Doctrine-Based Multi-Resolution Conversion for Distributed

Agent-Based Simulations

Raul Ceretta Nunes

1 a

, Guilherme Miollo

, Edison Pignaton de Freitas

2 b

and Luis Alvaro Lima Silva

1 c

Graduate Program in Computer Science, Federal University of Santa Maria, Santa Maria, Brazil

Institute of Informatics, Federal University of Rio Grande do Sul, Porto Alegre, Brazil

Keywords:

Distributed Simulation, Simulation Integration, Multi-Resolution Simulation, Agent-Based Simulation.

Abstract:

The treatment of different levels of simulation detail is a relevant component for integrating simulators in var-

ious application domains. These levels vary in federated multi-resolution simulations, from low-information

(low-resolution) to high-information (high-resolution). One of the challenges in this integration is the repre-

sentation and conversion of simulation data exchanged between the simulators. This work explores the use

of doctrine-based rules in the conversion to ensure correct simulation integration. These rules contain infor-

mation on how the multi-resolution conversion handlers should operate. To avoid abrupt changes from one

doctrine rule to another, this work also extends a doctrine description language to capture information for the

smooth transition between these rules. Experimental results demonstrate that it is possible to achieve simula-

tions that ﬂexibly deal with the required dynamism of a multi-resolution simulation environment.

1 INTRODUCTION

The interoperability between different resolution sim-

ulators is a fundamental issue for distributed simu-

lation applications. Such simulation systems’ inter-

operability is usually based on the exploration of the

High-Level Architecture (HLA) (IEEE, 2010), where

the integrated simulators are built to meet different

training objectives (Falcone et al., 2017). There are

simulators for operational training with high informa-

tion granularity models. For example, in (Zhou et al.,

2019), the goal is to allow trainees to perform detailed

operational activities in virtual environments. In con-

trast, simulation systems can focus on the execution

of more abstract simulation-based training activities

with low information granularity levels, such as the

ones required for strategic and tactical training. In

(Pozzer et al., 2022), for example, operational aspects

of the simulated problem are abstracted because they

have limited relevance for tactical and strategic train-

ing. The differences in these simulation goals most

often observed when simulation systems with differ-

ent resolution levels are used in the same integrated

https://orcid.org/0000-0003-3228-4071

https://orcid.org/0000-0003-4655-8889

https://orcid.org/0000-0002-6025-5270

simulation setup, each implemented according to dis-

tinct information levels of detail. The problem with

this approach is that simulated exercises involving

simulators with different resolution levels will likely

require customized integration and synchronization

methods, and the standard HLA architecture only su-

perﬁcially/partially answers this problem.

The works presented in (Kong and Xing, 2013)

and (Tolk, 2012) investigate multi-resolution con-

version solutions where the proposals are indepen-

dent of the communication architecture in the dis-

tributed simulation. From (Kong and Xing, 2013), a

multi-resolution modeling can be built from a multi-

resolution entity (MRE) method that maintains run-

ning information of different resolution models. The

challenge of MRE is to maintain consistency of the

interaction with different resolution models. The

work in (Tolk, 2012) details a solution for multi-

resolution conversion where agent-oriented aggrega-

tion/disaggregation methods are investigated. The

authors in (Paul et al., 2017) also use aggrega-

tion/disaggregation methods but propose exploring

user-written doctrine rules to detail the required infor-

mation for multi-resolution conversions. These works

are based on multi-resolution conversion handlers,

which deal with the required conversion ﬂexibility

for distributed simulation architectures. However, the

Nunes, R., Miollo, G., Pignaton de Freitas, E. and Silva, L.

Doctrine-Based Multi-Resolution Conversion for Distributed Agent-Based Simulations.

DOI: 10.5220/0012124700003546

In Proceedings of the 13th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH 2023), pages 377-384

ISBN: 978-989-758-668-2; ISSN: 2184-2841

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

377

literature still weakly addresses the dynamic multi-

resolution conversions and the consequent multi-

resolution mappings between low and high-resolution

multi-agent simulation systems (Macal, 2016).

