Hybrid Work Zone Information System

with Portable Changeable Message Signs

and Dedicated Short-Range Communication
U. Ibrahim, M. I. Hayee, and Eil Kwon
Several feasibility studies have been conducted (1520), and a few
prototypes have been demonstrated (2122) to justify the immense
impact of dedicated short-range communication (DSRC) as a key
enabling wireless communication technology in automated traffic
information systems in the work zone environment. In such systems,
DSRC-based communication can occur vehicle to infrastructure
(V2I), vehicle to vehicle (V2V), or both, given the application.
Normally, DSRC-based work zone traffic information systems
have two important components: (a) acquisition of traffic parameters
such as TT through the work zone and starting location of congestion
(SLoC), and (b) dissemination of these parameters to the vehicles as
they approach the work zone congestion area. Usually, both acquisition and dissemination of these parameters (e.g., TT and SLoC)
are accomplished with DSRC-based V2I or V2V communication, or
both. However, only those vehicles equipped to use DSRC technology
can take advantage of the disseminated safety information messages.
Thus such automated information systems for work zones may not
benefit those vehicles that lack DSRC capability. If a slow DSRC
market penetration rate is assumed in the beginning phase of future
DSRC deployment, there must be an efficient way to communicate
the work zone traffic parameters to all vehicles, with or without DSRC
capability.
To enable a smooth transition toward full deployment of DSRCbased work zone traffic information systems, integration of such
systems with existing technology is a necessary choice. Portable
changeable message signs (PCMSs) have been used extensively for
traffic control, and to display crucial travel-related information in
the work zone environment (23, 24). They are believed to command more motorist attention than static message signs and can be
dynamically configured at any time through both local and remote
means (25).
To acquire and disseminate travel parameters with DSRC-based
V2V or V2I communication, it is not necessary for all vehicles on the
road to be equipped with DSRC technology. As long as the traffic
parameters can be acquired and disseminated with less than a 100%
DSRC market penetration rate, these parameters can be communicated to vehicles not equipped with DSRC through PCMSs strategically placed alongside the road. To accomplish such communication,
a DSRC-PCMS interface needs to be developed. The PCMSs can
then be configured to receive and display new traffic parameters from
any nearby DSRC-equipped vehicle on the road.
This paper describes the development of a DSRC-based hybrid
information system for the work zone through the use of PCMSs. This
hybrid system is an extension of an already developed DSRC-based
travel information system, which fully relied on DSRC technology
for both acquisition and dissemination of TT and SLoC (22). In the

The future deployment of dedicated short-range communication (DSRC)

technology requires that applications with their bases in DSRC be integrated with existing traffic management techniques so that vehicles not
equipped with DSRC at the early stage of DSRC deployment can also
reap the potential benefits of DSRC technology. A hybrid traffic information system was successfully developed; it combines DSRC technology
and portable changeable message signs (PCMSs) for use in the work
zone environment to improve traffic mobility and thereby driver safety.
The developed system uses DSRC-based vehicle-to-infrastructure and
vehicle-to-vehicle communication to acquire travel safety parameters,
such as travel time (TT) and the starting location of congestion (SLoC),
and to disseminate these parameters to DSRC-equipped vehicles and
PCMSs, which are strategically placed along the roadside. Through the
use of the DSRC-PCMS interface developed and demonstrated in this
work, PCMSs can receive these travel safety parameters from nearby
DSRC-equipped vehicles on the road through DSRC-based vehicle-tovehicle communication, and display them for the drivers of vehicles that
lack DSRC capability. Such a system can be useful during the early
stage of DSRC deployment when DSRC market penetration is low. In
addition, a rigorous analysis was conducted to investigate the minimum
DSRC market penetration rate needed for successful functionality of
the developed system with respect to both acquisition and dissemination
of TT and SLoC. Through the use of a realistic traffic flow model, guidelines were developed to estimate a minimum DSRC penetration rate
needed to deploy the developed system for a variety of traffic scenarios
on a given work zone road.

