Enhancing Vulnerable Road User Awareness of Intelligent Transport

Systems Through Relay and Aggregation of Collective Perception

Messages with Road Side Units

Vincent Albert Wolff

Institute of Communication Technologies, Leibniz University Hannover, Appelstr. 9a, Hannover, Germany

Keywords:

Collective Perception, Collaborative Driving, V2X, V2I, Road Side Unit, Vulnerable Road User Awareness,

Vehicular Adhoc Networks, Road Safety.

Abstract:

The Collective Perception Service (CPS) allows connected vehicles to gain a more comprehensive picture of

their environment by sharing information about the dynamic state of objects with other vehicles and infras-

tructure. Objects detected by on-board sensors are shared through Vehicle-to-Vehicle (V2V) or Vehicle-to-

Infrastructure (V2I) communication. However, the range of V2V communication is limited, and Road Side

Units (RSUs) can be deployed to enhance the range and attenuate the negative effects of V2V signal propaga-

tion. We enhance the vehicular network by RSUs to aggregate and forward Collective Perception Messages

(CPMs) received from neighboring vehicles, thus improving the overall environmental perception and the per-

ception of Vulnerable Road Users (VRUs) in particular. Our simulation results, based on the ETSI ITS-G5

standard, demonstrate the effectiveness of the CPS in an urban intersection scenario, showing the positive

impact of additional V2I communication and the deployment of RSUs on vehicular perception of VRUs. The

addition of RSUs results in a signiﬁcant improvement in VRU perception, while packet loss on the network

channel increases moderately.

1 INTRODUCTION

Recent advancements in Advanced driver-assistance

systems (ADAS) and autonomous driving have led to

the development of a variety of services for Intelligent

Transportation System Stations (ITS-Ss), including

collision detection, emergency braking, lane-keeping

assistance, and many others. These services require

accurate and reliable information from ITS-Ss, which

must provide high-precision local sensor data and

a comprehensive Local Environment Model (LEM).

The LEM of each vehicle is updated by the on-board

sensor detections, as modern ITS-Ss are equipped

with Lidar, radar or camera sensors. Vehicle-to-

Everything (V2X) communication provides services

to enhance the LEM by receiving messages from

other stations about the position and dynamics of

one vehicle (Cooperative Awareness Message, CAM)

and by sharing the LEM with other ITS-Ss. This is

provided by the Collective Perception Service (CPS)

proposed by the European Telecommunications Stan-

dards Institute (ETSI).

https://orcid.org/0000-0003-4210-3035

The Collective Perception Service does not need

to rely on Vehicle-to-Vehicle (V2V) links exclusively.

The V2V communication can be assisted by Vehicle-

to-Infrastructure communication, which enables mes-

sage exchange between Road Side Units (RSUs) and

connected vehicles. The CPS has lots of potential to

decrease the risk of collisions, as it provides ITS-S

with data about other trafﬁc participants, objects and

obstacles in its surrounding. Especially Vulnerable

Road Users (VRUs), i.e. pedestrians and cyclists,

are more prone to critical accidents with high fatal-

ity rate (den Berghe, 2021). VRUs are not equipped

with sensors and have no communication capabili-

ties, making them all the more reliant on the V2X

communication, in particular Collective Perception.

The ETSI CPS (ETSI, 2019b) tackles this commu-

nication gap and provides the ITS-S with informa-

tion about current VRU dynamics and position which

is gathered by the on-board sensors of each station.

In current research the CPS is mostly investigated in

terms of its capabilities to improve the environmen-

tal awareness ratio (EAR) of vehicles and safety re-

lated metrics (Schiegg et al., 2021). While Collective

Perception has been shown to improve vehicle detec-

Wolff, V.

Enhancing Vulnerable Road User Awareness of Intelligent Transport Systems Through Relay and Aggregation of Collective Perception Messages with Road Side Units.

DOI: 10.5220/0011973300003479

In Proceedings of the 9th International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2023), pages 337-343

ISBN: 978-989-758-652-1; ISSN: 2184-495X

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

337

tion and increase safety by reducing blind spots, re-

lated work often overlooks the perception of VRUs.

