Method and Application of Operational Architecture Modeling Based on Information Flow Analysis

2.1 Analysis Viewpoints of DoDAF V2.0

DoDAF V2.0 mainly deploys around data, models and views, and emphasizes collection, organization and maintenance of data as well as information[4]. As kinds of architecture data selected, views describe collected data and outputting information during developing process, which are displayed visually by control panels, texts, figures, etc. Figure 1 gives all the analysis viewpoints of DoDAF V2.0.

The current researches on architecture modeling based on DoDAF V2.0 describe architectures visually and normatively, mainly from viewpoints of operations, system, data and information. However, most of them analyze operational process, system and information individually, not depicting complex information alternation relationships among real operational processes and implementing units exactly.

Figure 1

Analysis viewpoints of DoDAF V2.0

2.2 Views of Operational Architecture

Referring to multi-view theory of DoDAF V2.0[4], factors, such as operational requirement, system design and technology support, are considered synthetically to divide some parts of DoDAF V2.0 products into information flow view again when building operational architecture, namely, operational architecture view based on information flow analysis. The relationships between information flow view and original views are shown in Figure 2[5, 8]. Operational architecture modeling through information flow view in different battlefield environment, lays an emphasis on macroscopical control of the whole combat systems from viewpoints of operation, technology and system. In addition, it also reflects specific demands of information alternation and system functional connection noticed by different decision-makers especially.

Figure 2

Relationships between different views

2.3 Framework of Operational Architecture Modeling

The framework of operational architecture modeling based on information flow analysis, reflects the combat concept of information dominant and functional pattern of information flow driving materials flow and leading energy flow in systematic operations.

In specific modeling process, information flow is taken as main line and operational task as starting point. Following the order of “task-activity-node-information”, entities, relationships and attributes are defined, and mapping rules and modeling steps are normalized. Entities are decomposed hierarchically. Actions and information alternation process are analyzed accurately. Then, architecture model is built in which information alternation relationships are sorted, information exchanging attributes are explicated and information transferring process and alternation relationships between different activities and nodes are described clearly. Based on the analysis above, the framework of operational architecture modeling based on information flow analysis is given in Figure 3.

Figure 3

Framework of operational architecture modeling

The framework above consists of two parts: Entities and architecture models. Among them, entities are main elements of operational architecture, which provide basic support of various models. Mapping rules offer essential modeling principles.

3.1 Basic Definitions

3.1.1 Entities

According to the framework of operational architecture modeling, entities mainly consist of operational information, nodes, activities and tasks. Because modern warfare is a complex and huge system with typical networked characteristics, complex network theory can be used to define entities.

Definition 1

Entities are objects dealing information which are traced by task demands. They can be expressed as CE = {CE_i|i ∈ [1, m]}, where m is the number of entities[12].

Definition 2

Operational information is effective reflections of different battlefield objects’ characteristics, states, developments and changes which are produced in operational process. They can be expressed as CEinfo={Ii,jw|i∈[1,X0],j∈[1,I0]} where w is the particle size of operational views, X₀ and I₀ are the number of levels divided and information kinds, respectively.

Definition 3

Operational nodes are entities collecting, delivering and dealing data as well as information in operations. They can be expressed as CEnode={Ni,jn|i∈[1,J0],j∈[1,N0]} where n is the particle size of operational views, J₀ and N₀ are the number of divided levels and nodes, respectively[2].

Definition 4

Operational activities are ordered action sets of various nodes for specific tasks. They can be expressed as CEacti={Ai,ja|i∈[1,H0],j∈[1,A0]}, where a is the particle size of operational views, H₀ and A₀ are the number of levels divided and activities, respectively.

Definition 5

Operational tasks are ordered activity sets of various nodes for specific purposes. They can be expressed as CE_task = {T_i|i ∈ [1, T₀]}. As to any task T_i, there is equation T_i = A_{k, 1} ∪ A_k,2 ∪ ⋯ ∪ A_{k, i} where k ∈ [1, H₀], j ∈ [1, A₀].

Entities with certain kinds and limited number are basics of architecture views. Corresponding connection relationships of entities between other architecture views are shown in Figure 4[4, 8].