Work zone congestion can grow quickly and produce hazardous conditions, especially during rush hour (1, 2). This phenomenon highlights
the need for a traffic information system to notify drivers with timely
updates on travel time (TT) and congestion lengths (3, 4). Many
research reports by FHWA have clearly shown that an automated
traffic information system is highly desired to minimize backups
and improve safety (58). Intelligent transportation systems (ITSs)
and other such coexisting technologies are being pursued aggressively
to devise such automated traffic information systems (914).
U. Ibrahim, 73 MWAH, and M.I. Hayee, 271 MWAH, Department of Electrical and
Computer Engineering, and E. Kwon, 252 SCIV, Department of Civil Engineering,
University of Minnesota, Duluth, MN 55812. Corresponding author: U. Ibrahim,
ibra0130@d.umn.edu.
Transportation Research Record: Journal of the Transportation Research Board,
No. 2380, Transportation Research Board of the National Academies, Washington,
D.C., 2013, pp. 2935.
DOI: 10.3141/2380-04
29

30

Transportation Research Record 2380

newly developed hybrid system, acquisition of TT and SLoC is

still done through DSRC communication, but the DSRC-equipped
PCMSs are placed at strategic locations to display the updated traffic
parameters to vehicles that lack DSRC capability.
The main objective of the newly developed hybrid DSRC-PCMS
work zone information system is to acquire TT and SLoC through
DSRC technology, and to disseminate those parameters through
the use of a hybrid of DSRC and PCMS technologies to vehicles that
are farther away and that travel toward the work zone congestion.
Each time a new set of TT and SLoC is estimated, both of these
parameters are periodically disseminated to DSRC-equipped vehicles.
They also are disseminated to the DSRC-equipped PCMSs, strategically placed along the roadside, through the use of DSRC-based
V2V communication. Vehicles that have the DSRC capability can
take advantage of these parameters (TT and SLoC) directly through
the creation of an internal alert to the driver. Drivers of vehicles that
lack the DSRC capability can obtain this information if they look at
a roadside PCMS that displays the updated values of TT and SLoC.
Rigorous analysis was performed to formulate criteria to find the
minimum DSRC penetration rate needed for reliable functionality
of the developed system for both acquisition and dissemination of
travel parameters. Under realistic traffic flow conditions, the minimum
DSRC market penetration rate necessary to deploy the developed
hybrid information system for the work zone was found for both
nonrush and rush-hour traffic scenarios.

Hybrid DSRC-PCMS System Architecture

A conceptual diagram of the developed hybrid work zone information system depicts a typical work zone on a two-lane highway
where congestion is growing because of a lane closure (Figure 1).
In previous research, a work zone travel information system was
developed, which relied on DSRC-based V2I and V2V communication for both acquisition and dissemination of TT and SLoC (21).
DSRC-equipped PCMSs were added to the previously designed
system to make a hybrid DSRC-PCMS work zone travel information
system (Figure 1). Because DSRC-equipped PCMSs were added

RSU

to strategic locations on the roadside, all those vehicles that lack the
DSRC capability will now be able to take advantage of the updated
TT and SLoC information. Given the availability of PCMSs, they
can be located every couple of miles or just before an alternative route,
if present, or both. The detailed guidelines for the placement of
PCMSs for different traffic scenarios are provided in the Manual on
Uniform Traffic Control Devices (26).
In the developed system, a portable, DSRC-based, central roadside
unit (RSU) is installed at a height that provides a clear line of sight
for vehicles as they pass by and come into its direct, wireless access
range. The location of the central RSU with respect to the work
zone is determined such that its wireless access range on one side
coincides with the downstream end of the work zone or the ending
location of congestion (ELoC). In general, the ELoC is fixed, whereas
the SLoC may vary according to the incoming traffic flow at different times of the day.
The central RSU is installed and initialized with typical user input
parameters (e.g., ELoC, posted speed limit, direction) according to
the specific work zone environment (22). After its initialization,
the software of the central RSU will control the back-and-forth,
DSRC-based communication for all DSRC-equipped vehicles that
pass through the work zone congestion area and will use either V2I
or V2V communication, which will depend on whether a vehicle
is within or beyond its direct, wireless access range. The vehicle
hardware contains DSRC radio communication capability as well
as a Global Positioning System receiver. The latter is needed so that
the current location of the vehicle can be known.
After the central RSU is initialized, it selects a DSRC-equipped
vehicle to monitor its speed and location information to estimate
TT and SLoC. The RSU selects a DSRC-equipped vehicle, preferably
located well before the SLoC. The preferred area of selection before
the SLoC is shown as the desired region in Figure 1. Because the SLoC
could vary, given the traffic influx, the RSUs software is designed to
possess the capability to vary the desired region so that it always falls
well before the SLoC.
To engage a vehicle for traffic information acquisition, the central
RSU periodically transmits invitation messages to DSRC-equipped
vehicles through a combination of V2I and V2V communication.

DSR
equip C
pe
PCMS d

DSRC
equipped
PCMS

FIGURE 1 Conceptual architectural diagram of developed hybrid DSRC-PCMS information system for work zone.