In this work the focus is on enhancing the percep-

tion of VRUs, which is crucial for ensuring road traf-

ﬁc safety. As VRU safety is most critical in urban

environments, a scenario for an urban intersection is

created. In addition, the impact of RSUs is exam-

ined by placing an RSU at the intersection center.

The RSU helps to compensate urban-speciﬁc commu-

nication challenges such as multi-path signal prop-

agation, diffraction and shadowing due to building

and other obstacles. Maintaining a reliable commu-

nication link even with Non-Line-Of-Sight (NLOS)

communication has big impact on the CPS: Research

found that NLOS links signiﬁcantly reduce communi-

cation range and signal strength not only implement-

ing 802.11p on physical layer, but also for the upcom-

ing 802.11bd and 5G NR-V2X. (Anwar et al., 2019).

In this study, we simulate the Cooperative Per-

ception System (CPS) using the ETSI ITS-G5 (ETSI,

2020) and CPM (ETSI, 2019b) standards in an urban

setting. The simulated scenario incorporates typical

signal propagation characteristics at intersections and

varying densities of vehicles and VRUs. The service

performance is examined by the vehicular environ-

mental awareness of VRUs and network related met-

rics.

2 RELATED WORK

2.1 ETSI Collective Perception

The ETSI standardizes CPM and proposes generation

rules which deﬁne which objects and with which fre-

quency these objects will be added to a CPM. The

dynamic-based generation follow these rules:

1. The Object is detected for the ﬁrst time

2. Object has moved more than 4 meters since last

inclusion

3. Object velocity changed more than 0.5m/s since

last inclusion

4. Object heading angle has changed more than 4 de-

grees since last inclusion

5. Object is a VRU and was not included in a mes-

sage in the last 500ms

After the generation rules are applied, redundancy

mitigation techniques can be deployed to ﬁlter re-

dundant objects. There are frequency-based, entropy-

based, dynamic-based and conﬁdence-based rules de-

ﬁned, which combine object information of local de-

tection and received V2X-detection to determine the

priority to send one object (ETSI, 2019b). Redun-

dancy mitigation rules were studied by (Delooz et al.,

2022) in terms of environmental awareness, channel

load and information redundancy. They found that no

single redundancy mitigation technique reaches all re-

quirements and propose a ﬁltering algorithm to com-

bine different mitigation techniques.

2.2 Infrastructure-Assisted CPM

To enhance object detection reliability and range,

RSUs can be integrated in an VANet to assist the V2V

links. (Yu et al., 2022) provide an overview of related

studies on infrastructure-assisted vehicular services.

They focus on object perception and detection using

different type of on-board sensors (camera, Lidar and

a fused system).

There are two different operating options of RSUs

related to the CPS: RSUs can either detect objects

with local sensors and create CPMs from local detec-

tions or they relay received CPMs by other ITS-Ss.

The impact of RSUs on Collective Perception has

been studied in several forms: (Chtourou et al., 2021)

studied a RSU implementation with 802.11p in the

VEINS simulator. They conclude that RSU-assisted

CPS provides higher efﬁciency compared to CPS

based on V2V communication. (Pacella et al., 2021)

used a pure Road Side Unit Scenario, where the RSU

is equipped with sensors to broadcast CPMs. Their

focus is on latency measurements, where they esti-

mate an end-to-end delay of less than 250ms. (Gar-

lichs et al., 2020) implemented a relay mechanism to

aggregate CAMs received by vehicles and broadcast

the aggregated information in CPMs. They studied

that the driver reaction time can be improved while

the additional network load remains moderate.

2.3 Vulnerable Road User Safety

The safety of Vulnerable Road Users (VRUs), such

as pedestrians and cyclists, plays a critical role in the

concept of Collective Perception within transporta-

tion systems. As these individuals are typically not

equipped with Vehicle-to-Everything (V2X) commu-

nication modules, they must rely on vehicular safety

systems for protection. One approach to address this

issue is to integrate VRUs into a Vehicular Ad-hoc

Network (VANet) by allowing them to broadcast their

own messages regarding their location and move-

ments. The European Telecommunications Standards

Institute (ETSI) has standardized the Vulnerable Road

User Awareness Message (VAM) to facilitate commu-

nication between VRUs and vehicles (ETSI, 2019a).

