Learning Center
Plans & pricing Sign in
Sign Out



									            Simulation of the dynamic interactions between terror
                  and anti-terror organizational structures

                                    Stanislaw Raczynski
                          Universidad Panamericana, Mexico City

Keywords: simulation, modeling, terrorism, discrete event, agent-oriented, social
simulation, soft systems

A discrete-event model of the dynamics of certain social structures is presented. The
structures include terror organizations, anti-terror and terror-supporting structures. The
simulation shows the process of creating the structures and their interactions. As the result
we can see how the structure size changes and how the interactions work, including the
process of destroying terror organization links by the anti-terror agents. The simulation is
agent-oriented and uses the PASION simulation system.

The US State Department defines terrorism as "premeditated, politically motivated violence
perpetrated against noncombatant targets by sub-national groups or clandestine agents,
usually intended to influence an audience". There are other definitions and opinions, but all
of them coincide in the characteristics of premeditated and planned in advance, politically
motivated action. Undoubtedly such actions need to be supported by some kind of
organization with well defined structure and operation rules. Human societies have always
been developing social structures. The difference between human-created structures and
those formed by animal populations lies only in the degree of sophistication. Our
organizations can support religions, scientific development and politics, as well as
destruction, crime and racial or religious extermination.
A social organizational structure is not just a set of individuals interacting with each other.
Such structure should be treated as a system. This means that it may have (and nearly
always has) an objective or "superior" goal, and intents to satisfy it. This objective may
even be opposite to the objectives of every one of the structure members, but, in many
cases known from history, it cannot be influenced or changed by the members. A social
structure acts as a new agent, using its members as nothing more than a medium to achieve
the goal.
For example, a goal of any political party is always to get power and nothing more. Many
trade union organizations have lost their original goal (defending the interests of workers)
and beceme also a power-seeking structures. The interaction between different social
structures is an interesting problem and can be simulated, to some extend, of course.
Consult an article I wrote more than 30 years ago ( Raczynski, 1980 ) on the simulation of
interactions between political, administrative and trade-union structures in the communist
totalitarian system. For other approach to terrorism simulation consult Smith, 2001,2002.

As for the recent war on terror, it is a real, serious war. Serious means that the aim of at
least one of the parts is extermination, not domination or territorial expansion. Note that the
conventional (rather medieval) concept of a "battlefield" has no sense in this war. The war
is distributed over the whole globe and the structures involved are always global.

Modeling and simulation of soft systems where the human factor is crucial to the system
behavior is a difficult and risky task. The models can hardly be validated, unless we have
sufficient historical data. Even if we can prove that the model behavior for the past data is
correct, this does not mean that it will be valid for the future trajectories. So, it is a common
practice to use invalid models of soft systems (in fact, the authors of such models normally
do not care about the model validity). This does not mean that the model, even if it is
invalid, cannot provide relevant and interesting information. What we should take into
account is that the numeric results can be false. The model user must be aware of this, and
rather look for a behavioral properties that repeat for different data sets and model
parameters, using his experience in intuition.
Our model is agent-oriented. Let recall what object- and agent-oriented simulation is. In the
object-oriented simulation the model components are represented by objects in the machine
memory. An object is a set formed by data structures and methods. The data structures are
defined due to the programming language actually used. For example, if the object is a car,
the data may contain the car type, model, year of production, weight, motor type, color etc.
The methods associated with the object are procedures related to this particular data set and
to the rules of the car behavior, for example the whole set of equations of the car movement,
procedures that display the required data, or even an interface with virtual reality tools that
can show the animated car image.
In agent-oriented simulation the model components are also objects, called agents. But,
unlike conventional objects, the agents are equipped with some kind of intelligence. They
can take decisions, communicate with other agents and negotiate, if their local goals are
different and need some compromise to be taken. In object- and agent-oriented simulation
languages the user defines one or more agent (object) types. These are type declarations,
not agents. The agents are being created during the program run. The resulting agents are
called instants of the corresponding type. After being created, the agent may be activated,
and executes its methods (also called events in the simulation terminology).
Our model is open; this means it can be easily modified and expanded. Here we present the
actual version, which is merely the kernel of possible future developments. ("Structure" and
"organization" are synonyms in the following).
All structures in this model are hierarchical. Each structure has a leader (head) and a
hierarchical "tree" of subordinates. Three kinds of structures can be created.
Terrorist structures, anti-terror structure and terror-supporting structure.
Terrorist structures (may be one or more) are those whose members are able to carry out a
terrorist act.
The members of the anti-terror structure are able to neutralize members of a terrorist
structure, destroying a part (one or more branches) of the terrorist structure.
The terror-supporting structure is not a terrorist one, and it does not participate in the terror
acts directly. This structure looks for a possible links with terrorist structures. If such a link