Ibrahim, Hayee, and Kwon

31

Incoming INFO
messages

WORK
ZONE
AHEAD

Alternating
messages

Serial connection
DSRC Unit

SC4
Controller

TRAVEL
TIME
36 MINS
QUEUE
AHEAD
0.7 MI

Sign-controller
connection

FIGURE 2 Schematic diagram of DSRC-PCMS hardware interface.

DSRC-equipped vehicles located in the desired region will respond

to the invitation messages through acknowledgments to the RSU.
One of the responding DSRC-equipped vehicles is selected to acquire
traffic data by the central RSU. The selected DSRC-equipped vehicle
then periodically sends back to the central RSU its location and
speed information as it passes through the congestion area. When
the selected DSRC-equipped vehicle travels beyond the ELoC, the
central RSU estimates the TT and SLoC on the basis of the data
received by this vehicle with the use of a threshold-based definition
of congestion (22).
Once TT and SLoC are estimated, the information message that
contains updated TT and SLoC parameters is broadcast to all DSRCequipped vehicles and PCMSs, periodically (e.g., every few seconds).
However, the information message is updated with a new set of values
of TT and SLoC only when a selected DSRC-equipped vehicle travels beyond the ELoC, which completes the information exchange
cycle with the central RSU as highlighted above. Normally, only
one vehicle is selected and monitored at a time. However, if the TT
turns out to be greater than a predefined threshold, then a new vehicle
is selected after every update time so that a new set of TT and SLoC
can be acquired every update time. Usually, an update time needs
to be selected large enough in which SLoC or TT can appreciably
change.
During this whole process of TT and SLoC estimation, many
messages are exchanged between the selected DSRC-equipped
vehicles and the central RSU with the DSRC-based V2I or V2V
communication, or both. The Society of Automotive Engineers
has specified safety message composition for the DSRC communication in its draft standard SAE J2735. In the work reported here,
message formats were used that comply with such standards and
contain the mandatory fields for such message types as traveler
information message and basic safety message. The messages that
the protocols use in the work reported here contain the data fields as
specified in the J2735 standard, and the entire message is encoded
and communicated according to the standardized protocol. In addition, in the back-and-forth communication between the central RSU
and DSRC-equipped vehicles, all information is exchanged without
retention of any vehicle identity information to maintain privacy.

the updated parameters on its display matrix for the benefit of passing drivers. While the DSRC-PCMS interface undergoes design, the
message format is maintained according to the guidelines for the use
of PCMSs in the work zone, as suggested by the Manual on Uniform
Traffic Control Devices.
A schematic of the DSRC-PCMS interface is shown in Figure 2.
The DSRC-PCMS hardware interface design is accomplished through
the use of the RS232 serial port connection between the DSRC unit
and the PCMS. In the work reported here, a PCMS device was used,
which was made by ADDCO LLC (an Imago Company), Saint Paul,
Minnesota. This particular PCMS type is the one sold most in North
America, and it is fully compliant with the national transportation
communications for ITS protocol (27). The PCMS comes with a
proprietary logic controller, called SC4, and uses a modified higher
data link layer control language to let the external agents communicate
with the controller.
As soon as the central RSU acquires a new set of TT and SLoC,
it updates the information message, which is broadcast periodically to all DSRC-equipped vehicles and DSRC-equipped PCMSs.
Once the updated information message is received by the DSRC
unit of the PCMS, the data are formatted in an encoded message
with proper higher data link layer control, which is communicated
serially to the PCMS controller (i.e., SC4). The PCMS controller
processes the encoded message to create a display pixel map, which
is then sent to the display matrix in proper format to light the
corresponding LEDs.
The message continues to be displayed on the display matrix until
an updated information message is received that contains new TT
and SLoC values. The actual display message will contain the TT and
the distance to the queue ahead. The DSRC device of the PCMS has
a Global Positioning System capability that helps calculate the distance from its current position to the SLoC (i.e., the distance to the
queue ahead). This distance and the TT were displayed alternately
with a few seconds interval (a 3-s interval was used in this study)
because both of these parameters could not fit into the three lines by
eight characters display matrix. The two typical alternate messages
are shown as an inset in Figure 2.