The main contribution of this work is to pro-

pose a solution to multi-resolution conversion prob-

lems. This solution is based on doctrine rules

that specify how to make customized agent aggre-

gation/disaggregation conversions between low and

high-resolution agent-based simulators. These con-

verters receive low-granularity information from a

low-resolution level simulator (e.g., containing ag-

gregate units such as battalions) and convert them to

a high-granularity simulator (e.g., convert an aggre-

gate unit into its associated physical entities). This

work also proposes the exploration of dynamic multi-

resolution conversion handlers, providing a solution

to smooth the simulation changes between discrete

states, allowing these changes to be realistically de-

veloped in simulators with different resolution levels.

2 A DOCTRINE-BASED

MULTI-RESOLUTION

CONVERSION SOLUTION

The work in (Pozzer et al., 2022) presents a con-

crete instance of integrated multi-resolution simula-

tion. The SIS-ASTROS simulator develops a low-

resolution simulation for recognition, choice, and oc-

cupation of tactical battery positions to promote tacti-

cal training. These virtual tactical activities are simu-

lated in a high level of detail in this simulation system.

However, the SIS-ASTROS simulator is a compo-

nent of an Integrated Simulation System, where con-

structive, virtual tactical, and virtual technical (opera-

tional) simulators are interconnected to build a virtual

real-world setting for the augmented training of per-

sonnel with distinct military echelons.

A common simulation scenario in the proposed

SIS-ASTROS Integrated Simulation System occurs

when the constructive simulator, for higher military

echelons, controls agents inserted in the virtual tacti-

cal exercises. The multi-resolution conversion issues

appear because the simulated agents (or aggregates)

do not need to perform detailed tactical behaviors in

the constructive simulations. When converting the

simulation information from the constructive simula-

tor to the virtual tactical, the constructive view only

provides broad and discretized information to con-

trol the agents mapped in the other simulation sys-

tem. As the constructive simulator keeps the property

of these mapped agents, it cannot deal with the differ-

ent agent behaviors that occur when these entities fol-

low alternative simulation doctrines. These doctrines

guide the execution of the required agent behaviors

in the various tactical positions of the simulated exer-

cises. However, the agent’s doctrine behavior changes

throughout the simulations are important elements for

the ﬂuency and realism of the more detailed virtual

tactical simulations.

The solution for distributed multi-resolution sim-

ulation proposed in this article considers that at least

two simulators are integrated, low and high-resolution

ones. Figure 1 shows that unit X is aggregated in

the low-resolution simulator. The division of unit X

into its individual elements is performed according

to the speciﬁed disaggregation rule. Unit X is de-

tailed as circularly dispersed Y entities on the high-

resolution simulator’s terrain. Figure 1 also shows

that the high-resolution simulator represents terrain

elements in higher levels of detail (e.g., various veg-

etation types). These elements are taken into account

at the time the disaggregation occurs.

Figure 1: The disaggregation of simulated units from the

low to the high-resolution simulator: the unity X is split

into several entities Y.

In addition to the terrain differences between the

involved simulators, the used military doctrines must

be respected during the resolution exchange. As

shown in Figure 1, the doctrine deﬁnes that the unit X

disaggregation corresponds to ﬁve entities Y. The doc-

trine also deﬁnes that a circular displacement should

be used when these Y entities are placed in the terrain

managed by the high-resolution simulator.

Even if the simulation information is exchanged

between simulators to ensure the correct resolution

conversion, the problem of integrating the simulators

when the simulation situation changes persist. In the

high-resolution simulator, the Y entities must mod-

ify the situation at the simulation run time if the unit

moves to a new area. This exchange also involves

SIMULTECH 2023 - 13th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

378

altering the doctrines that guide the simulation of Y

entities. Although this exchange of doctrines to guide

the Y entities’ behaviors in the high-resolution simu-

lator are relevant to produce faithful simulations, this

exchange of doctrines presents an irrelevant level of

detail to the low-resolution simulations.