Figure 4

Entities connections between different architecture views

3.1.2 Relationships and Attributes

Definition 6

Relationships are flow of data or information between different entities. They are usually noted by information requirement lines and expressed as CR = {CR_i|i ∈ [1, n]}, where n is the number of connection relationships.

Definition 7

Attributes are characteristics determined by entities and relationships in information exchanges. They can be expressed as CP = {CP_i|i ∈ [1, t]}, where t is the number of attributes.

Generally, entities don’t exist solely. Besides, they connect each other with specific attributes. Basic grammar rules of architecture modeling are commonly constituted by entities, relationships and their attributes, which are foundations of architecture’s grammar normality and information integrity. The relationships between entities are shown in Figure 5. A group of three-variable relationships is composed of operational activities, nodes and information around a specific task. It can be expressed as CR = ϒ (CE_acti, CE_node, CE_info), which is the key of operational architecture modeling.

Figure 5

Relationships between entities

3.1.3 Mapping Rules

Definition 8

Mapping rules are basic criterions for different entities when generating associated relationships. They include four kinds, namely, activity → activity, activity→node, node→activity and node→node[8–12], and can be expressed as CF = {CF_i|i ∈ [1, p]} where p is the number of rules.

Rule 1 Activity→activity information mapping

Activity→activity information mapping reflects the input, output and control information of an operational activity determining particular alternation information between activities. It is a direct mapping and can be expressed as a three-variable function of original activity, target activity and the information alternated, namely, CF1=ΔCEactio,CEactit,CEinfo.

1. Set view particle size as θ.

2. Select information Ii,jθ and determine the generating and consuming activities are An,pθ and An,cθ, respectively.

3. Take information Ii,jθ as output data constraint of activity An,pq and input or control data constraint of activity An,cθ, respectively.

4. Complete the activityactivity information mapping between An,pθandAn,cθ, and note it as CF₁ : An,pθ→An,cθ.

Then, Rule 1 can be shown as Figure 6.

Figure 6

Activity→activity information mapping

Rule 2 Activity→node information mapping

Activity→node information mapping reflects output pointing nodes of operational activities determining all the nodes involved in activities. It is a composition mapping and can be expressed as a four-variable function of activity, the input, output and control nodes, namely, CF2=ΓCEacti,CEnoder,CEnodez,CEnodec.

1. Set view particle size as ψ.

2. Select activity Ai,jψ, and determine the nodes triggering Ai,jψareNt,1ψ,Nt,2ψ,⋯,Nt,nψ. They can be expressed as Nt,rψ and noted as Nt,rψ=Nt,1ψ∪Nt,2ψ∪⋯∪Nt,nψ.

3. Similarly, the nodes set controlling Ai,jψisNt,zψ and the output nodes set is Nt,cψ.

4. Take nodes set Nt,rψ,Nt,cψandNt,zψ as the input, control and output nodes set of activity Ai,jψ, respectively.

5. Complete the activitynode information mapping of Ai,jψ and note it as CF₂ : Ai,jψ→Nt,rψ∪Nt,zψ∪Nt,cψ.

Then, Rule 2 can be shown as

Figure 7

Activity→node information mapping

Rule 3 Node→activity information mapping

Node→activity information mapping determines the activities with given node involved. It is a composition mapping and can be expressed as a multi-variable function of node and activities it involved, namely, CF3=Π(CEnode,CEacti1,CEacti2,⋯,CEactin).

1. Set view particle size as υ.

2. Select node Ni,jυ and the activities involved are Ah,1υ,Ah,2υ,⋯,Ah,nυ They can be expressed as set Ahυ and noted as Ahυ=Ah,1υ∪Ah,2υ∪⋯∪Ah,nυ.

3. Associate node Ni,jυ with all the elements in activity set Ahυ correspondingly.

4. Complete the nodeactivity information mapping of Ni,jυ and note it as CF₃ : Ni,jϑ→Ah,1ϑ∪Ah,2ϑ∪⋯∪Ah,nϑ.

Then, Rule 3 can be shown as Figure 8.

Figure 8

Node→activity information mapping

Rule 4 Node→node information mapping

Node→node information mapping determines associating relationship and alternation information among nodes. It is a direct mapping and can be expressed as a three-variable function of generating node, consuming node and the information alternated, namely, CF₄ = HCEnodeg,CEnodec,CEinfo.

