis. In sec ion II we highligh some o he
simula o s ha a e used o e alua e he pe o mance o he
p oposed wo k. Sec ion III summa izes highligh s o some
ela ed wo k on connec ed au onomous and hyb id ehicles,
ede a ed lea ning and pla ooning. Since we conside he
p esence o hyb id au onomous elec ic ehicles in ou
p oposals. Finally, sec ion IV depic s inno a i e ways in
collabo a ing he ehicles using pla ooning and e ec i e ways
in wi eless cha ging o ehicles while a elling, which will be
aken in o accoun in he design o ou FL amewo k. Fig.1.
ep esen s he basic a chi ec u e o he p oposed amewo k
pla ooning and wi eless cha ging o HACV based on FL.
2. SIMULATION FRAMEWORK
To a oid eal wo ld conges ion and collision wi h
pedes ians, ehicles and o he objec s, AVs and CAVs a e
es ed using mul iple simula o s ha p o ide be e accu acy
and eliable esul s. Nume ous simula o s such as SUMO,
OMNET++, VENTOS, Veins, Plexe, e c. a e oday a ailable.
A b ie o e iew abou main simula o s used o
expe imen a ion in ehicula communica ions, is p esen ed
below.
2.1. SUMO
Simula ion o U ban Mobili y (SUMO) [24] is an open
sou ce, mic oscopic a ic simula o so wa e and one o he
widely used so wa e o simula ing he ITS and ou e planning
o a pa icula loca ion. SUMO also ollows he physics and
dynamics o a ious ypes and quan i ies o ehicles [24].
Gene a ion o oad ne wo k is possible wi h ne con e ,
ne gene a e and ne edi whe eas a ic gene a ion is possible
wi h DUA ou e , JTR ou e OD2 ips, MA ou e , and
DF ou e . Once ehicle ou ing and a ic da a a e classi ied,
communica ion be ween ehicles and in as uc u e a e
classi ied using OMNET++.
2.2. OMNET++ and Veins
Objec i e Modula Ne wo k Tes bed in C++ (OMNET++)
[25] is a disc e e e en simula o wi h an in eg a ed
de elopmen en i onmen (IDE) based on C++ language ha
helps in building ne wo k simula ions. This consis s o a es ed
and p ede ined amewo k o de e mine and simula e a ious
ne wo ks, such as INET (TCP/IP s ack), Veins/MIXIM
(WAVE/DSRC s ack), con olle a ea ne wo k (CAN) and
SimuLTE (LTE cellula ne wo k). One among hem is Veins
[32], which in eg a ed wi h OMNET++ and SUMO o o m a
ull simula ion amewo k o ehicula ne wo ks. Veins
includes a oad and communica ion ne wo k ha helps o
simula e VANET applica ions o mobili y ehicles.
2.3. VENTOS AND PLEXE
Vehicula Ne wo k Open Simula o (VENTOS) [26] is one
o collabo a i e a ic low analysis so wa e ha allows
esea che s o implemen algo i hms and p o ocols abou
DSRC. I is an open-sou ce simula o ha helps he V2X
communi y o wo k wi h anspo a ion, ehicula ne wo k and
con ol heo y esea ch ields. VENTOS a chi ec u e has
addi ional ea u es o coope a i e adap i e c uise con ol
(CACC) in SUMO h ough TRACI, ha helps in pla ooning o
he ehicles. VENTOS also suppo s eins o V2X
communica ion wi h IEEE 802.11p [27]. Ano he open-sou ce
simula o ha is simila o VENTOS is PLEXE [28] which is
mos ly used o Pla ooning o CACC o he ehicles. As Plexe
is an ex ension o eins amewo k, his helps in simula ing
ealis ic en i onmen s, main aining he physics and dynamics
o he ehicles along wi h an in e - ehicle communica ion
p o ocol s ack.
3. RELATED WORK
The ollowing pape s add ess key issues o he esea ch
a ea ha need o be ocused on: The wo k [5] implemen ed
longi udinal au onomous d i ing o hyb id elec ic ehicles.
Along wi h access o V2X/V2V is made simple using game
heo y app oach o longi udinal au onomous d i ing con ol
amewo k. This helps in achie ing mul iple objec i es by
coo dina ing in e nal combus ion engine, elec ic mo o and
ehicle b aking. The simula ion esul s om he expe imen s
sugges educ ion in uel consump ion, imp o ing d i ing
com o and ca - ollowing abili y. The au ho s o [29]
de eloped a V2V communica ion ne wo k o imp o e a uck
pla ooning ha enables sha ing o da a, e.g., speed o ehicle.