DSRC-PCMS Interface

Traffic Flow and Vehicle Density

for Hybrid System

To integrate PCMS into the previously developed work zone information system, a DSRC interface with PCMS was developed and
field demonstrated. With this interface, a PCMS could receive an
information message that contained TT and SLoC from a nearby
DSRC-equipped vehicle through V2V communication, and show

In a hybrid DSRC-PCMS system, the role of the central RSU

is critical because it engages with other vehicles on the road to
acquire travel parameters such as TT and SLoC. The acquisition
of TT and SLoC is accomplished through the selection of a vehicle
in the desired region (Figure 1) to be monitored periodically for its

32

speed and location information. Both DSRC-based V2I and V2V

communication are needed because the desired region is quite far
and well beyond the direct, wireless access range of the central
RSU. Similarly, the acquired parameters need to be disseminated
to the DSRC-equipped vehicles and PCMSs, well beyond the
SLoC. Thus V2V communication is key to accomplishing this
task as well.
The reliable acquisition of TT and SLoC requires that a DSRCequipped vehicle be found and selected in a timely manner whenever
the central RSU starts to look for a new DSRC-equipped vehicle in
the desired region. Similarly, the reliable dissemination of the TT
and SLoC requires the availability of enough DSRC-equipped vehicles to facilitate message propagation through DSRC-based V2V
communication. Thus a minimum traffic flow rate and a minimum
vehicle density for a given DSRC penetration rate are needed to
successfully accomplish the tasks of acquisition and dissemination
of TT and SLoC.
Determination of the minimum traffic flow rate, along with the
DSRC penetration rate, ensures that a DSRC-equipped vehicle can
be found and selected whenever the central RSU needs to update
TT and SLoC. Similarly, a minimum traffic density, along with the
DSRC penetration rate, ensures that enough DSRC-equipped vehicles
are on the road to facilitate the V2V communication needed to acquire
and disseminate TT and SLoC.
Traffic flow and vehicle density give rise to statistical distribution of vehicles in time and space, respectively. The most commonly
followed stochastic traffic model is Cowans headway model (28).
According to Cowans model, vehicle distribution in time and space is
Poisson in nature. In general, Poisson is applicable when the vehicle
density is light so that free traffic conditions exist (i.e., the arrival of
a given vehicle is not affected by any other vehicle that precedes it).
However, when the vehicle density grows enough to lead to congested
traffic conditions, the vehicle distribution in time and space becomes
uniform instead of Poisson. In the research described here, the relevant vehicle distribution models were applied to determine the
minimum traffic flow and vehicle density needed for reliable acquisition and dissemination of TT and SLoC. Thus the minimum required
DSRC market penetration rate was found for reliable functionality
of the developed system.

Traffic Flow Rate and TT and SLoC Acquisition

The reliable acquisition of TT and SLoC depends on the central RSUs
ability to timely find and select a DSRC-equipped vehicle within the
desired region, which in turn depends on the total number of vehicles
that cross the desired region in a given time (i.e., traffic flow and DSRC
penetration rate). If the traffic flow or the DSRC market penetration
rate (or both) is small, the central RSU may have to wait for a long time
before a DSRC-equipped vehicle passes through the desired region,
and the acquisition cycle may not proceed efficiently.
Because the desired region is always located well before the
congestion starts, the traffic flow through the region is free as opposed
to a bounded flow, which builds up gradually after the desired region
and leads toward the congested flow around SLoC. During rush hour,
this situation may not exist because the congestion stretches for a
much longer distance. In that case, the SLoC is located at a point
from which TT needs to be calculated because there is no SLoC
in reality.
The traffic flow of incoming vehicles determines how many vehicles will cross the desired region in a given time. Given the free-flow

Transportation Research Record 2380

condition for a given traffic flow rate, q vps, a total of qT vehicles

will cross the center point of the desired region in time T. If a
Poisson arrival distribution is assumed, the probability (p) that exactly
n number of vehicles cross the center point of the desired region in
time T is described by Equation 1 (29):
p ( n ) = ( qT )

e qT
n!

(1)

From this equation the probability that no vehicle crosses (n = 0)

the center point of the desired region will determine the cumulative
probability of time headway, h, as described in Equation 2.
p ( h T ) = 1 e qT

(2)

Equation 2 gives the proportion of the total number of vehicles (qT)

in time T with time headway h T. Therefore, the total number
of vehicles (N) with time headway h T, is given by Equation 3.

