A DECENTRALIZED ROUTE GUIDANCE ALGORITHM IN

URBAN TRANSPORTATION NETWORKS

Ludovica Adacher, Gaia Nicosia

Dipartimento di Informatica e Automazione, Università Roma Tre, Rome, Italy

Keywords: Shortest Path, Intelligent Trans

portation Systems, Route Guidance.

Abstract: In the last decades, due to the increasing car traffic and the limited capacity of urban networks, algorithms

for traffic management and route guidance are becoming more and more important. GPS technology can be

used for fleet monitoring in urban or suburban areas, from a central monitoring station and may provide

useful information concerning the movement of all vehicles. Current route guidance systems are simple

from an algorithmic point of view (they compute shortest paths to the destination), but they have to deal

with huge size networks. For this reason, a decentralized approach, in which each vehicle independently

calculates its own route, is desirable. Naturally, to limit the congestion due the vehicles decisions, an

estimate on the different possible routes is required. Hence, we propose a decentralized algorithm in which

each vehicle computes its own route on the basis of the traffic information provided by the reference station.

Moreover, we propose a method for forcing vehicles to choose different paths and for informing the

reference station on the routes of all vehicles, so that traffic forecast is updated.

1 INTRODUCTION

The steadily growing car traffic and the limited

capacity of our streets demonstrate the necessity of

designing methods for better traffic management.

Studies [Jahn et al. 1999] show that an individual

“blind” choice of routes leads to travel times that are

between 6% and 19% longer than necessary.

Electronic and sensory devices are becoming more

popular and they provide or will shortly provide

detailed information about the actual traffic flows,

thus making available the necessary data to employ

better means of traffic management. So, the focus is

on developing Intelligent Transportation Systems

that are capable of better managing existing capacity

and encouraging more efficient vehicle routing over

time and space.

Many vehicles get equipped with the so-called

ute guidance systems. They guide the driver from

the origin to the destination by visual and acoustic

indicators. These systems in order to compute their

routes need digital maps, the current position

obtained by Global Position Systems (GPS), and

possibly up-to-date traffic data, which are broadcast

by radio or cellular phone. Most inefficiencies

caused by human route choice can be reduced by

route guidance systems, obtaining a better use of the

whole network. Unfortunately, many simulations

also predict that these benefits will be lost once the

number of equipped vehicles exceeds a certain

threshold. In fact, it is possible that such system

actually causes congestion, if the algorithms try to

minimize the individual journey time of each driver

separately, without taking into account the effects of

their own route recommendation. Thus, the need for

integrated algorithms that actually take into

consideration the overall road usage (which can be

viewed as the sum of all individual journey times)

has been recognized.

2 BASIC STRUCTURE OF ROUTE

GUIDANCE PROBLEMS

In the literature there are essential two different

approaches to route guidance problems: one

researches the system optimum, the other the user

equilibrium. When the interest is in minimizing the

total travel time, the best is to consider a global

perspective and compute the system optimum. In this

case, the existing road network may carry more

traffic [Wardrop 1952]. Unfortunately, this policy

may route some drivers on unacceptable long paths

in order to use shorter path for many other drivers.

The length of a route in the system optimum can be

318

Adacher L. and Nicosia G. (2004).

A DECENTRALIZED ROUTE GUIDANCE ALGORITHM IN URBAN TRANSPORTATION NETWORKS.

In Proceedings of the First International Conference on Informatics in Control, Automation and Robotics, pages 318-321

DOI: 10.5220/0001145603180321

 SciTePress

Figure 1: Conceptualized distributed transportation network.

significantly higher than in the user equilibrium

[Roughgarden and Tardos 2000]. The main

drawback of this approach is that after a while a user

may be unsatisfied and stop using the route guidance

system.

On the other hand, when minimizing individual

journey times of each driver, we can be far from the

system equilibrium. Many algorithms try to

minimize individual journey times, without taking

into account the effects of their route

recommendation. Under current systems, many

drivers could be given exactly the same route

recommendation. Therefore, assuming similar mean

speed, some drivers may always stay together on

their trip (platooning), and this may possibly lead to

congestion. While Adler and Blue [1998] call this

phenomenon oversaturation, Ben-Alkiva. De Palma

and Kaysi [1991] call it overreaction, since too

many drivers follow the recommended route. To

deal with this problem, it is essential to split

platoons over several paths (multiple path routing).