1. Set view particle size as π.

2. Select information Ii,jπ and determine the generating and consuming nodes are Nm,pπ and Nm,cπ respectively.

3. Protract information requirement line from node Nm,pπtoNm,cπ and correspond it with information Ii,jπ.

4. Complete the node→node information mapping between node Nm,pπandNm,cπ and note it as CF4:Nm,pπ→Nm,cπ.

Then, Rule 4 can be shown as Figure 9.

Figure 9

Node→node information mapping

3.2 Modeling Flow

According to mapping rules, activities and information are hierarchically decomposed to build models of operational activity, node and information alternation, respectively. The specific flow is shown in Figure 10[15].

Figure 10

Basic flow of operational architecture modeling

Step 1 Clarify operational task and decompose operational activities and nodes hierarchically according to the difference of view particle sizes.

Step 2 According to mappings of activity→activity, activity→node and node→activity, determine the information alternated between activities, and sort nodes involved in activities, and clarify activities with nodes involved and build operational activity model combined with hierarchical structures of activities and nodes.

Step 3 According to node→node mapping, determine the information alternated between operational nodes and build operational node model combined with hierarchical structure of nodes.

Step 4 According to models of operational activity and node, build information alternation model.

Step 5 According to three models above, design operational architecture based on information flow analysis.

3.3 Model Construction

3.3.1 Operational Activity Model

Operational activity model reflects hierarchical information alternating situation among operational activities. It can be expressed as a three-variable function of activity, information and node, which can be noted as OAM = Θ ( CE_acti, CE_info, CE_node). When view particle size is set as χ, activities can be divided into ε levels where ε ≤ H₀. Here, activity set is CEacti=An,tχ|n∈[1,ε],t∈[1,A0], information set is CEinfo=Ip,qχ|p∈[1,X0],q∈[1,I0] and node set implementing activities is CEnode=Nf,lχ|f∈[1,H0],l∈[1,N0]. When view particle size is set as δ, activities can be divided into (γ levels where δ < γ and ε < γ ≤ H₀. Here, activity set is CEacti=Ab,cδ|b∈[1,γ],c∈[1,A0], information set is CE_info = Iw,vδ|w∈[1,X0],v∈[1,I0] and node set is CEnode=Nz,xδ|z∈[1,H0],x∈[1,N0]. The operational activity model is shown in Figure 11.

Figure 11

Operational activity model

3.3.2 Operational Node Model

Operational node model reflects hierarchical information requirement relationships among operational nodes. It can be expressed as a two-variable function of node and information, which can be noted as ONM = Φ (CE_node, CE_info). When view particle size is set as α, nodes can be divided into ϕ levels where ϕ ≤ J₀. Here, node set is and information set among nodes is CEnode=Nm,kα|m∈[1,ϕ],k∈[1,N0]andCEinfo=Ii,jα|i∈[1,X0],j∈[1,I0]. When view particle size is set as β, nodes can be divided into φ levels where β ≤ α and ϕ < φ ≤ J₀. Here, node set is CEnode=Ne,sβ|e∈[1,φ],s∈[1,N0] and information set is CE_info = Ig,hβ|g∈[1,X0],h∈[1,I0]. Operational node model Figure 12.

Figure 12

Operational node model

3.3.3 Information Alternation Model

Information alternation model reflects information alternating relationships among nodes in operational process. It can be expressed as a three-variable function of information, activity and node, which can be noted as OIM = Ψ (CE_info, CE_node, CE_acti). When view particle size is set as ε, select information Ii,jε and the generating and consuming node and activity sets are Nm,pε,Nm,cε,Am,pε and respectively. Therefore, the information alternation model can be expressed as Nm,pε|Am,pε→Nm,cε∪Am,cε. Similarly, when view particle size is set as (ξ (ξ < ε), it can be expressed as Nm,pξ|Am,pξ→Nm,cξ∪Am,cξ. The information alternation model is shown in Figure 13.