( qT )(1 e qT ) = N

(3)

If it is assumed that the DSRC penetration rate is k (fraction of the

total number of vehicles), the number of DSRC-equipped vehicles
that cross the center of the desired region in time T is kN. Equation 3
is numerically solved to find T in which one DSRC-equipped vehicle
crosses the center of the desired region, for a given traffic flow, q,
with different values of DSRC penetration rate, k. For example, if
the penetration rate is 10% (k = 0.1), Equation 3 is solved for N = 10,
with the assumption that one out of every 10 vehicles, on average,
that crosses the center of the desired region in time T, will be DSRC
equipped. The results are shown in Figure 3 in which T (i.e., average time lapse between two adjacent DSRC-equipped vehicles that
cross the center of the desired region) is plotted versus q for different
values of k.
Figure 3 shows the average time lapse between two adjacent
DSRC-equipped vehicles that cross the center of the desired region
in free-flow traffic conditions (i.e., Poisson temporal distribution
is assumed). However, if the congestion extends much longer distances so that free-flow traffic conditions do not hold true anymore,
the vehicle temporal distribution effectively becomes narrowly uniform around average time headway. If the narrow, uniform, temporal
distribution is considered, the average time lapse between two
adjacent DSRC-equipped vehicles that cross the center of the desired
region will become 1/(kq), which is comparable to the results shown
in Figure 3. Thus Figure 3 can be represented for congested and
free-flow conditions to find out the average time lapse between two
DSRC-equipped vehicles that cross the center of the desired region
for any traffic flow rate q and DSRC penetration rate k.
Figure 3 can help estimate the required DSRC penetration rate at
a specific traffic flow rate, which is needed for the central RSU to
reliably find and select a DSRC-equipped vehicle in any given time
interval. This time interval should be a small fraction of the update
time after which the central RSU starts to search for a new DSRCequipped vehicle in the desired region. If the update time is assumed
to be 10 min, then 30 s is chosen as the time interval in which the
central RSU will search for a DSRC-equipped vehicle. The 30-s
time interval is reasonable because it is 5% of the update time of
10 min. If an update is needed more often, a shorter time interval
needs to be considered.

Average time lapse between two adjacent DSRC-equipped

vehicles crossing the center of desired region (seconds)

Ibrahim, Hayee, and Kwon

33

60
DSRC Penetration Rates
50

10%
20%

40

30%
50%

30

20

10

200

400

600

1000
1200
800
Traffic flow (veh/hour)

1400

1600

1800

2000

FIGURE 3 Average time lapse between two adjacent DSRC-equipped vehicles as they cross center of desired region as function
of traffic flow for different DSRC penetration rates.

Vehicle Density and TT

and SLoC Dissemination

in length (L) with a space headway less than L for a given vehicle
density (D).

As described earlier, to disseminate TT and SLoC, a minimum vehicle density is needed to sustain DSRC-based V2V communication
for a given DSRC penetration rate. For any given vehicle density,
the vehicles on the road are spatially distributed in random fashion.
Like the temporal distribution or time headway, the spatial distribution or space headway can be derived from the Poisson distribution
for free-flow traffic conditions. An analytical approach similar to
the one developed in the previous section for time headway can be
used to create a modified equation for the total number of vehicles

( DL ) (1 e DL ) = N

(4)

If it is assumed that the DSRC penetration rate is k, there will be kN

DSRC-equipped vehicles present on each road section of length L.
Equation 4 is numerically solved to determine the average distance
L in which one DSRC-equipped vehicle is present for different
DSRC penetration rates. The results are shown in Figure 4, in which
the average distance between two adjacent DSRC-equipped vehicles is shown versus vehicle density for different DSRC penetration

Average distance between two adjacent

DSRC-equipped vehicles (meters)

500
450
DSRC Penetration Rates
400

10%

350

20%
30%
50%

300
250
200
150
100
50
0

10

20

30

40
50
60
Vehicle density (veh/mile)

70

80

FIGURE 4 Average distance between two adjacent DSRC-equipped vehicles as function of vehicle density for different DSRC
penetration rates.