is established, the structure which is linked to receives a support. This means its power (this
is explained further on) grows and the ability to carry out a terrorist act also grows.
The main component is an agent. It is an entity that can move over a plain region. The
agent has a type attribute.
Agent types are as follows.
Type 1 - a "neutral" object. It only moves, but do not enter in any social structure. In fact,
the neutral agents are not necessary in the actual model, because they do not influence other
agents or structures. However, they exist to simulate a non-engaged population. In the
future model expansion they can be equipped with new behavioral properties. In the actual
version, a "terrorist" agent can be neutralized. As in the model there is no birth-and-death
process, a neutralized terrorist does not disappear, but it become a neutral object.
Type 2 - a potential terrorist. This agent can become a member of a terrorist organization.
Type 3 - a potential anti-terror agent. It can become a member of an anti-terror organization.
Type 4 - a terror-supporting agent. It can become a member of a terror-supporting
In this model one anti-terror, one terror-supporting and multiple terror structures are
The agents move inside a rectangular region. The movements are random, but there exist
some attracting spots in the region (cities). The attracting forces are strong enough to make
the agents to concentrate, after some model time, around the cities (five cities are created in
this model).
Structure creation is spontaneous, and it is a result of the rules of agent behavior. If an
agent is of type 2 (potential terrorist), then it permanently looks for other agents of the same
type to capture them as its subordinates, until it has enough subordinates (the limit of 10 is
fixed). For each of the subordinates, the agent, who has captured it, becomes its superior.
This forms a hierarchical, tree-like structure. If an agent has one or more subordinates and
no superiors, he is the head (leader) of the organization. But, even being the structure head,
it can be captured as a subordinate by a member or leader of other terror structure. This
makes the structures to merge in larger organizations. Any capturing (linking) event may
occur when the two involved agents are close enough (the “contact” distance). Each
structure has a power parameter. In the actual model the power is equal to the number of
the structure members. However, it can be increased if there exist a link between the terror
structure the terror-supporting structure. The power of the structure influences the ability of
capturing new members, and the ability to carry out a terror act.
Structure interactions
As stated before, the interaction between terror structures only consists in merging. The
interaction between anti-terror and terror structures is as follows. The members of the anti-
terror structure have no information about terror structures, until a terror structure is
infiltrated by anti-terror agents, or some of the members of a terror structure collaborate
with anti-terror agents. The infiltration event occurs, when an anti-terror agent is captured
as a subordinate member of a terror structure. There is a small probability that this occurs,
namely when the capturing agent recognizes by mistake an anti-terror agent as a potential
terrorist. On the other hand, any member of a terror structure can become a collaborator
(this does not mean that it becomes a member of the anti-terror organization). If there exist
collaborators or infiltrated agents, the terror structure may be attacked. The attack does not
mean the structure destruction. It consists in a “local” destruction; the immediate superior
of the infiltrated member and all its subordinates are neutralized. This means that their type