Many authors have proposed different approaches to

deal with multiple path routing. Rilett and Van

Aerde [1996] suggest adding individual random

error terms to the road travel times broadcast by a

central controller, in order to cause the in-vehicle

computers to choose different paths. Lee [1994]

computes k-shortest every ten minutes and then

distributes drivers over them every two minutes,

considering the current travel times on these paths.

Another popular approach is to route drivers

along the paths of so-called user equilibrium, so that

no driver can get a quicker path through the network

by unilaterally changing his route [Fresz T. L.,

1985]. This concept was introduced to model natural

driver behaviour and it has been studied extensively

in the literature.

In our work we assume that connected users are

provided with an Intelligent Traveller Information

System (ITIS) capable of providing route guidance

and/or traffic advice both pre-trip and while en-

route. ITIS is a term coined by Adler and Blue

[2002] to describe next generation information

devices that can gather and process information as

well as learn and represent user preferences and

behaviour.

3 PROBLEM DEFINITON AND

PROPERTIES

A traffic network, represented by a direct graph G=

(N,A), consists of a set of |N| of nodes and a set of |A|

links. Consider a situation in which a vehicle with

ITIS (Intelligent Traveller Information System, see

Figure 1) is currently travelling on link (i,s) towards

the destination node d, we want find on which link

(s,j) the vehicle should enter next, so as to minimize

the expected travel time to the destination node d.

It is assumed that the local controller (i.e. a

vehicle with ITIS) has available complete

information on the topology of sub network G′ (the

network representing all candidate paths from origin

o to destination d of the interested vehicle) and

current estimates of travel times on individual links.

Traffic flows have two important features that

make them difficult to study mathematically. One is

“congestion”, and the other is “time”. Congestion

captures the fact that travel times increase with the

amount of flow on the streets, while time refers to

the movement of vehicles along paths as “flow over

time”.

Congestion implies that transit time t

is not a

constant, but monotonically increases with the

augmentation of the flow value x

Flow variation over time is an important feature

in network flow problem arising in various

applications such as road or air traffic control,

production systems, and communication network.

A DECENTRALIZED ROUTE GUIDANCE ALGORITHM IN URBAN TRANSPORTATION NETWORKS

319

Figure 2: A very simple example.

Ford and Fulkerson [1958] observe that flow-over-

time problems in a given network with transit times

on the arcs can be transformed into equivalent static

flow problems in the corresponding time expanded

network.

4 DECENTRALIZED APPROACH

In many different areas of manufacturing, traffic

network, medicine, software engineering and etc.,

the decentralized approach reaches good

performances. Obviously, when all data are

available and of reasonable size, it is always possible

to use a centralized approach.

Naturally, a user will buy a route guidance

system only if he can obtain a benefit, but a system

optimum is not always a good solution from a user

point of view. Probably, a driver is induced to use a

route guidance system, if he/she is confident and

sure that the recommended route is always a good

route for him/her. In our work, we consider a

decentralized decision approach, where each driver

selects his/her route, on the basis of the traffic on the

network.

We consider a hierarchical structure with two

different levels: at a high level there is a reference

station, that represents a sort of real time database,

where all information related to the traffic on the

network are available; at the local level is present a

local controller (Intelligent Traveller Information

System ITIS). The reference station is responsible for

management functions such as collecting and storing

data, gathered from the network, and disseminating

traveller information.

Figure 1 depicts the set of travellers and flow

entities seeking to travel through the network.

4.1 Path computation

Each vehicle with ITIS calculates its path from the

origin o to destination d on the basis of some

information that is exchanged with the reference

station. For each link of the traffic network we know

the estimated minimum transit time t

min

and the

estimated maximum transit time t

max

. These two

parameters are calculated at the high level on the

basis of real-time, collected and stored data. For

each link e

min

is a lower bound on the transit time

(dependent on flow) on e;

max

is an upper bound on the transit time

(dependent on flow) on e.

As we will discuss in the next section, the value

max

can be calculated also by considering the

vehicles that are in the network, but have not

decided their route yet.