90

100

34

Transportation Research Record 2380

rates. Equation 4 is applicable to free-flow conditions, but if the

traffic flow is congested, the average distance between two adjacent
DSRC-equipped vehicles becomes 1/kD for a given vehicle density
(D) and DSRC penetration rate (k).
As in time headway analysis, the average distance between two
adjacent DSRC-equipped vehicles remained the same for congested
and free-flow conditions for a given vehicle density, D, and DSRC
penetration rate, k. Therefore, Figure 4 represents the average distance between two DSRC-equipped vehicles for both free and congested traffic flow scenarios. However, a difference arises in the
spatial distribution of the DSRC-equipped vehicles, which is narrowly
uniform in congested flow and Poisson in free flow.
Figure 4 can help estimate the required DSRC penetration rate
needed to sustain the V2V communication needed for dissemination
of TT and SLoC. Although, on average, the distance between any
two adjacent DSRC-equipped vehicles is L, the crucial length to
consider to sustain V2V communication is 2L because the maximum
distance between two adjacent DSRC-equipped vehicles may be
2L in some sections of the road. This scenario is the worst case, in
which a DSRC-equipped vehicle is present on the extreme left side
of the road section of length L, and on the adjacent road section of
the same length, a DSRC-equipped vehicle is present on the extreme
right side, which makes the distance between the two DSRC-equipped
vehicles 2L.
If the direct wireless range of the DSRC units is considered to
be 500 m, the distance between any two adjacent DSRC-equipped
vehicles should be no more than 500 m to sustain the V2V communication. Thus the average distance between the two DSRCequipped vehicles should be 250 m. However, there might arise a
reason (e.g., temporary loss of the line of sight, shutdown of a given
vehicles DSRC unit) that might cause the V2V communication
chain to be interrupted and impair the reliable dissemination of TT
and SLoC. One way to get around this situation is to consider an
average distance of 125 m instead of 250 m between two adjacent
DSRC-equipped vehicles to double the number of DSRC-equipped
vehicles available for V2V communication. However, if a practical
work zone road situation is assumed in which there are, in general,
two lanes in each direction, the number of vehicles available for
V2V communication will, in fact, be two, including DSRC-equipped
vehicles for both lanes. Furthermore, DSRC-equipped vehicles in
the two lanes in the opposite direction could also help sustain V2V
communication. Here the criterion was an average distance between
two DSRC-equipped vehicles of 250 m to reliably sustain the V2V
communication needed for dissemination of TT and SLoC.

DSRC Penetration Rate

The required traffic flow and vehicle density for a given DSRC
penetration rate can be estimated from Figures 3 and 4, respectively.
As explained earlier, a 30-s interval is the criterion to find a DSRCequipped vehicle to acquire TT and SLoC, and an average distance
of 250 m between the two DSRC-equipped vehicles is the criterion to
sustain V2V communication to disseminate TT and SLoC. For example, for a 10% DSRC market penetration rate, a 30-s interval criterion
means that the traffic flow rate should be 1,200 vps (Figure 3). Similarly, for a 10% DSRC market penetration rate, the criterion of a 250 m
average distance between two DSRC-equipped vehicles suggests that
the traffic density should be 65 vpm for a given lane (Figure 4). With
the same method, the required traffic flow and densities for different
DSRC market penetration rates are estimated for different values
of DSRC penetration rate, and the results are shown in Figure 5.
Figure 5 can help estimate the required DSRC penetration rate for the
developed system to reliably function on a given work zone road for
known traffic conditions (i.e., traffic flow and vehicle density).
As an example, analysis was done on the real-time data collected
on a two-lane road section in the MinneapolisSaint Paul, Minnesota,
metropolitan area (southbound I-169), which contained flow, density,
and speed information under driving conditions during a typical work
day. With the use of Figure 5, the required DSRC penetration rates
were determined for this practical scenario. Nonrush and rush hour
traffic data were analyzed separately.
Data for the nonrush hour (10 a.m. to 2 p.m.) suggested that during
this time, traffic flow ranged from 1,000 to 1,500 vph with corresponding traffic densities of 21 to 26 vpm for a given lane, which thereby
maintained an average speed of 55 mph (i.e., free-flow condition).
In this situation, a DSRC penetration rate of about 15% (i.e., worstcase traffic flow of 1,000 vph) is required, as estimated from Figure 5.
Similarly, a DSRC penetration rate of about 35% (i.e., for worst-case
traffic density of 21 vpm) is required as estimated from Figure 5.
Therefore, a DSRC penetration rate of at least 35% (dictated by
minimum density) is needed for the developed system to successfully
work under this scenario.
Similarly, data for the rush hour (6 to 10 a.m.) suggested that
the traffic flow dominantly ranged from 1,300 to 1,800 vphpl,
with vehicle densities that ranged from 30 to 80 vpmpl. Moderately
higher flow than during nonrush hour, but with significantly higher
density, meant that the average speed decreased from 55 to 25 mph
(i.e., congestion had developed). However, in this situation, both
worst-case traffic flows (1,300 vph) and vehicle density (30 vpm) were

FIGURE 5 Required traffic flow and density for given DSRC penetration rate.