is changed to type 1 (neutral agent) and all their links in the structure disappear. If the
structure is small, this may result in structure destruction. For bigger structures this causes a
considerable damage to the structure, that normally brakes into two ore more unlinked
The terror-supporting structure looks for possible links between its members and members
of terror structures. If a link is established, the power of the linked terror structure grows.
The terror-supporting structure cannot be attacked or destroyed.
All the above actions, like member capturing, collaborating, infiltrating, linking and others,
depend of the corresponding necessary conditions and on a random factor. This random
factor, that influences the probability of the event, is controlled by a corresponding model
parameter. This allows disabling some events or carrying out simulation experiments with
different levels of the probability factors.
Finally, terror acts are simulated. However, these events are not the most important ones in
our model. No explosions, casualties or deaths are simulated. If a terror act occurs, it is only
indicated graphically on the screen and counted. Note that the main purpose of the present
simulations are the interactions between social structures involved, and not the terror acts.
In the future developments the terror acts will be simulated with more details, taking into
account not only their probabilities, but also magnitude. This can be related to availability
of the weapons of mass destruction and the advances of mass destruction technology, which
sooner or later will be available for everyone, including terrorists. Some more details on the
simulated model events are given in the following section.

All the mechanisms included in our model and described above in general terms can be
implemented in any simulation language that supports object-oriented discrete event
simulation. The present implementation has been done using the PASION (PAScal
simulatION) language. To learn more about PASION consult . See also my previous articles (Raczynski …).
Recall that PASION program consists of a sequence of process declarations. Each process
contains a set of events. This process/event structure is ideal to describe the behavior of
model agents. The main process named AGENT defines the behavior of our agents. In the
main program segment several hundreds of agents are created as instants of the AGENT
process. Each agent is given attributes. They are as follows.
Agent type (1,2,3 or 4) for neutral, potential terrorist, anti-terrorist, terror-supporting,
Position and velocity. Those are agent coordinates and its actual velocity in the region of
H – the time-step for the MOVE event.
Colab – a logical attribute, set true if the agent is a collaborator (only for agent type 2).
Inf – a logical attribute, set true if the agent is an anti-terror one, infiltrate to a terror
Sup – the pointers to the superior agents; may be in the organization number 1,2 and 3
(terror, anti-terror and terror-supporting, respectively). As an infiltrated agent may have two
superiors, this is an array. The array index tells to which structure the superior belongs.
Note that an anti-terror agent can infiltrate to one terror structure only.
Sub - – the pointers to the subordinate agents, may be in the organization number 1,2 and 3
(terror, anti-terror and terror-supporting, respectively). As an infiltrated agent may have

subordinates both in terror and anti-terror structure, this is an array with dimensions 3x10
(first index being the organization index, the other identifies the subordinate agent).
Top – this identifies the leader of the structure to which the agent belongs. It is an array
with one index (organization number).

Initially, all the links are cleared, so no structures exist in the set of agents. Once activated,
the agents begin to move and look for other agents to link them to, and executing other
The AGENT process has the following events.
MOVE – the agent movement. This event also makes an agent of type 2 (a terrorist)
collaborate with anti-terror structures. This occurs with some given probability. If the agent
is already collaborating, then it can invoke the DESTROY event.
TERR1 – looking for subordinates for agents of type 2. This creates terror structures
TERRACT – committing a terror act
DESTROY – this neutralizes a terror agent and destroys all its links
ANTITERROR – creates links for agents of type 3, creating the anti-terror structure
TERRORSUPPORT - creates links for agents of type 4, creating the terror-supporting

There is also other process named WORLD, with only one instant that manages the
graphical display of the model state, recalculates the power of each organization and
gathers the model statistics. The WORLD process also creates the links between terror-
supporting and terror structures and actualizes the TOP attribute of all agents.

                                              The program generates an animated image
                                              of the actual situation, showing the
                                              movements of all agents and the structure
                                              of each organization. The graphical
                                              symbols are shown on figure 1. The links
                                              that form hierarchical structures are shown
                                              as lines that connect the agents. The last
                                              item on figure 1 shows the event of
                                              destroying a part of a terror organization.
                                              As stated before, we do not simulate any
                                              birth-and-death process. The agent
                                              population is fixed, whit given percentage
                                              of potential terrorists, anti-terror agents and
                                              terror-supporting agents. The main rule is
                                              that an agent cannot take any action if it is
                                              not a member of one of the organizational
                                              It should be noted that the model is open to
                                              any changes and extensions. The actually
  Figure 1. Graphical symbols                 included mechanisms should be treated as a
                                              basic and simple rules needed on the very
begiining of the research. Such tools as PASION permit to easily add new events or
processes to the model.