However, it is important to note that the widespread

VEHITS 2023 - 9th International Conference on Vehicle Technology and Intelligent Transport Systems

338

Object #1

Object #2

Object #3

CPM

...

Object #1

Object #4

Object #5

CPM

...

Object #2

Object #6

CPM

...

CPM

...

Object #1

Object #2

Object #3

Object #4

Object #5

Object #6

RSU

Figure 1: Aggregation of CPMs in the RSU.

adoption of this technology cannot be guaranteed, and

the accuracy and reliability of VRU communication

devices, such as smartphones, must be established be-

fore they can be effectively utilized.

(Willecke et al., 2021) simulate ETSI Collective

Perception using the Artery framework and show that

the CPS can signiﬁcantly improve vehicular aware-

ness of VRUs while maintaining a stable network

load. Also alternative parameterizations for CPS are

analyzed to indicate different tracking accuracies for

person objects. (Lobo et al., 2022) study VAM and

a combination of CPM and VAM, which outperforms

implementing one standalone service.

3 CONTRIBUTION

Although urban scenarios have already been studied

in regards to CPM, there is a lack in research of the

combination of VRUs and RSUs in characteristic ur-

ban environments. As current research of RSUs sug-

gest a placement at intersections centers (Wang et al.,

2017), it is worth to study the impact of RSUs on

VRU awareness. As stated in 2.2 related work found

that the relay mechanism can be beneﬁcial imple-

menting the aggregation of Cooperative Awareness

(CA) messages, which indicates that the relay and ag-

gregation of CPMs can be a next step to be deployed.

Our goal is to demonstrate the impact of a re-

lay and aggregation system on the perception ratio of

VRUs. Most related work focuses on general object

perception and seldom takes into account the unique

safety concerns of VRUs. Compared to ITS-equipped

vehicles, VRUs lack the ability to broadcast their

location and current movement via the CA service,

which makes them more dependent on the CPS.

We also investigate the impact of the additional

RSU message dissemination on the overall network

channel load, as the CPS bandwidth is limited and

competes with other V2X technologies. In addi-

tion, different vehicle densities and market penetra-

tion rates are tested to analyze the impact of increas-

ing channel congestion. The simulation is conducted

using the ETSI ITS-G5 communication model and

ETSI CPM generation rules according to 2.1, provid-

ing a realistic communication environment and ensur-

ing comparability with other studies.

This simulation study examines the combination

of V2I and V2V links in order to fully evaluate the

impact of Road Side Units (RSUs) on the perception

of the environment. To provide a comprehensive anal-

ysis, we compare the results of the implementation

with both V2I and V2V links enabled to those ob-

tained using pure V2V communication without RSUs.

4 IMPLEMENTATION

The proposed implementation utilizes an RSU relay

and aggregation mechanism, which is described in

the following section. The urban scenario poses sig-

niﬁcant challenges for signal propagation, hence the

GemV2 model (Boban et al., 2014), which supports

different types of signal links, is used. Additionally,

the characteristics of the chosen road scenario and

trafﬁc simulation will be discussed.

4.1 CPM Aggregation and Relay

Each car generates its own CPMs based on dynamic

ETSI rules described in 2.1, which are broadcasted

to ITS-Ss in communication range. When the CPMs

are received by the RSU, they are aggregated to build

newly generated CPMs, ideally containing a greater

number of objects in the PerceivedOb jectContainer.

The RSU relays the messages with a ﬁxed frequency

of 100ms. Figure 1 shows an example of the aggrega-

tion, where the RSU receives CPMs by 3 cars and ag-

gregates the including objects to one new CPM. The

LEM of the RSU is created out of the LEMs of the

vehicles in the VaNET according to equation 1.