Figure 13

Operational information alternation model

4.1 Process of Air Assault Operation

Air assault operation is the operational action which attacks enemy targets on the ground or at sea from air. It mainly includes intelligence reconnaissance, electronic interfering and assault operation. In the intelligence reconnaissance, reconnoitering forces, such as radar, electronic confronting and technology reconnoitering, are applied synthetically to collect intelligence about enemy aviation bases, combat plans and air-defense weapons, especially precise locations, shape characteristics, number and air-defense capability of big radars and air-defense missiles positions. In the electronic interfering, enemy electronic equipments are suppressed, interfered and deceived severely. In addition, important radar stations, command and control (C2) systems and electromagnetism targets are attacked by anti-radiation. As the result, a strong interfering region is formed to blind enemy radars and weaken C2 systems and air-defense weapon systems. In the assault operation, under the order and coordination of command post, each operational unit attacks enemy key targets and important positions following scheduled plans, especially enemy air-defense systems and air counterattack systems. During and after the first assault, reconnoitering means, such as satellite, aviation and ground reconnoitering, are used to check out attack effects and enemies’ deployment changes timely. Operational units can adjust subsequent assault targets and methods by analyzing reconnaissance results. The important targets damaged are prior to be attacked continuously and preferentially for thorough ruin and being suppressed long time[12]. The basic operational process is shown in Figure 14.

Figure 14

Basic process of air assault operation

Air assault operation (AAO) can be divided into intelligence reconnaissance (IR), electronic confronting (EC) and assault operation (AO). Operational nodes can be divided into intelligence security nodes (ISN), C2 nodes (C2N) and fire attack node (FAN). The ISN include radar force (RF), electronic confronting force (ECF), intelligence center (IC) and warning satellite (WS). The C2N include united command post (UCP) and regional command post (RCP). The FAN is aviation force (AF). Operational information can be divided into intelligence security information (IIS), combat command information (ICC) and weapon control information (IWC).

4.2 Operational Activity Model

According to the difference of view particle sizes, IR can be divided into warning & detecting (WD), identification (ID) and information amalgamation (IA). EC can be divided into electronic reconnaissance (ER), electronic interference (EI) and fire destruct (FD). AO can be divided into standoff assault (SA) and precision assault (PA). AF node can be divided into command post (ACP), warning & command plane (WCP), interfering formation (IF), suppressing formation (SF), assaulting formation (AF) and covering formation (CF). RF node can be divided into command post (DCP), camp a (CA)b (CB) and c (CC). The hierarchical structures of activities and nodes are shown in Figure 15.

Figure 15

Hierarchical structures of activities and nodes

According to mapping rule 1, namely, activity→activity information mapping, two inner mappings of intelligence reconnaissance activity, two inner mappings of electronic confronting activity, one inner mapping of assault operation activity and three mappings between three kinds of operational activities are built. The eight activity→activity mappings are shown in Table 1.

According to mapping rules 2 and 3, namely, activity→node and node→activity information mappings, eleven activity→node mappings and sixteen node→activity mappings are built. The twenty-seven mappings among activities and nodes are shown in Table 2.

Based on the analysis above, activity model for air assault operation is built as shown in Figure 16.

Figure 16

Hierarchical structures of activities and nodes

4.3 Operational Node Model

According to mapping rule 4, namely, node→node information mapping, twenty-three inner mappings of fire attack node set, fourteen inner mappings of intelligence security nodes set, two inner mappings of C2 nodes and twenty-one across mappings among three operational nodes set are built. The sixty node→node mappings are shown in Table 3.

Based on the analysis above, node model for air assault operation is built as shown in Figure 17.

Figure 17

Node model for air assault operation

4.4 Information Alternation Model

Based on the operational activity and node models above, information alternation model is built with main line of information types. In this model, according to implementation process of air assault operation, original nodes, targets nodes, generating activities and consuming activities of different information are analyzed orderly. Information transferring course and alternation relationships are sorted clearly. Information alternation model for air assault operation is shown in Figure 18.

Figure 18

Information alternation model for air assault operation

Taking information flow as main line, architecture modeling considers synthetically various factors, such as operations, information, system and technology, and clarifies operational process analysis further, which outstand significant effects of different information and realize the architecture developing strategy of information flow centered and process oriented. Compared with architectures built simply from viewpoints of operations, service and capability before[15–17]), the operational architecture above displays elements more comprehensively, analyzes modeling results more visually, depicts information relationships more clearly and describes requirements more normally, which promote the use and exchange of architecture information effectively and expatiate the operational concept of information dominant in modern warfare adequately.