Asynchronous Message- Passing and Inter

36 International Journal of Knowledge-Based Organizations, 4(4), 36-50, October-December 2014

Asynchronous MessagePassing and Inter-Application Communication Software for Process Improvement in Complex Systems Darko Galinec, Zagreb Polytechnic for Technical Sciences, Zagreb, Croatia Ljerka Luić, Karlovac University of Applied Sciences, Karlovac, Croatia

ABSTRACT Complex systems nowadays are usually supported by previously developed or inherited application systems, which have been developed on the basis of business functions standard model. Such application systems cover particular business fields, however, since they have been developed in an independent and uncoordinated way, there are modest possibilities of interacted automated data exchange because of heterogeneous data. Because of increasing data exchange needs, enforcement of development and upgrade of programmatic procedures is required. In this connection it is necessary to find ways and means to integrate all existing application systems in given circumstances and render possible exchange and usage of all needed data in a reliable and consistent way. The problem can be solved in different ways. Asynchronous message-passing and inter-application communication software appliance imply commencement of automated data exchange among existing organizations application systems, providing for full scale control and management of mutually exchangeable data, internally as well as externally. In this connection contributing model of the running application systems integrated by means of Message-Oriented Middleware (MOM) of the complex system has been created. Keywords:

Application, Database, Information System, Integration, Message-Oriented Middleware (MOM), Process Improvement

1. INTRODUCTION With complex business systems and their accompanying information systems, which support the implementation of their business functions, there is a need for constant mutual

data exchange. In this case automated processes of one information system produce data that are used by another information system or several of them. Here, complex system is referred to as any system featuring a large number of interacting components (agents, processes,

DOI: 10.4018/ijkbo.2014100103 Copyright © 2014, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

International Journal of Knowledge-Based Organizations, 4(4), 36-50, October-December 2014 37

etc.) whose aggregate activity is nonlinear (not derivable from the summations of the activity of individual components) and typically exhibits hierarchical self-organization due to certain (selective) circumstances. This definition applies to systems from a wide array of scientific disciplines. Indeed, the sciences of complexity are necessarily based on interdisciplinary research (Rocha, 2010). Generation, usage and maintenance of procedures which support processes and operations of data exchange present an additional effort for information staff and an additional expense for the organization. This is so due to the constant need for data exchange, in circumstances of frequent changes in the inner and outer environment of the complex system. This kind of data exchange slows down other business processes in the system as well – those depend on data which need exchange, because this procedure cannot be executed in real time. The dimension of this problem is more related to organizational domain than to the technical; in solving it we must take into account the restriction which is obvious in the fact that within every information system, alongside the allowed (and desirable) business process reengineering in the complex system, the data structure of the master data system should be maintained. In the case examined and elaborated by the author, databases can physically be centralized or distributed, but from the point of view of business functions and processes they are logically separated and the object-related communication is conducted through the existing and specifically created programmatic procedures for efficient database updates. There are no common database tables, and the data, when transferred from one system into another, change their format and structure - they are being processed through new group of specific procedures according to the system they have entered, keeping their information content. Different information systems programmatic solutions are also independent of each other and support the work of independent databases as parts of these systems. There is a networking infrastructure which enables the interconnection

between information systems and work for users of all information systems. It also provides information assurance. The need for integration of information systems exists because part of the data created by one system processes (output values) is used for processing within one or more other systems (input values). The basic integration condition of the information system must be releasable: with information systems which are complex and distributive in organization, as well as in those which are organized centrally, managing the data of the complex system must be centralized, i.e. master data management (MDM) must be achieved. Technical interoperability is relative to the norms and standards used for the interconnection of computer systems and services. It includes open interfaces, network and security services, middleware, integration presentation and data exchange. Semantic interoperability is relative to data meaning. Thanks to this level exchanged data have the same meaning at the starting point and destination, pieces of information originating from various information resources are linked in a meaningful way. Process interoperability is relative to business goals, business process modeling and cooperation achievement between different units whose structure and work mode are not necessarily congruous. To fulfill system user needs and to reinstall available and simple user services it is necessary to establish process interoperability. The objective of the research is to provide information systems interoperability by the means of message-oriented middleware. This paper is structured as follows. Section 1 and 2 discuss the necessity to integrate information systems of the organization; the concept of solution model is given as well. Section 3 explains technical aspects of examined integration solution at logical level, including new model of the running application systems, integrated by means of message-oriented middleware of the complex military system.