LEM

RSU

∑

i=1

LEM

Vehicle(i)

(1)

In our RSU service implementation, VRUs are

given the highest priority when generating and aggre-

gating CPMs. After the type of object, the next pri-

Enhancing Vulnerable Road User Awareness of Intelligent Transport Systems Through Relay and Aggregation of Collective Perception

Messages with Road Side Units

339

ority is the age of the information. This ensures that

the most recent sensor measurements are prioritized

when generating and aggregating CPMs. The infor-

mation shared among connected vehicles is as up-to-

date as possible.

The use of Road Side Units (RSUs) in this imple-

mentation allows for enhanced range and improved

signal propagation of the CPMs, which can improve

the overall perception and safety of connected vehi-

cles in urban environments.

4.2 Signal Propagation Modelling

Simulating V2X communication requires a signal

propagation model to implement realistic communi-

cation characteristics. For the simulation we choose

the GemV2 geometry-based model for simulating

Vehicle-To-Everything (V2X) communication. It

considers three types of links: line-of-sight (LOS)

links, non-line-of-sight (NLOS) links due to vehicles,

and NLOS links due to static objects.

In GemV2, the LOS links are modeled using a

two-ray propagation model. This model considers the

direct path between the transmitter and receiver, as

well as a reﬂected path off of a reﬂecting object, such

as a building or vehicle. The reﬂected path can affect

the signal strength and the performance of the V2V

communication system.

Furthermore GemV2 considers NLOS links due to

vehicles and static objects. In these cases, the signal

must pass through or around the obstructing objects,

which can cause attenuation, delay, and other impair-

ments. GemV2 takes into account the geometry of the

environment and the objects within it to accurately

model the NLOS links and their effects on the V2V

communication system. (Boban et al., 2014)

Equation 1 deﬁnes the well-known two-ray

ground reﬂection model, which models the V2X sig-

nal propagation. (Sommer and Dressler, 2011)

TOT

| =

LOS

cos(ω

(t −

LOS

))

ground

cos(ω

(t −

ground

))

(1)

For the pass loss PL, the following equation ap-

plies:

PL(d) = PL(d

) + 10 γ log

(

) (2)

The factor γ is set differently for LOS, NLOSb

(building) and NLOSv (vehicle) links to model the

signal propagation according to different types of ob-

stacles. GemV2 has the ability to efﬁciently scale to

Figure 2: SUMO scenario of the simulation.

high-density scenarios, where there are many vehicles

and obstacles that can affect the V2V communication.

GemV2 uses geometric calculations and approxima-

tions to quickly and accurately simulate V2V commu-

nication in these scenarios.

5 SIMULATION

The simulation is built to evaluate the performance

of a vehicle-to-everything (V2X) communication sys-

tem in accordance with the ETSI ITS-G5 (ETSI,

2020) standard. The simulation environment is con-

structed using the open-source framework Artery

(Riebl et al., 2015). The trafﬁc simulation is imple-

mented in a scenario created using the Simulation of

Urban Mobility (SUMO) (Krajzewicz et al., 2002).

A two-lane road in each direction is simulated, with

an intersection arm length of 150 meters. Two dif-

ferent density modes are evaluated, one with a den-

Table 1: Simulation parameters.

Parameter Value

Enabled services ETSI CA, ETSI CPS

Physical layer IEEE 802.11p

Bit rate 6Mbit/s

Carrier frequency 5.9GHz

Bandwidth 10MHz

Channel G5-CCH

Transmission power 200 mW

Signal threshold -85 dBm

Noise threshold -65 dBm

V2X propagation model GEMV2

Penetration rate 100%

DCC Disabled

RSU max antenna gain 10dB

RSU antenna type Constant gain antenna

Vehicle sensor set 85m 360°camera

VEHITS 2023 - 9th International Conference on Vehicle Technology and Intelligent Transport Systems

340

Figure 3: EAR of VRUs for low density scenario

sity of 75 vehicles per kilometer and the other with a

density of 150 vehicles per kilometer. The number of

VRUs is set proportional to the vehicular density. The

VRUs consist of cyclists and pedestrians, that share

the sidewalk. The scenario implemented in SUMO

is illustrated in Figure 2. Obstacles (grey areas) are

placed on each intersection arm to simulate the im-

pact of buildings at intersections. The evaluation is

conducted at the intersection center, and the impact of

network load and sensor detections by the outer junc-

tions is mitigated by ﬁltering the data to only consider

vehicles in the area of interest. The RSU is placed at

the intersection center.