An example of doctrine changes appearing in sim-

ulation exercises is presented in Figure 2, which

shows that Y entities in the high-resolution simula-

tor must move to goal A. As they are still under the

control of the low-resolution simulator, these enti-

ties must perform the navigation behavior actions re-

quired by the X unit in the low-resolution simula-

tor. To arrive at A, this unit follows a planned route

(dashed line in Figure 2). However, according to the

simulated doctrine in the low-resolution simulator, the

movement of Y entities must follow a convoy forma-

tion along the dotted line path in Figure 2. When they

arrive at goal A, their arrangement on that terrain lo-

cation can also differ from the one used in the tacti-

cal position initially occupied by these entities. The

result is the need to dynamically deal with the fol-

lowing simulation situation changes: the ﬁrst occurs

when the simulated entities leave the current tactical

position, and the second when they arrive at a destina-

tion tactical position (goal A), in addition to executing

doctrine-based convoy movement behaviors.

In summary, multi-resolution conversion depends

on two issues: 1) the representation of additional in-

formation to regulate the simulation differences be-

tween the low- and high-resolution simulators; and 2)

the mapping of simulation changes that occur at sim-

ulation run-time. This work proposes an XML-based

representation language to express doctrine rules for

mapping low and high-resolution distributed simula-

tions to approach the ﬁrst issue. It relies on dynamic

multi-resolution conversion handlers to manage the

simulation changes of the integrated agents to ap-

proach the second issue. Therefore, the research re-

lies on the representation of doctrine rules to better

specify multi-resolution conversions.

2.1 Multi-Resolution

Aggregation/Disaggregation

Conversion Rules

The multi-resolution conversion solution adopted in

this work is based on previous advances found in the

literature (Li et al., 2008) and (Paul et al., 2017). It ex-

plores the representation and computation of doctrine

rules to guide the distributed simulations. The pro-

posed solution extends the expressiveness of the doc-

trine rule description language detailed in (Paul et al.,

2017). The multi-resolution doctrine rules aim to pro-

Figure 2: Low and high-resolution simulators showing dif-

ferent views of a simulated agent movement situation.

vide the information necessary for the high-resolution

simulator to faithfully simulate real-world agent be-

haviors, where the simulated entities and their simu-

lated behaviors are controlled by the low-resolution

simulator. Although the low-resolution simulator

owns these agents, and this ownership remains so

during the course of the integrated simulations, these

agents make a fundamental part of the simulations ex-

ecuted by the high-resolution simulator.

When implementing a new multi-resolution doc-

trine rule language, rule modeling concepts must be

deﬁned. In our work, these concepts were motivated

by a case study developed in the SIS-ASTROS GMF

project (Pozzer et al., 2022). One of the aims of this

project is to integrate a high-resolution virtual tacti-

cal simulator with a constructive simulator while ex-

ecuting distributed simulation exercises. The inte-

gration rules rely on deﬁning these domain concepts

to detail different simulation situations. In this case

study, the main attributes for specifying the multi-

resolution conversion rules are the following: Object

identiﬁcation; Unity composition; Simulation situa-

tion; Formation organization; Agent position; Dis-

tance between agents; Distance between agents and

terrain features; Central agent; and Agent orientation.

DoctrineRules: Fig. 3 shows that DoctrineRules is

the root concept in the rule speciﬁcation model. This

concept is detailed by Entities and Units, organizing

these elements in the doctrine rule representation.

Entities: Figure 4 shows that Entities are composed

of one or more Entity elements. The attributes of

these concepts are:

• name: identiﬁes the simulated object;

Doctrine-Based Multi-Resolution Conversion for Distributed Agent-Based Simulations

379

Figure 3: The representation of doctrine rules.

• FOMClass: describes the object type;

• enumeration: a unique numerical attribute for rep-

resenting the simulated doctrines.

Figure 4: The Entity concept and its attributes.

Units: Figure 5 shows that the Units concept con-

sists of one or more Unit elements. Each Unit is mod-

eled by Composition and Behavior elements. The

Unit attributes are the same of the Entity whenever

the object is an entity and has the same purpose.

Figure 5: The Units concept, its subconcepts and attributes.