Ibrahim, Hayee, and Kwon

large enough to warrant successful functionality of the developed

system, with a DSRC penetration rate of a little less than 20%.
Conclusions
This paper describes a newly developed hybrid PCMSDSRC
information system for the work zone. The developed system can
acquire important travel parameters such as TT and SLoC through
DSRC-based V2I and V2V communication and then periodically
broadcast them to drivers of DSRC-equipped vehicles as well as
to DSRC-equipped PCMSs. In this hybrid system, DSRC-equipped
PCMSs are strategically placed alongside the work zone road. They
are treated just like DSRC-equipped vehicles as recipients of information messages, except that they can display the received information
messages to many passing drivers who lack the DSRC capability. For
this purpose, a DSRC-PCMS interface was developed to help PCMSs
receive safety messages, which contained TT and SLoC from a nearby
DSRC-equipped vehicle through DSRC-based V2V communication.
Rigorous analysis was conducted to investigate the minimum
DSRC market penetration rate needed for the developed hybrid
system to successfully acquire and disseminate TT and SLoC. The
results of this analysis, when applied to a practical road scenario,
indicated that a market penetration rate that ranged from 20% to
35% was needed for the system to work, with the lower rate needed
for rush hour conditions. Although this analysis was specific to one
road situation, the implication was that the required DSRC penetration
rate during rush hour in general was less than the DSRC penetration
rate required in nonrush hour for the developed system to reliably
work. The vehicle density to sustain DSRC-based V2V communication was much higher in rush hour, which was a limiting factor to
determine the minimum DSRC penetration rate needed for reliable
dissemination of the information message.
References
1. Lajunen, T., D. Parker, and H. Summala. Does Traffic Congestion
Increase Driver Aggression? Transportation Research Part F: Traffic
Psychology and Behavior, Vol. 2, No. 4, 1999, pp. 225236.
2. Van Eenennaam, E.M., and G. Heijenk. Providing Over-the-Horizon
Awareness to Driver Support Systems. Proc., 4th IEEE Workshop on
Vehicle to Vehicle Communications, 2008.
3. Wegener, A., M. Pirkowski, M. Raya, H. Hellbrck, S. Fischer, and
J.P. Hubaux. TraCI: An Interface for Coupling Road Traffic and Network
Simulators. Proc., 11th Communications and Networking Simulation
Symposium, New York, 2008, pp. 155163.
4. Marfia, G., and M. Roccetti. Vehicular Congestion Modeling and Estimation for Advanced Traveler Information Systems. Proc., International
Federation for Information Processing Wireless Days, Venice, Italy,
Oct. 2022, 2010, pp. 15.
5. Luttrell, T., M. Robinson, J. Rephlo, R. Haas, J. Srour, R. Benekohal,
J.S. Oh, and T. Scriba. Comparative Analysis Report: The Benefits of
Using Intelligent Transportation Systems in Work Zones. FHWAHOP-09-002. FHWA, U.S. Department of Transportation, 2008.
6. Chu, L., H.-K. Kim, Y. Chung, and W.W. Recker. Evaluation of Effectiveness of Automated Work Zone Information Systems. In Transportation
Research Record: Journal of the Transportation Research Board, No. 1911,
Transportation Research Board of the National Academies, Washington,
D.C., 2005, pp. 7381.
7. Intelligent Transportation Systems in Work Zones: A Cross-Cutting
Study. Integrated Work Zone Systems for Improving Travel Conditions
and Safety. FHWA-OP-02-025. FHWA, U.S. Department of Transportation, 2002.
8. Real-Time Work Zone Traffic Control System Using an Automated
Traffic Information System to Reduce Congestion and Improve Safety
During Reconstruction of the I-55 Lake Springfield Bridge in Illinois.