Simulation parameters are set according to table

1. The physical layer of the system is based on the

IEEE 802.11p standard, which speciﬁes the use of

the 5.9 GHz frequency band for wireless communi-

cation in intelligent transportation systems. The chan-

nel used in the simulation has a bandwidth of 10MHz.

The transmission power is set to 200 mW. The signal

and noise thresholds are -85 dBm and -65 dBm, re-

spectively, which represent the reconstruction thresh-

olds of the signal from noise. The V2X propagation

model used in this simulation is GemV2, described

in 4.2. The penetration rate, which is the proportion

of vehicles equipped with V2X technology, is set to

a range of values: 5%, 10%, 25%, 50%, and 100%.

The Decentralized Congestion Control (DCC) is dis-

abled, which means that the system does not control

the ﬂow of data on Facilities or Access layer. The

type of antenna used is a constant gain antenna (omni-

directional), thus the gain of the antenna does not vary

with frequency or direction. In the simulation ve-

hicles are equipped with 360-degree camera sensors

with a range of 85m as on-board sensors, while the

RSU does not have any environmental sensors.

As ITS-Ss can be equipped with multiple services,

besides the CPS the CA service is enabled.

Figure 4: EAR of VRUs for high density scenario.

6 RESULTS

For the evaluation of our simulation, perception and

network related metrics are investigated. We adapt the

Environmental Awareness Ratio (EAR) metric from

(Schiegg et al., 2019), which was originally intro-

duced for objects of any type. Figure 3 and 4 shows

the EAR of the vehicles in respect to VRUs accord-

ing to equation 3. It is deﬁned as the ration between

the sum of detected VRUs within a speciﬁc range and

the total number of VRUs in the range. The range of

interest is set to 200m in the simulation. The tracking

age of information is set to a maximum of 1 second.

Tracks with older updates are not considered.

EAR(V RU s) =

∑

i=1

V RU

Detected,i

∑

i=1

V RU

Range,i

(3)

Figure 3 and 4 shows the EAR for low- and high-

density scenarios, respectively. In both cases, it is

clearly demonstrated that the EAR increases as the

penetration rate of equipped vehicles increases. This

outcome is expected as the number of vehicles capa-

ble of sharing locally-sensed objects increases. The

impact of the RSU on the EAR is signiﬁcant. For ex-

ample, with no RSU and a 5% penetration rate, the

mean EAR jumps from 0.26 to 0.62. This trend is

consistent for higher penetration rates, though the rel-

ative increase becomes smaller. In the low-density

scenario, a median of 100% EAR is achieved at a pen-

etration rate of 50%. In contrast, for the high-density

scenario, the median is only reached at a penetration

rate of 100% with the RSU enabled. Another impor-

tant observation is that when the RSU is turned on, the

lower quartile value is signiﬁcantly improved, provid-

ing a higher minimum coverage of detected objects.

It is also noteworthy that the overall EAR is lower for

the high-density scenario. This could be due to the

fact that while one vehicle may detect more VRUs

Enhancing Vulnerable Road User Awareness of Intelligent Transport Systems Through Relay and Aggregation of Collective Perception

Messages with Road Side Units

341

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

PDR

0.0

0.2

0.4

0.6

0.8

1.0

CDF

Penetration Rate

25 %, No RSU

25%, RSU

100%, No RSU

100 %, RSU

Figure 5: Cumulative Distribution Function of PDR for low

and high penetration rate, RSU turned on and off.

in the high-density case, the total number of VRUs

is also increased. Another possible reason is that with

an increasing number of objects, the obstruction of the

ﬁeld of view sensor increases. Other reasons could be

network-related and warrant further investigation.