He e, wo an ennas we e placed a he on and ea o he PL
ehicle ha p o ides good esul s in a eal scena io. In [30] he
au ho s conside ed an ad anced ehicula ne wo king called
Fede a ed Vehicula Ne wo king ha consis s o DSRC and
mmWa e communica ion o s able and scalable pe o mance
and suppo s machine lea ning (ML) and FL. Also, hey ha e
es ed se e al ou ing algo i hms using Ne wo kX, a g aph-
based simula o . The au ho s in [31] p oposed a new FL
algo i hm o designing a CAV con olle o p edic and
pe o m eal ime decisions. The au ho s p oposed a no el
dynamic ede a ed p oximal (DFP) algo i hm o FL aining.
Simula ion esul s show ha he con olle helps in iden i ying
he accu a e change o e speed in di e en a ic scena ios
and also, helps in de eloping wi eless connec i i y o CAVs.
4. ONGOING WORK
To imp o e he ene gy e iciency o ehicles, main aining
he sa e y o ehicula da a and educing he CO2 emission a e
he main a eas ha a e being ocused on his hesis. The s a e
o a (SOA) echnology ha is being implemen ed in he
ongoing wo k is ollowed in ou s ages: (i) Fi s s age is o
design and implemen a simula ion amewo k o hyb id
elec ic ehicle echnology wi h sha ed connec ed au onomous
ehicles ha helps in main aining he change in powe ain
om engine o elec ic, based on he low o a ic, ype o
oad ne wo k such as u ban a ea and highway. (ii) In he second
s age, communica ion be ween ehicles is made o a clus e
o ehicles o ming a pla oon. Implemen ing HEAV
echnology wi h CACC o pla ooning helps in educing uel
consump ion while ehicles a e in pla oon. This also helps in
imp o ing he s a e o cha ge (SOC) o he ba e y ha helps
279
in cha ging ba e ies while ehicles a e connec ed in pla oons.
(iii) The hi d s age o ou ongoing wo k is o u ilize 5G o
ad anced LTE combined wi h IEEE 802.11p/DSRC
echnologies ha can sa egua d he ehicles om collision and
main ain high speeds and p ope communica ion be ween
ehicles wi hou any dis u bance. (i ) Finally, o sus ain he
p i acy o da a ha a e being sha ed be ween ehicles and
RSU, ad anced machine lea ning a chi ec u es such as FL will
be employed.
5. ACKNOWLEDGEMENTS
This wo k was suppo ed by he Spanish Go e nmen
unde esea ch p ojec “Enhancing Communica ion P o ocols
wi h Machine Lea ning while P o ec ing Sensi i e Da a
(COMPROMISE)” PID2020-113795RB-
C31/AEI/10.13039/501100011033.
REFERENCES
[1] “Taxonomy and De ini ions o Te ms Rela ed o D i ing Au oma ion
Sys ems o On Road Mo o Vehicles”, SAE S anda d J3016, 2018.
[2] M. Ky iakidis, R. Happee, J.C. de Win e , “Public opinion on au oma ed
d i ing: esul s o an in e na ional ques ionnai e among 5000
esponden s”, T anspo Res. F: T a ic Psychol. Beha ., ol. 32, pp.
127-140, 2015.
[3] Daniela Paddeu, Ioannis Tsou os, G aham Pa khu s , Amalia
Polydo opoulou, Ian She gold, “A s udy o use s p e e ences a e a
b ie exposu e in a Sha ed Au onomous Vehicle (SAV)”,
T anspo a ion Resea ch P ocedia, ol. 52, pp. 533-540, 2021,
h ps://doi.o g/10.1016/j. p o.2021.01.063.
[4] Tony Z. Zhang, T. Donna Chen, “Sma cha ging managemen o sha ed
au onomous elec ic ehicle lee s: A Puge Sound case s udy”,
T anspo a ion Resea ch Pa D: T anspo and En i onmen , ol. 78,
2020, h ps://doi.o g/10.1016/j. d.2019.11.013.
[5] Shuo Cheng, Liang Li, Xiang Chen, Sheng-nan Fang, Xiang-yu Wang,
Xiu-heng Wu, Wei-bing Li, “Longi udinal au onomous d i ing based
on game heo y o in elligen hyb id elec ic ehicles wi h
connec i i y”, Applied Ene gy, ol. 268, 2020,
h ps://doi.o g/10.1016/j.apene gy.2020.115030.