35

Intelligent Transportation Systems in Work Zones: A Case Study. FHWAHOP-04-018. FHWA, U.S. Department of Transportation, 2004.
9. Bushman, R., and C. Berthelot. Effect of Intelligent Transportation
Systems in Work Zones: Evaluation of North Carolina Smart Work
Zones. North Carolina Department of Transportation, Raleigh, 2004.
10. Bscher, M., P. Coulton, C. Efstratiou, H. Gellersen, D. Hemment, R.
Mehmood, and D. Sangiorgi. Intelligent Mobility Systems: Some SocioTechnical Challenges and Opportunities. In Communications Infrastructure: Systems and Applications in Europe. (R. Mehmood, E. Cerqueira, R.
Piesiewicz, and I. Chlamtac, eds.), Springer, London, pp. 140152, 2009.
11. Intelligent Transportation Systems in Work Zones: A Case Study
(Dynamic Lane Merge System). FHWA-HOP-04-033. FHWA, U.S.
Department of Transportation, 2004.
12. Lee, E.-B., and C. Kim. Automated Work Zone Information System on
Urban Freeway Rehabilitation: California Implementation. In Transportation Research Record: Journal of the Transportation Research Board,
No. 1948, Transportation Research Board of the National Academies,
2005, pp. 7785.
13. Fontaine, M.D. Guidelines for Application of Portable Work Zone Intelligent Transportation Systems. In Transportation Research Record: Journal
of the Transportation Research Board, No. 1824, Transportation Research
Board of the National Academies, Washington, D.C., 2003, pp. 1522.
14. Alazawi, Z., S. Altowaijri, R. Mehmood, and M.B. Abdljabar. Intelligent
Disaster Management System Based on Cloud-Enabled Vehicular Networks. Proc., 11th International Conference on ITS Telecommunications,
St. Petersburg, Russia, August 2325, 2011, pp. 361368.
15. Xu, Q., T. Mak, J. Ko, and R. Sengupta. Vehicle-to Vehicle Safety Messaging in DSRC. Proc., 1st ACM Workshop on Vehicular Ad Hoc Networks,
Philadelphia, Pa., Oct. 1, 2004.
16. Qian, Y., K. Lu, and N. Moayeri. A Secure VANET Mac Protocol for
DSRC Applications. Proc., IEEE Globecom, New Orleans, Nov. 2008.
17. Kandarpa, A.R., M. Chenzaie, M. Dorfman, J. Anderson, J. Marousek,
I. Schworer, J. Beal, C. Anderson, T. Weil, and F. Perry. Final Report.
Vehicle Infrastructure Integration Proof of Concept Test. Executive
Summary. Booz Allen Hamilton, McLean, Va., February 2009.
18. Hsu, C., C. Liang, L. Ke, and F. Huang. Verification of On-Line Vehicle
Collision Avoidance Warning System Using DSRC. World Academy of
Science, Engineering and Technology, Vol. 55, July 2009.
19. Misener, J.A., and S.E. Shladover. PATH Investigations in VehicleRoadside Cooperation and Safety: A Foundation for Safety and VehicleInfrastructure Integration Research. Proc., IEEE Intelligent Transportation
Systems Conference, Sept. 916, 2006.
20. Liu, Y., O. Ozguner, and E. Ekici. Performance Evaluation of Intersection
Warning System Using a Vehicle Traffic And Wireless Simulator. Proc.,
IEEE Intelligent Vehicles Symposium, June 68, Las Vegas, Nev., 2005,
pp. 171176.
21. Maitipe, B.R., M.I. Hayee, and E. Kwon. Development and Field
Demonstration of DSRC-Based V2I Traffic Information System for the
Work Zone. Final Report. University of Minnesota, Center for Transportation Studies, ITS Institute, Duluth, 2010.
22. Maitipe, B.R., U. Ibrahim, M.I. Hayee, and E. Kwon. Vehicle-toInfrastructure and Vehicle-to-Vehicle Information System in Work Zones:
Dedicated Short-Range Communications. In Transportation Research
Record: Journal of the Transportation Research Board, No. 2324, Transportation Research Board of the National Academies, Washington, D.C.,
2012, pp. 125132.
23. Weng, J., and Q. Meng. Rear-End Crash Potential Estimation in the
Work Zone Merging Areas. Journal of Advanced Transportation, 2012.
24. Fang, C. Portable Intelligent Traffic Management System for Work
Zones and Incident Management Systems: Best Practice Review. Proc.
11th International IEEE Conference on Intelligent Transportation
Systems, Oct. 1215, 2008, pp. 563568.
25. Ullman, G.L., C.L. Dudek, and B.L. Ullman. Development of a Field
Guide for Portable Changeable Message Sign Use in Work Zones. Texas
Transportation Institute, College Station, 2005.
26. Manual on Uniform Traffic Control Devices. FHWA, U.S. Department
of Transportation, 2003.
27. ADDCO LLC Website. Model DLH1000-ALS. http://www.addco.
com/-en/products/traffic/portables/als/. Accessed: Jan. 2012.
28. Cowan, R.J. Useful Headway Models. Transportation Research Vol. 9,
No. 6, 1975, pp. 371375.
29. Gartner, N., C. Messer., and A.K. Rathi. Revised Monograph on Traffic
Flow Theory, Unsignalized Intersections. Presented at 84th Annual Meeting of the Transportation Research Board, Washington, D.C., 2005.
The Intelligent Transportation Systems Committee peer-reviewed this paper.