It is observed that the VRU awareness does not

reach 100% without the RSU enabled. At ﬁrst glance

this is not a result to be expected as previous stud-

ies found that the EAR enhanced by the CPS covers

100% of detected objects even with penetration rates

of 25% and beyond (Schiegg et al., 2021). However

this work focus on the perception of VRUs. Com-

pared to other type of trafﬁc participants, they are of-

ten hidden by the vehicular on-board sensors due to

their small surface. Furthermore, no difference has

been made between VRUs on the sidewalk or VRUs

crossing the street. Therefore a number of undetected

VRUs could be on the sidewalk. However it is worth

to consider all VRUs in the EAR: Pedestrians or cy-

clists could intend to cross the street or behave unex-

pectedly. Another aspect that decreases the EAR is

the inclusion of obstacles in the simulation by plac-

ing buildings at all intersection arms: The on-board

360 degree sensor can be blocked and the commu-

nication range is attenuated due to the NLOS signal

propagation. The evaluation of network-related pa-

rameters is illustrated in ﬁgure 5, which shows the

Packet Delivery Ratio (PDR) for a low (25%) and

high (100%) penetration rate, represented as a Cumu-

lative Distribution Function (CDF). In the graph, the

vehicle density is ﬁxed at the low density scenario to

simplify the analysis by reducing the number of pa-

rameters. This type of graph allows for a clear vi-

sualization of the distribution of PDR and the mini-

mum PDR (CDF(x) → 0). The graph compares the

PDR with and without the enabled RSU. It can be ob-

served that, for the same penetration rate, there is a

signiﬁcant increase in network load with the RSU en-

abled, resulting in a lower PDR: The minimum PDR

decreases from 0.75 to 0.39 (25% penetration rate)

and from 0.57 to 0.29 (100% penetration rate), re-

spectively. The 50th percentile drops from 0.89 (25%

penetration rate) to 0.56 and from 0.76 to 0.42 (100%

penetration rate), respectively. Additionally, it is evi-

dent that the variance of the distribution increases as

the PDR decreases, indicating a less reliable system

overall. The observations of PDR also provide insight

into another potential reason for the decreased EAR

in the high-density scenario. As the PDR decreases,

the communication channel becomes increasingly sat-

urated, resulting in a greater number of packet losses.

In an over-congested channel, not all information will

be shared between the ITS-Ss, leading to a reduction

in the EAR.

7 CONCLUSION

The results of the simulation according to ETSI stan-

dards show that an RSU at urban intersections can im-

prove the vehicular perception of VRUs by enabling

the aggregation and dissemination of CPMs to ITS-

Ss. However, it should be noted that this implemen-

tation also incurs a cost in the form of increased net-

work congestion, leading to a decrease in the Packet

Delivery Ratio (PDR) on the communication channel.

Despite this decrease in PDR, the results indicate that

the overall improvement in Environmental Awareness

Ratio (EAR) when the RSU is activated outweighs

the decrease in PDR, resulting in a net beneﬁt from

the utilization of the RSU. It is also revealed that im-

provement of VRU awareness is a challenging task in

highly-congested urban scenarios: High CPS penetra-

tion rates are needed to achieve proper EAR. As the

results are promising, in future work additional sce-

narios can be investigated. Special RSU CPM gen-

eration rules could be introduced to enhance the net

beneﬁt of an RSU at urban intersections, leading to a

more controlled dissemination strategy by the RSU.

This could result in an decrease of network channel

load while maintaining a high EAR.

ACKNOWLEDGEMENTS

This publication is funded by the Lower Saxony

Ministry of Science and Culture under grant num-

ber ZN3493 within the Lower Saxony “Vorab“ of the

Volkswagen Foundation and supported by the Center

for Digital Innovations (ZDIN).

VEHITS 2023 - 9th International Conference on Vehicle Technology and Intelligent Transport Systems

342

REFERENCES

Anwar, W., Franchi, N., and Fettweis, G. (2019). Physical

Layer Evaluation of V2X Communications Technolo-

gies. In 2019 IEEE 90th Vehicular Technology Confer-

ence (VTC2019-Fall), pages 1–7, Honolulu, HI, USA.