The structures are created due to the structure creating events in the agent process. Each
agent is permanently looking for other ones that can be captured as its sub-ordinate. The
agent being captured cannot have already been captured by other agent (it cannot have the
"superior" in the corresponding structure). However the agent that has one or more
subordinates and no superiors (a leader) can be captured as a subordinate. This way the
structures grow. As mentioned before, we can have only one anti-terror structure and on
terror-supporting structure. However, multiple terror organizations can exist. The relation
between a terror and terror-supporting organization is defined by the links between
members of the two organizations. If such links exist, then the linked terror organization
receives an additional power from the power-supporting organization. This power
parameter influences the ability to commit a terror act and to capture new members. So,
more powerful organizations grow faster.
The struggle between anti-terror and terror organizations consists in destroying a part of a
terror organization, erasing the corresponding links. The unlinked agents become "neutral"
(recall that no killings or deaths are simulated). A neutral agent does nothing. However,
there is a probability that such agent can become a terrorist again and incorporates to other
terror organization. The event of destroying terror links can be done only if there exist anti-
terror agents infiltrated into a terror organization, or there exist a collaborating terrorist. In
other words, this activity is possible only if there is some information about a terror
organization available by means of intelligence activities. If a terror structure is being
attacked, it does not disappear. Only some links, linked to collaborator or infiltrated agent
are destroyed, and the linked agents neutralized. In our model neither the anti-terror
organization nor the terror-supporting structure can be attacked. The attacked structure may
result to be damaged. This consists in erasing several links (not all the structure). This
erasing process starts with the superior of the infiltrated or collaborating agent and goes
down to the last (lower) level in the structure. However, the attempt to damage the structure
may fail. The probability of the failure depends on the power of the attacked structure. Note
that the power of a terrorist organization is a function of the size of the organization and on
the number of links to the terror-supporting structure, if such has been created. There is no
room here to give the detailed description of the entire algorithm, so only relevant and
qualitative characteristics are indicated. The whole PASION code contains about 800 lines
of code. After translation this expands to more than 1400 lines of Delphi code that uses
some PASION units giving total of several thousand code lines.


In all experiments the model was completely abstract, without using any real data. So, no
quantitative conclusions can be taken from the results. However, changing the model
parameters one can observe interesting behavioral model properties, and carry out many
“what-if” experiments.
   The experiments have been done with total of 500 agents, approximately one third of
them being potential terrorists, one third anti-terror and one third potential anti-terror agents.
The whole model is controlled by a number of parameters. The model is controlled by a
series of parameters, most of them fixed in the code. The parameters define the model
behavior, mainly by means of probabilities of the occurrence of certain events. In the
following, the parameters named “rate” are not exactly the rates or probabilities, but they
can control the corresponding probabilities. In general, if a parameter is set equal to zero,

the corresponding event or model behavior is being disabled. The parameters that can be
defined by the user at run time are as follows.
Terror organization creation rate – controls the ability to create terror organization
Anti-terror organization creation rate – controls the ability to create anti-terror organization
Infiltration rate – controls the probability that an anti-terror agent becomes a part of terror
Collaboration rate – controls the probability that a member of a terror structure
collaborates with the anti-terror organization
Terror-supporting creation rate – controls the ability to create a terror-supporting
Linking terror-supporting with terror organization – controls the probability of creating
the links between terror-supporting and terror structure.
Terror act rate - controls the probability of committing a terror act by a terror organization
Rate of returning to terror activity – controls the probability that a neutralized terrorist
become a terrorist again.
Even such limited number of interaction rules and controlling parameters allows carrying
our lot of simulation experiments. Firs of all, we can investigate how each parameter
influence the model behavior. Changing the rules of interaction (this needs an intervention
in the model code) one can see how relevant certain model events and what the result of the
changes is. What we show here is merely a small example of a simulation experiment.
The experiment consists in enabling and disabling the actions of the terror-supporting
structure. Disabling can be done by simply setting the “Terror-supporting creation rate”
equal to zero.
Our simulations start with all the structures cleared. This is not a realistic situation.
However, this can be interpreted as if we start with an irrelevant terrorism problem, and
then the problem suddenly appears. Anyway, other experiments can be carried out,
skipping the early “warm-up” period and looking for the model behavior for greater
simulation time. As stated before, the model is rather abstract, so the time unit is not
defined. In this program the time interval for the agent movements was set equal to 7 that
can be interpreted as 7 days, if one wants to locate the events in real time. The final
simulation time was equal to 3650 time units.