Composition: Figure 6 shows that the Composi-

tion concept indicates which entities, among those de-

clared, are part of a Unit. It consists of one or more

Entity elements. Its attributes are:

• type: identiﬁes the entity type; it should corre-

spond to a previous deﬁned name of an Entity;

• id: a unique numerical ID attribute identifying an

Entity of a Unit. When multiple entity types exist,

this attribute distinguishes them.

Figure 6: The Composition concept, its subconcepts.

Behavior: The Behavior concept indicates how

units should behave in each simulated doctrine. This

concept consists of one or more Situation elements

and assumes an entity corresponds to a simulation

agent. Its attributes are:

• name: identiﬁes the doctrine situation;

• formation: formation organization adopted by the

Entities of a Unit in the simulated situation;

• entityDist: if necessary, indicates the distance be-

tween entities in the simulated situation;

• terrainFeatureDist: indicates the distance the enti-

ties should keep from the terrain features observed

in the simulated situation;

• entityRef: indicates the coordinate of the refer-

ence entity, which is used as reference for or-

ganizing other entities in the simulated terrain.

This reference should be the related Unit on low-

resolution simulator.

The Entity concept associated with the Situation con-

cept is used to describe information that is speciﬁc

to the entity within the simulated situation. The at-

tributes that describe these concepts are:

• id: a unique numerical attribute identifying an En-

tity in the simulated situation. This numerical id is

related to elements of the Composition concepts;

• position: indicated the terrain position the Entity

is located in the simulated situation (and the ter-

rain formation organization adopted by them);

• orientation: if necessary, it indicates the agent ori-

entation in the simulated situation, e.g., North,

South, etc.

2.2 The Representation of Dynamic

Multi-Resolution Conversion Rules

The dynamic multi-resolution conversion rules are

related to agent behavior changes occurring at simu-

lation run-time. The dynamism concerns the smooth-

ing that must occur whenever the doctrine rules guid-

ing the agents’ simulations change. To approach this

problem, this work proposes determining situation

change zones to address these events. When instan-

tiating these zones in a simulated terrain, the deter-

mined zones indicate the areas where the simulated

entity behaviors change from one doctrine to another.

In this doctrine-based multi-resolution conversion

solution the situation change zones correspond to

doctrine transition zones that are theoretical areas

mapped into the simulated terrain at the simulation

run-time. Figure 7 illustrates this zone from a park-

ing situation to a movement situation in response to a

movement order. When low and high-resolution sim-

ulators are integrated, these zones are created when-

ever agent movement orders are issued in the low-

SIMULTECH 2023 - 13th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

380

resolution simulator. When this order is executed, the

agent moves from one terrain location to another.

Figure 7: A situation change zone where occur a parking to

movement doctrine transition.

Fig. 8 illustrates a unit movement in a simulation

in response to a movement order in the low-resolution

simulator. When the movement command is emitted,

the ﬁrst terrain zone is automatically determined in

the high-resolution simulator. In the start zone, the

disaggregated entities are under the doctrine rule be-

havior that must be used in that simulated tactical po-

sition (parking). When moving to leave this position,

the unit executes a dynamic convoy movement behav-

ior to change from parking to movement doctrine rule

behavior. Consequently, the agents start following a

different doctrine rule, a doctrine transition rule that

activates an autonomous behavior. In the end zone,

a similar behavior appears. The entities must change

the executed doctrine rule (movement to parking).

With the deﬁnition of the simulation situation

transition zones mapped to the simulated terrain, the

proposed dynamic multi-resolution conversion fol-

lows the activities presented in Figure 9. Notice that

the high-resolution simulator is responsible for exe-

cuting the movement behaviors of the simulated unit.

The low-resolution simulator controls the agents’

movement positions in this distributed simulation sce-

nario. When the unit moves in the low-resolution

simulator, all entities mapped into the high-resolution

simulator must move too. The low-resolution simu-

lator also determines the target position. Thus, the

Figure 8: Start and end movement zones determination and

their corresponding behaviors.

unit’s movement should be mapped from the low to

the high-resolution simulator. To approach this prob-

lem, the units’ terrain coordinates are continuously

updated and mapped from one simulator to another.

According to the doctrine rules used by the high-

resolution simulator, the unit terrain coordinates are

used to determine whether any of the simulated units

entered a situation transition zone.