Each vehicle v , when computing its optimal

route from o

to d

, assumes that the transit time on

arc e is:

= α

min

+ (1- α

) t

max

The parameter

is a characteristic of the ITIS,

it is a real number

∈[0,1]. This characteristic

index

of the vehicle allows us to determine

different paths and avoid congestion on some links.

A very simple example in which different paths

corresponds to different values for

is depicted in

Figure 2.

4.1.1 Potential flow evaluation

We introduce the concept of potential flow as a way

of providing an estimate on transit times in the

immediate future by considering vehicles with ITIS

that are in the network, but have not decided their

route yet. Then, the potential flow represents all

users that have requested information on the traffic

of the network but have not calculated their personal

route yet. Naturally, it is necessary to have an

evaluation of these potential users to eliminate the

platooning effect.

Here, we consider two different ways to calculate

potential flow (in Figure 3 there is an example of the

two different potential flow evaluations). Since

traffic networks are usually very large and since

ICINCO 2004 - INTELLIGENT CONTROL SYSTEMS AND OPTIMIZATION

320

Figure 3: An example of potential flow evaluation in cases (a) and (b).

detailed information might not be necessary, we

divide the network G in different areas or sectors

and we proceed as follows.

Case (a): For each vehicle v we consider all paths

not exceeding a certain length from its origin o

destination d

. A vehicle v is part of a potential

flow for a sector s if an arc of its path is in sector

Case (b): For each vehicle v we consider a fixed

dimension (D) neighborhood of its origin o

and

its destination d

. This dimension D is related to

the dimension of the sector. The neighbourhoods

grow until they meet each other and the

dimension of the overlapping area is at least D (in

Figure 4, there is an example with D=3). A

vehicle v is a part of a potential flow for every arc

in the neighbourhoods.

We are currently running some tests in order to

evaluate which level of information is requested to

obtain a good solution and to determine the best

tradeoffs between detailed information and running

times.

5 CONCLUSIONS

In this work we are trying to utilize decentralized

approach for a hard problem like the route-guidance

problem. We have proposed a simple approach to

find a compromise between single users satisfaction

and global utilization of the network.

We are still at a preliminary stage of our

research. We are setting up a simulation, using

algorithm A* to calculate shortest paths, in order to

evaluate how the different choices for

lead to

different paths.

REFERENCES

Adler J. L., Blue V. J., 1998. Toward the design of

intelligent traveller information systems.

Transportation Research C, 6:157-172.

Adler J. L., Blue V. J, 2002. A cooperative Multi-agent

transportation management and route guidance

system. Transportation Research C. 433-454.

Ben-Akiva M. E., De Palma A. , and Kaysi I., 1991.

Dynamic network models and driver . Transportation

Research A, 25:251-266.

Ford L. R., Fulkerson D. R., 1958. Constructing maximal

dynamic flows from static flows. Operations Research

6: 418-433.

Fresz T. L., 1985. Transportation network equilibrium

design and aggregation: key development and

research opportunities. Transportation Research A.,

19:413-427.

Jahn O., Möhring R. H., Schulz A. S., Stier Moses N. E.,

2002. System Optimal Routing of Traffic Flows with

User Constraints in Networks with Congestion,

Technical Report 754/2002 Technische Universität

Berlin.

Jahn O., Möhring R. H., Schulz A. S., 1999. Optimal

Routing of Traffic flows with length restrictions in

networks with congestion. Technical Report 658/1999

Technische Universität Berlin.

Lee C. K., 1994. A multiple-path routing strategy for route

guidance systems. Transportation research C, 2:185-

195.

Rilett L.R., van Aerde M. W., 1991. Modeling distributed

real-time guidance strategies in a traffic network that

exhibits the Braess paradox. In proceedings of IEEE

Vehicle Navigation and Information Systems

Conference.

Roughgarden T., Tardos E., 2000. How bad is selfish

routing?. In proc. of the 41

annual IEEE, symposium

on foundation of computer science FOX’00.

Wardrop J. G., 1952. Some theoretical aspects of road

traffic research ,Proceeding of the Institution of civil

engineers.

A DECENTRALIZED ROUTE GUIDANCE ALGORITHM IN URBAN TRANSPORTATION NETWORKS

321