          Figure 2. An screen shot of an early stage of the model run

Let us start with the terror-supporting structure enabled. Figure 2 shows an early stage of
structure creation. There are several small terror organizations, one small anti-terror and
one terror-supporting organization. All the agents move on the screen and the structure
links are being created. On figure 3 we can see the event of destroying a part of a terror
structure. Bold red lines are the links being destroyed. Figure 4 shows the average
trajectory for the number of active terrorists, taken from 50 repetitions of the simulation run.
By an active terrorist we understand a terrorist who belongs to a terror organization. The

   Figure 3. Destroying a terror organization. Bold line shows links
   being destroyed.

shadowed area is the region where the number may be situated, with probability equal to
0.9. Figure 5 depicts the same trajectory with the terror-supporting organization disabled.
What is interesting, that these curves nearly always have a maximum. After reaching the
maximum, the number of terrorists decays. The numerical results were as follows.
Terror-supporting enabled: maximum terrorists 72, final 36, total terror acts 31
Terror-supporting disabled: maximum terrorists 70, final 10, total terror acts 6
But these are only the average values. We must remember that the average value taken
from simulation experiments with random factors is not very informative. Note that, for
example, for the trajectory of the number of terror acts (terror-supporting enabled), the
confidence interval with probability 0.9 is between 6 and 50 (see figure 7), with the average
of 31.

   Figure 4. Average and confidence intervals for the number of active terrorists.
   Terror-supporting structure enabled.
Figure 6 is a 3-dimensional image (a standard PASION output) of the probability density
function for the number of accumulated terror acts. The horizontal axes are the time and the
number of terror acts, and the vertical axis is the probability density function.

   Figure 5. Average and confidence intervals for the number of active terrorists.
   Terror-supporting structure disabled.

Figure 7 shows the same trajectory as a 2D plot, indicating the limits of the confidence
intervals for probability level 0.9 (shadowed region).

Figure 6. 3D image of probability density function. Terror-supporting structure

Figure 5. Average and confidence intervals for the number of terror acts. Terror-
supporting structure enabled.

The presented simulation is a first stage of research that should be continued and may result
in much more detailed and realistic approach to the terror dynamics modeling. Perhaps
similar models can be developed using the System Dynamics approach, with more global
view. But, using System Dynamics we will not be able to see the structure creation and the
agent movements. Note that this geometrical approach, including attraction centers (cities)
can be very important and can hardly be introduced into SD models. Our experiments have
been done on a PC. Obviously, using multi-processing and distributed simulation one can
deal with similar models hundreds of times faster and greater.
As PASION models are open for changes, we can easily add new processes and events. In
particular, to simulate the model dynamics for longer model time intervals a birth-and-
death process should be added to renew the involved human resources. Also a process of
improving and proliferation of the weapons of mass destruction should be added to control
the impact of terror acts.


Raczynski, S, 1980, Simulation of interaction between certain hierarchical structures,
Proceedings of the conference "Simulation of Large Systems", Universitat Bielefeld,
Bielefeld, 1980.

Smith, Roger, 2002, Counter Terrorism Simulation: A New Breed fo Federqation,
Simulation Interoperability Workshop, Orlando, Florida, Spring 2002

Smith, Roger, 2001, Modeling and Simulation Adds Insight on Terrorism, Signal
Magazine, December 2001. Armed forces Communications and Electronics Association

PASION system home page: http//


To top