The low- and high-resolution simulator agent

movement algorithms are executed when agents are

within the simulation situation change zone. How-

ever, they cooperate to achieve the required realistic

movement behavior in the high-resolution simulator.

Then the multi-resolution doctrine conversion rules

proposed in this work detail which terrain positions

the units should be. When the simulated units en-

ter the change zone determined in the simulated ter-

rain, a doctrine rule relevant to that simulation situ-

ation change is selected, determining the new simu-

lated doctrine these agents should start executing.

3 A CASE STUDY

A case study was developed to assess the ﬂexibility

and dynamism of the proposed multi-resolution con-

version solution. This case study involves integrating

simulated exercises developed in two distinct simu-

lators: a low and a high-resolution simulator. Impor-

tantly, the low-resolution simulator contains a simula-

tion scenario that should be realistically mapped into

the high-resolution simulator. To do so, implemen-

tations of multi-resolution conversion handlers are

used so that agents’ aggregations are converted (via

the RTI structure) from the low- to the high-resolution

simulator, reaching the high-resolution simulator in a

proper doctrine-based disaggregated format.

In the developed case study, the low-resolution

simulator contains an aggregation of vehicles com-

prising a military battery. This aggregation in-

volves multiple instances of three types of elements

(i.e., military vehicles) named in this case study as

Unid1, Unid2, and Unid3. Unlike the low-resolution

simulator, the high-resolution one must simulate

these entities individually. The implemented multi-

resolution conversion handlers execute the speciﬁed

multi-resolution doctrine rules to properly split the

aggregated objects into their individual components.

When the units are disaggregated and represented

in the high-resolution simulator, they must occupy the

same tactical position as in the low-resolution simu-

lated terrain. In this tactical position, agents must be

displaced in the terrain according to a particular static

formation organization detailed by the military doc-

Doctrine-Based Multi-Resolution Conversion for Distributed Agent-Based Simulations

381

Figure 9: Activity diagram explored dynamic multi-resolution conversion handlers.

trine. Therefore, the different disaggregated units in-

volved in this case study, named U1EntA, U1EntB,

U1EntC, U2EntA, U2EntB, and U3EntA, are posi-

tioned on the simulated terrain accordingly.

When agents’ aggregations receive a command

to occupy a new tactical position at run-time, these

agents must perform a doctrine-based motion behav-

ior. The transition from a static position occupation

doctrine to a dynamic movement one is simpliﬁed

in the low-resolution simulator. This simulator just

needs to simulate an aggregate of agents (i.e. the

whole battery) moving from one place to another. In

this simulation level, the movement details of each

unit in the aggregated entity are irrelevant.

As the low-resolution simulator controls the ag-

gregate entity, the problem is that the transition be-

tween these static and dynamic doctrines must oc-

cur in an organized and realistic manner in the

high-resolution simulator. So the proposed multi-

resolution conversion handler should detail how to

smoothly move between these doctrines while map-

ping these aggregate entities into individual units in

the high-resolution simulation.

3.1 The Multi-Resolution

Aggregation/Disaggregation

Conversion

All the multi-resolution doctrine rules are checked

during the simulation exercise’s representation. Ini-

tially, declarations of the involved simulation entities

and units are analyzed.

We highlight the doctrine rules’ modeling allows

unique identiﬁers for each entity within the simulated

situations. It also allows handling doctrine rules indi-

vidually. The optional attributes added in the multi-

resolution doctrine rule representation favor captur-

ing the speciﬁcs of simulated agent behaviors. The

following attributes present examples of these simu-

lation situations:

• entityDist: determines the distance between the

vehicles in a convoy movement formation.

• terrainFeatureDist: determines the distance con-

voy vehicles must keep from obstacles and other

simulated terrain features;

• entityRef: describes a central agent (e.g., a com-

mand vehicle) used by simulated doctrines, which

are static and dynamic agent organizations involv-

ing a central element;

• orientation: determines the geographic direction

in which the simulated agents are presented in 3D

terrain scenarios.