[6] Shaik Amjad, R. Rud amoo hy, P. Sadagopan, S. Neelak ishnan,
"Implemen a ion and e alua ion o change-o e speed in plug-in hyb id
elec ic wo wheele ", Ene gy, ol. 109, 2016, pp. 858-865,
h ps://doi.o g/10.1016/j.ene gy.2016.05.025.
[7] Union, E., 2021. “EU Clima e Ac ion and he Eu opean G een Deal”,
Accessed 11-02-2021 2021.
h ps://ec.eu opa.eu/clima/policies/s a egies/2050_en.
[8] Fa ooq She , Sihong Chen, Ali Raza, Tahi Rasheed, Omid Razmkhah,
Tazien Rashid, Piyya M. Ra i-ul-Shan, Begum E en, "No el s a egies
o educe engine emissions and imp o e ene gy e iciency in hyb id
ehicles", Cleane Enginee ing and Technology, ol. 2, 2021,
h ps://doi.o g/10.1016/j.cle .2021.100074.
[9] Ca l Be genheml, Hen ik Pe e sson, E ik Coelingh, C is o e Englund,
S e en Shlado e , Sadayuki Tsugawa, "O e iew o pla ooning
sys ems", 19 h ITS Wo ld Cong ess, Vienna, Aus ia, 22/26 Oc obe
2012.
[10] Heikoop, D., de Win e , J., an A em, B., S an on, N. A., "E ec s o
pla ooning on signal-de ec ion pe o mance, wo kload, and s ess: A
d i ing simula o s udy", Applied E gonomics: human ac o s in
echnology and socie y, ol. 60, pp. 116-127, 2017,
h ps://doi.o g/10.1016/j.ape go.2016.10.016.
[11] A. Sa ke , C. Qiu, H. Shen, "Quick and au onomous pla oon main enance
in ehicle dynamics o dis ibu ed ehicle pla oon ne wo ks", in: P oc.
o 2nd In e na ional Con e ence on In e ne -o -Things Design and
Implemen a ion, pp. 203–208, 2017.
[12] Banks, V.A., S an on, N.A., Ha ey, C., "Sub-sys ems on he oad o
ehicle au oma ion: hands and ee ee bu no ‘mind’ ee d i ing",
Sa e y Science, ol. 62, pp. 505-514, 2014,
h p://dx.doi.o g/10.1016/j.ssci.2013.10.014.
[13] P.K. Singh, G.S. Tabjul, M. Im an, S.K. Nandi, S. Nandi, "Impac o
secu i y a acks on coope a i e d i ing use case: CACC pla ooning",
in: P oc. o TENCON, pp. 0138–0143, 2018.
[14] Hong Zhang, Xinxin Lu, "Vehicle communica ion ne wo k in in elligen
anspo a ion sys ems based on In e ne o Things", Compu e
Communica ions, ol. 160, pp. 799–806, 2020.
[15] Michael Lee, T a is A kison, "VANET applica ions: Pas , p esen , and
u u e", Vehicula Communica ions, ol. 28, 2021,
h ps://doi.o g/10.1016/j. ehcom.2020.100310.
[16] Dahiya, A., Chauhan, R., "A compa a i e s udy o MANET and VANET
en i onmen ", Jou nal o compu ing, ol. 2, no. 7, pp. 87-92, 2010.
[17] Y. Shi, Q. Han, W. Shen, X. Wang, "A Mul i-Laye Collabo a ion
F amewo k o Indus ial Pa ks wi h 5G Vehicle- o-E e y hing
Ne wo ks", Enginee ing, 2021,
h ps://doi.o g/10.1016/j.eng.2020.12.021
[18] Gue a a L, Aua Cheein F. "The ole o 5G echnologies: challenges in
sma ci ies and in elligen anspo a ion sys ems", Sus ainabili y,
12(16):6469, 2020, h ps://doi.o g/10.3390/su12166469.
[19] G. A ani i, C. Campolo, M. Condoluci, A. Ie a and A. Molina o, "LTE
o ehicula ne wo king: a su ey", in IEEE Communica ions
Magazine, ol. 51, No. 5, pp. 148-157, May 2013, doi:
10.1109/MCOM.2013.6515060.