Boban, M., Barros, J., and Tonguz, O. (2014). Geometry-

Based Vehicle-to-Vehicle Channel Modeling for

Large-Scale Simulation. IEEE Transactions on Vehic-

ular Technology, 63(9):4146–4164.

Chtourou, A., Merdrignac, P., and Shagdar, O. (2021). Col-

lective perception service for connected vehicles and

roadside infrastructure. In 2021 IEEE 93rd Vehicu-

lar Technology Conference (VTC2021-Spring), pages

1–5.

Delooz, Q., Willecke, A., Garlichs, K., Hagau, A.-C., Wolf,

L., Vinel, A., and Festag, A. (2022). Analysis and

Evaluation of Information Redundancy Mitigation for

V2X Collective Perception. IEEE Access, 10:47076–

47093.

den Berghe, W. V. (2021). European Road Safety Observa-

tory Road Safety Thematic Report - Speed.

ETSI (2019a). ETSI TR 103 300-1 Intelligent Transport

System (ITS); Vulnerable Road Users (VRU) Aware-

ness.

ETSI (2019b). ETSI TR 103 562 v2.1.1 Intelligent Trans-

port System (ITS); Vehicular communications; Ba-

sic set of applications; Analysis of the Collective-

Perception Service (CPS).

ETSI (2020). ITS-G5 Access layer speciﬁcation for Intel-

ligent Transport Systems operating in the 5 ghz fre-

quency band.

Garlichs, K., Kaya, C., and Wolf, L. (2020). Leverag-

ing the Collective Perception Service for CAM Infor-

mation Aggregation at Intersections. In 2020 IEEE

92nd Vehicular Technology Conference (VTC2020-

Fall), pages 1–5.

Krajzewicz, D., Hertkorn, G., R

ossel, C., and Wagner, P.

(2002). SUMO - An Open-source Trafﬁc Simulation.

In Proceedings of the 4th middle East Symposium on

Simulation and Modelling (MESM20002), pages 183–

187.

Lobo, S., Festag, A., and Facchi, C. (2022). Enhancing the

safety of vulnerable road users: Messaging protocols

for v2x communication.

Pacella, F., Bonetto, E., Castillo, G. A. G., Brevi, D., and

Scopigno, R. (2021). Implementation and Latency As-

sessment of a Prototype for C-ITS Collective Percep-

tion. In 2021 IEEE International Mediterranean Con-

ference on Communications and Networking (Medit-

Com), pages 100–105.

Riebl, R., G

unther, H.-J., Facchi, C., and Wolf, L. (2015).

Artery: Extending Veins for VANET applications. In

2015 International Conference on Models and Tech-

nologies for Intelligent Transportation Systems (MT-

ITS), pages 450–456.

Schiegg, F., Brahmi, N., and Llatser, I. (2019). Analytical

Performance Evaluation of the Collective Perception

Service in C-V2X Mode 4 Networks. pages 181–188.

Schiegg, F. A., Llatser, I., Bischoff, D., and Volk, G. (2021).

Collective Perception: A Safety Perspective. Sensors,

21(1).

Sommer, C. and Dressler, F. (2011). Using the Right Two-

Ray Model? A Measurement Based Evaluation of

PHY Models in VANETs. In Proc. ACM MobiCom,

pages 1–3.

Wang, Z., Zheng, J., Wu, Y., and Mitton, N. (2017). A

Centrality-based RSU Deployment Approach for Ve-

hicular Ad Hoc Networks. In 2017 IEEE International

Conference on Communications (ICC), pages 1–5.

Willecke, A., Garlichs, K., Schulze, F., and Wolf, L. C.

(2021). Vulnerable Road Users Are Important As

Well: Persons in the Collective Perception Service. In

2021 IEEE Vehicular Networking Conference (VNC),

pages 24–31.

Yu, G., Li, H., Wang, Y., Chen, P., and Zhou, B. (2022).

A review on cooperative perception and control sup-

ported infrastructure-vehicle system. Green Energy

and Intelligent Transportation, 1(3):100023.

Enhancing Vulnerable Road User Awareness of Intelligent Transport Systems Through Relay and Aggregation of Collective Perception

Messages with Road Side Units

343