The proposed multi-resolution doctrine rules repre-

sent the simulation information using the XML lan-

guage. Starting from the proposals presented in (Paul

et al., 2017), the XML representation allows multi-

resolution conversion handlers to better process the

data required to process the multi-resolution doctrine

rules. As experienced in our project, such XML for-

mat is simple to understand, favoring the doctrine

rules’ modeling, revision, and updating.

3.2 The Dynamic Multi-Resolution

Conversion

In the developed case study, the integration of simu-

lators was tested with and without the use of dynamic

multi-resolution conversion implementations. These

handlers map run-time simulation situations, where

agent behavior information is mapped from the low

to the high-resolution simulator.

The dynamic multi-resolution conversion is

based on the following simulation information: the

convoy formation organization must be adopted by

the simulated agents; there is a simulation command

indicating the agents should act in such a way in the

simulations; the agents are geographically oriented in

the simulated terrain; the agents must consider the ter-

rain elements when executing the simulated doctrine

rules; and the agents must keep minimum and maxi-

mum distances between them.

SIMULTECH 2023 - 13th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

382

In the performed implementations, an Integration

module is used in the our case study. This module

uses a target coordinate, received from low-resolution

simulator via HLA, and calculates the movement pa-

rameters required by the agents of high-resolution

simulator. These parameters are stored in the Vehicle-

Behaviour handler, responsible for moving the agent

in the simulated environment. These parameters are

called steeringParameters (implemented as Steering

Behavior algorithms (Reynolds et al., 1999)). They

are necessary for the agent to transit through the sim-

ulated terrain to destinations avoiding collisions with

static and dynamic obstacles. The algorithm for mov-

ing agents in the high-resolution simulator is called

VehicleBehaviour. The setup parameters for these

modules are the following:

• Move deﬁnes whether the agent should move. If

false, the agent stops the movement;

• MoveForward deﬁnes whether the agent moves

forward or backward;

• DesiredSpeed required movement speed in km/h.

• SteeringCoefﬁcient deﬁned the movement direc-

tion, varying from -1 to 1. Negative values make

the agent turn left, and positives right. If the value

is zero, the agent moves in a straight line.

An usual example of a dynamic simulation situation

occurs when the aggregated Unid1 has to move from

one tactical position to another. The Unid1 vehicles

are in a marching column formation to simulate the

desired doctrine movement behavior. When the vehi-

cles reach the destination, they occupy the new tac-

tical position accordingly. It means these simulated

vehicles are displaced in another terrain formation, a

ready line (side-by-side) formation.

This simulation scenario can be detailed in ﬁve-

time frames (i.e., simulation states). First, the Unid1

vehicles are out of a situation change zone and mov-

ing to it according a doctrine rule that specify column

formation. Second, the Unid1 vehicles arrive in the

situation change zone and getting in it. Third, the ve-

hicles are inside the situation change zone and move

in it. Fourth, the Unid1 vehicles arrive at the situation

change zone border and cross it. Finally, the Unid1

vehicles arrive a new tactical position where the for-

mation doctrine rule speciﬁes an in line formation

(side-by-side). As presented by the high-resolution

simulator, Figure 10 details how these simulated ve-

hicles (agents) behaviors can be mapped from the low

to the high-resolution simulation scenario.

Fig. 10(a) presents a 2D projection of the high-

resolution simulator without executing this dynamic

multi-resolution conversion. Fig. 10(b) presents the

same simulation situation where the high-resolution

simulator uses the simulation information received

from the dynamic multi-resolution conversions.

The four agents are organized in a linear forma-

tion in the ﬁrst simulation time step. They are moving

toward a simulation situation change zone. There is

no difference between the simulated scenarios with

and without the dynamic simulation treatment in this

simulation period of time. The agents adopt a new

formation organization in the second simulation time

step. However, there is a gradual transition from one

formation organization to another, controlled by the

dynamic multi-resolution handlers. It is possible to

observe the agents continue to move in the same ways

when the dynamic simulation conversion is not ex-

ecuting. It means they do not change the doctrine

movement behavior they are executing. The agents

almost reach their destination positions in the third

simulation time step. These positions are located in-

side the tactical position they should occupy. When

the dynamic simulation treatment is not explored, the

simulated agents keep the same movement behavior

even when they almost reach their destination. In the

fourth (and ﬁrst) simulation time steps, the simulated

scenarios with and without dynamic multi-resolution

simulation handlers are identical in low and high-

resolution simulators. However, there is a smoother

and more realistic transition in the executed agents’

navigation behaviors with the execution of the pro-

posed dynamic multi-resolution handlers.