Copyright © 2014, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.


Possibilities to enlarge the elaborated approach on future integration processes in relation to complex systems external circumstances are examined in Section 4. Conclusive Section 5 reveals benefits which can be achieved by application of proposed concept of the information systems integration and complex systems process improvement. The approach itself is open for enlargement, dynamic adjustments and extensions.

2. ORGANIZATION ANALYSIS, BUSINESS GOALS AND INTEGRATION APROACH Usually, public sector and defense organizations face challenges that differ from the challenges for private firms. They have to meet multiple, often conflicting goals and they are subject to constraints of a financial, legal, contractual, personnel and institutional nature. Radical processfocused change in public sector and defense organizations can only be achieved through deep changes in their bureaucratic practices. Renovation in the public sector and defense mostly emphasizes quality and productivity improvements, the elimination of bureaucracy, process simplification and the reduction of processing times. In public administrative processes ontology-based organizational memory systems are especially important: many existing sources of knowledge, laws, comments on laws, specific regulations, old similar cases, available case-specific documents and information etc., are prevalent in different places and in different forms and representations, at several degrees of formality, and are related through many links (Papavassiliou et al., 2003). Nowadays large enterprises are confronting new and significant challenges including: customer expectations for more comprehensive and timely access to markets and information, growth in business volumes, and changes to service delivery models (Fowler, 2002). As stated by Kovačić et. al. (2006, p.4), very complex and process-oriented nature of business has led organizations to use process modeling methods

and tools as a means of managing the complexity of these systems, and to aid in achieving business goals. Business process modeling (BPM) has now been in the public domain for four decades, but it is only in the late 1990s that integrated business process modeling tools have been developed. Integrated process modeling tools must be capable of showing interconnections between the activities and conducting a decomposition of the processes. These tools must help users to conduct “what-if” analyses and to identify and map no-value steps, costs, and process performance (bottleneck analysis). They should be able to develop AS-IS and TOBE models of business processes, which represent both existing and alternative processes. They can be used to predict characteristics that cannot be directly measured, and can also predict economic and performance data that would otherwise be too expensive or impossible to acquire. Each BPM software application is defined by a mix of several components. The most important components of BPM tools are: (1) process modeling and design; (2) process monitoring; (3) process operation (automation and integration); (4) technology platforms and interfaces. However, the common characteristic of BPM tools is the ability to develop, use and maintain the business process repository. On the business and technological levels one has to plan, analyze and shape data and processes by which the exchange is done, and to make a programmatic solution for linking and intercommunication of the information systems to be integrated. The programmatic system must have 3 main parts, functioning as follows: •

Data mapping system: Which copies master entity data of the information systems exchanging data. For the automating of data transfer and linking two bases it is necessary to predict, model and shape all the data and procedures necessary for the data copying between systems to be consistent. This part of the system must ensure data dynamics (creating, viewing, editing and erasing data), because it is subject to



•

•

change. It is the basis for logical linking and exchange, and for physical copying and storage of synonymous data, designated differently within each of the existing transaction systems. This kind of data is the result of organizational procedures from the previous period, and the data values which their identifiers point at are necessary for data exchange and the processing required. This part of the programmatic system must ensure simple data handling, and give the system administrator complete control and management of data and the programmatic solution; Message-oriented middleware (asynchronous message-passing and interapplication communication software): This system ensures data transfer between information systems, definition of data format and linking paths, use of the existing and development of missing data-extracting programmatic procedures; besides it insures monitoring of the message transfer system. A message is typically composed of three basic parts, the headers, the properties, and the message payload or body (Chappel, 2005); Applications for automatic interfacing of existing information systems: In the roles of sender/receiver in the message transfer system. It enables the sending/receiving, adjusting and processing of the messages (data) received.