[20] A. Bu bano-Ab il, R.M. Edwa ds, V. Rangel-Licea, R. Aquino-San os,
M. Lopez-Gue e o, R.S. Kalawsky, M. Behja i, "Modeling and
analysis o LTE connec i i y in a high mobili y ehicula
en i onmen ", Compu e s & Elec ical Enginee ing, ol. 68, pp. 322-
336, 2018, h ps://doi.o g/10.1016/j.compeleceng.2018.04.007.
[21] S. A allone, N. Pasquino, S. Zinno and D. Casillo, "Sma phone-based
measu emen s o LTE ne wo k pe o mance", 2017 IEEE In e na ional
Ins umen a ion and Measu emen Technology Con e ence (I2MTC),
pp. 1-6, 2017, doi: 10.1109/I2MTC.2017.7969854.
[22] Jakub Konečný, H. B endan McMahan, Felix X. Yu, Pe e Rich á ik,
Ananda Thee ha Su esh, Da e Bacon, "Fede a ed Lea ning: S a egies
o Imp o ing Communica ion E iciency", a Xi :1610.05492, 2016.
[23] Y. Lu, X. Huang, Y. Dai, S. Maha jan and Y. Zhang, "Fede a ed Lea ning
o Da a P i acy P ese a ion in Vehicula Cybe -Physical Sys ems",
IEEE Ne wo k, ol. 34, no. 3, pp. 50-56, May/June 2020, doi:
10.1109/MNET.011.1900317.
[24] Daniel K ajzewicz, 2010, "T a ic Simula ion wi h SUMO – Simula ion
o U ban Mobili y", In e na ional Se ies in Ope a ions Resea ch &
Managemen Science, Fundamen als o T a ic Simula ion, chap e 0,
pp. 269-293, Sp inge , DOI: 10.1007/978-1-4419-6142-6_7.
[25] C. Somme , R. Ge man and F. D essle , "Bidi ec ionally Coupled
Ne wo k and Road T a ic Simula ion o Imp o ed IVC Analysis",
IEEE T ansac ions on Mobile Compu ing, ol. 10, no. 1, pp. 3-15, Jan.
2011, doi: 10.1109/TMC.2010.133.
[26] Amoozadeh, M., “Vehicula Ne wo k Open Simula o (VENTOS)”,
h p://maniam.gi hub.io/VENTOS.
[27] Mani Amoozadeh, B yan Ching, Chen-Nee Chuah, Dipak Ghosal, H.
Michael Zhang, “VENTOS: Vehicula Ne wo k Open Simula o wi h
Ha dwa e-in- he-Loop Suppo ", P ocedia Compu e Science, ol. 151,
2019, pp. 61-68, h ps://doi.o g/10.1016/j.p ocs.2019.04.012.
[28] M. Sega a, S. Joe e , B. Bloessl, C. Somme , F. D essle and R. L. Cigno,
"Plexe: A pla ooning ex ension o Veins", 2014 IEEE Vehicula
Ne wo king Con e ence (VNC), pp. 53-60, 2014, doi:
10.1109/VNC.2014.7013309.
[29] Ca l Be genhem, E ik Hedin, Daniel Ska in, "Vehicle- o-Vehicle
Communica ion o a Pla ooning Sys em", P ocedia - Social and
Beha io al Sciences, ol. 48, pp. 1222 – 1233, 2012.
[30] J. Posne , L. Tseng, M. Aloqaily and Y. Ja a weh, "Fede a ed Lea ning in
Vehicula Ne wo ks: Oppo uni ies and Solu ions", IEEE Ne wo k, ol.
35, no. 2, pp. 152-159, 2021, doi: 10.1109/MNET.011.2000430.
[31] Tengchan Zeng, Mingzhe Chen, Mehdi Bennis, "Fede a ed Lea ning on
he Road: Au onomous Con olle Design o Connec ed and
Au onomous Vehicles", a Xi :2102.03401, 5 Feb 2021.
[32] I ikha Ahmad, Ra idah Md Noo , Ismail Ahmedy, Syed Adeel Ali Shah,
Ib a Yaqoob, Ejaz Ahmed, Muhammad Im an, "VANET–LTE based
he e ogeneous ehicula clus e ing o d i ing assis ance and ou e
planning applica ions", Compu e Ne wo ks, ol. 145, pp. 128-140,
2018, h ps://doi.o g/10.1016/j.comne .2018.08.018.
280