4 FINAL REMARKS

This work proposes a doctrine-based approach to

solve the multi-resolution conversion problem in dis-

tributed simulation systems. To integrate simulators

with different resolution levels, the proposed doc-

trine rule-based solution allows the representation of

a greater variety of scenarios in which the transition

between different simulators is required. This allows

the representation of more speciﬁc simulation situa-

tions involving the transition of military doctrines into

simulated terrain situations with different levels of de-

tail in each integrated simulator. The multi-resolution

solution proposed in this work allows more realistic

integrated simulations, further increasing the role of

doctrine-based rules in providing precise simulation

data for resolution conversions. Directions for future

work are focused on enhancing doctrine-based inte-

gration and, consequently, producing smoother and

more realistic integrated simulations.

Doctrine-Based Multi-Resolution Conversion for Distributed Agent-Based Simulations

383

(a) High-resolution simulator where the dynamic multi-resolution conversion is not executed.

(b) High-resolution simulator where the dynamic multi-resolution conversion is executed.

Figure 10: The dynamic multi-resolution conversion between simulators.

ACKNOWLEDGEMENTS

We thank the Brazilian Army (Strategic Program

ASTROS) for ﬁnancial support through the SIS-

ASTROS GMF project (TED 20-EME-003-00).

REFERENCES

Falcone, A., Garro, A., Anagnostou, A., and Taylor, S. J. E.

(2017). An introduction to developing federations

with the high level architecture (hla). In 2017 Win-

ter Simulation Conference (WSC), pages 617–631.

IEEE (2010). Ieee standard for modeling and simulation

(m&s) high level architecture (hla)– federate inter-

face speciﬁcation. IEEE Std 1516.1-2010 (Revision

of IEEE Std 1516.1-2000), pages 1–378.

Kong, Y. and Xing, L. (2013). Study of multi-resolution

modelling framework in joint operations simulation.

Proceedings of the 2013 International Conference

on Information Tecnology and Computer Application

Engineering (ITCAE 2013).

Li, M., Bi, C., Liu, X., and Niu, D. (2008). Research on

rules of aggregation and disaggregation for aircraft en-

tity models in air force combat simulation. In 2008

Asia Simulation Conference - 7th International Con-

ference on System Simulation and Scientiﬁc Comput-

ing, pages 1286–1290. IEEE.

Macal, C. (2016). Everything you need to know about

agent-based modelling and simulation. Journal of

Simulation, 10:144–156.

Paul, R. L., Nunes, R. C., Oliveira, V. D., and Kunde,

D. (2017). Doctrine based multi-resolution hla dis-

tributed simulation. Proceedings of the Symposium on

Applied Computing, pages 59–64.

Pozzer, C., Martins, J., Fontoura, L., Silva, L., Rutzig, M.,

Nunes, R., and Pignaton de Freitas, E. (2022). Sis-

astros: An integrated simulation system for the ar-

tillery saturation rocket system (astros). In 12th In-

ternational Conference on Simulation and Modeling

Methodologies, Technologies and Applications (SI-

MULTECH 2022), pages 194–201. SCITEPRESS –

Science and Technology Publications, Lda.

Reynolds, C. W. et al. (1999). Steering behaviors for au-

tonomous characters. In Game developers conference,

volume 1999, pages 763–782. Citeseer.

Tolk, A. (2012). Engineering principles of combat model-

ing and distributed simulation. Wiley.

Zhou, L., Mao, Z., Jing, F., Chen, Y., Gao, M., Wang, J.,

and Liu, Z. (2019). Design of satellite communica-

tion simulation training system based on hla, vr and

opnet. In 2019 International Conference on Robots &

Intelligent System (ICRIS), pages 125–128.

SIMULTECH 2023 - 13th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

384