Information systems exchanging data must have an in-built functionality of sending and receiving data, with the option of their transformation before exchange. They must (on the sending side) include procedures for data selection in both directions, protected transfer of data through the linking path, entry to the data mapping system and their transformation, continuation of safe transfer, receiving and distribution of data into appropriate application modules (on the receiving side). The transformed data (adapted to the master data of the sender’s or recipient’s information system) must fill in the tables of the appropriate

modules in the processing application. In case there are not any, tables and modules of the processing application must be defined and built, and implemented in the existing information systems - in the databases and programmatic solutions part. All information system application modules (the newly-made and the existing) must eventually make an interconnected and consistent whole - an integrated programmatic and device system.

2.1. Review of Previous Work For integration of two business functions or two business systems it is necessary to connect their business processes with application support and data exchange. Processes appertaining to one application system create data which will be used by another application system. First and the key reason for the integration of business systems applications are user business needs for business processes and information flow, and changes in business processes which occur during business transactions. The next integration reason is related to the differences in technologies by means of which applications are made. Integration should be carried out to connect technologically different applications. Because of process complexity, which includes breakdown of the existing business processes and applications, business processes changes on the basis of business needs and user requirements, modeling of such processes and new applications and their connection is necessary for successful completion of Enterprise Application Integration (EAI) project. Applications have been developed on the basis of business system function architecture (for example, automation of production, procurement i.e. applications for their support). EAI related requirements also appear with business combinations - integration of business subjects into a whole (fusion, consolidation, acquisition). In such cases information systems applications, of the subjects which are going to be merged, are integrated. From early sixties to the late seventies of the last century, business systems



applications were simple in view of shape and functionality. Business system data integration was not considered at all, the aim was to support manual procedures by PC. During eighties the need to integrate applications within business systems was recognized. There were attempts to reshape existing applications to make them suitable for integration. Since during nineties Enterprise Resource Planning (ERP) applications prevailed, the existing applications and data were adjusted to ERP system. The aforesaid could be done only through EAI introduction as a logical sequence of events. Later on the advantages of integration of multiple business processes through existing applications have been conceived. Other factors which have contributed to the EAI development include growth of applications intended for supply chain management (SCM) support and business to business integration (B2B), applications for modern business processes support and web application integration. Modern business requires process centric approach, i.e. end-to-end (E2E) management and control of business system. The process includes various applications intended for the support of different business processes and functions. Contemporary example of automation (Todea, 2005) and integration of military process and application systems describes the U.S. Army’s Tank-automotive and Armaments Command (TACOM) Life Cycle Management Command’s sensor solution (Soft32.com, 2005). It is integrated with IBM middleware, and offers its headquarters staff the ability to gain the same on-demand access to mission critical information that can currently be viewed in remote regional and field locations. As a result, it can look forward to improving logistical support and maintenance for military vehicles. The solution uses embedded sensors in military vehicles that send signals from the field to IBM middleware and business partner applications in central locations to remotely diagnose repairs, and to determine fuel and ammunition replenishment needs. Currently, troops are required to make routine in-person inspec-

tions of military ground vehicles, sometimes in battlefield. The successful pilot demonstrates that automating the process has the potential to improve troop productivity and safety. The pilot was demonstrated on the Stryker Brigade’s tactical wheeled vehicles. Embedded sensors in Stryker Brigade military vehicles, using IBM WebSphere MQ Series™ as a messaging interface, transmit data from IBM WebSphere Microbroker™ and IBM DB2 Everyplace™ by relaying information via wireless networks to computers in numerous trucks that transport and replenish parts from various locations. Data from the field can then be transmitted from computers in the trucks, or local checkpoints, to central headquarters locations via satellite. The headquarters can monitor military readiness on numerous business process applications based on IBM WebSphere Portal™ and IBM DB2 Universal Database™ software. When a vehicle is dispatched, the operator can turn on a health monitoring maintenance system which provides data about its readiness. This also triggers a Global Positioning System (GPS) that identifies the location of the vehicle. Colored icons will appear if a repair problem arises, immediately alerting the vehicle operator to notify a supervisor that there is a need to make a decision to proceed with a mission or to work with the maintenance team to diagnose the problem and automatically order parts. Data on the repair history of the vehicle is stored in an IBM DB2 Universal Database™. Transforming a manual process which is labor intensive and sometimes dangerous to an automated process can improve operational efficiency, cut costs, and keep the troops out of harm’s way.

2.2. Benefits Expected Interface systems will be construed as subsystems to the existing information systems within which there is need for data exchange. It will enable sending data from the source information system, use of the message-oriented middleware and their receiving on the part of the target information system, and the other way



round. The received data will gain the shape adapted to the master data of the source/target information system, fill in the corresponding tables of the right processing modules. Tables and processing modules will be defined, built and implemented into the existing information systems - databases and programmatic solutions. Information systems integration tasks will be executed through methodological development procedure stages and steps, by recognizing data and process groups - defining the object system business technology, and changing, expanding or cancellation of the existing business processes, with the opportunity to introduce new ones. This way we will get an integrated information system, by connecting the existing information systems which have the characteristic of data independence and the master data. The organization’s integrated information system will be open - in the future it will enable data exchange with new information systems. Its openness should in the end lead to the possibility of constructing and realization of an integrated information system in the organization, by connecting the existing heterogeneous information systems and those emerging.

2.3. Development of the Complex System Integration Solution The reason for integration of existing applications within the organization is to enable the support of newer, faster, and more accurate business processes, providing meaningful and consistent information management. A business process is defined as a subset of business activities performed by the organization to achieve the goals for which it has been created. Activities correspond to different stages of process execution. In order to be initiated, some activities require particular artifacts or events as an input which may be taken directly from the environment or produced as outputs by other activities. Thus, an event is a passive element of the process that reflects a signal in a business environment which triggers the execution of an activity. Data object is an

instance containing a collection of data and methods for operating on that data [Kovačić et. al., (2006), p.4]. In the process of implementing the integration procedure, which should be conducted in accordance with the regulations for the designing and building of information systems (through the stages of planning, analysis, defining, building, introduction and implementation) there will be experts of a diverse project team, composed of third-party experts and the organization’s information staff. It is ensured that the latter gain the knowledge of the existing complex system functions, processes, activities (procedures) and data - business technology, system control and management through control and management of key data and processes. Organizations require security functionality in five main areas (Dang Van Mien, 2003): 1. Authentication: Are the users of an application who they say they are? 2. Integrity: Is a received message identical to the one that was sent? 3. Privacy and confidentiality: Can you ensure that only the intended recipient can read a message? 4. Audit and non-repudiation: Can messages be audited, so that nobody can deny having sent or received a given message? 5. Availability: Can a message be sent at any time, from anywhere? Successful integration, in accord with the order of development priorities and corresponding to the required conditions for information protection and safety, will ensure the possibility of consequent integration of emerging information systems, both in inner and outer environments.

2.4. Methodology Nowadays, there are several EAI methodologies. One of them (Cognizant Technology Solutions, 2005) mentions the following methodological steps:



1. Estimate of EAI necessity resulting in the respective Estimate Report; 2. Strategic EAI planning and implementation, resulting in Strategic Report; 3. Development of application and technical architecture, resulting in Detailed Implementation Plan; 4. EAI implementation, monitored through Testing Reports and Quality Assurance Survey. Galinec & Vidović (2008) present new enterprise application integration (EAI) methodology which is driven by business processes and is carried out in phases. EAI should be carried out in five methodological phases: 1. Business processes modeling and planning of EAI process needs according to business level (planning); 2. Analysis of communication and semantic requirements with transactions-requirements towards EAI; 3. Providing for interoperability through three-level EAI model (design); 4. EAI development by means of adequate technology (construction); 5. EAI implementation and acceptance.

3. TECHNOLOGICAL AND FUNCTIONAL ASPECTS OF INTEGRATION SOLUTION Business integration platform can be used to (IBM Corporation, 2003b): 1. Create and deploy new business processes; 2. Synchronize business events in multiple systems; 3. Integrate applications on diverse platforms; 4. Transform information formats “en-route” (along the way) between applications; 5. Link people into a new business approach. Detailed performance management and activity monitoring is essential for managing an

integrated business environment and ensuring efficient and effective operation. A real-time, continuous view of all activities enables immediate reaction to any changes in operating performance (Magic Software Enterprises, 2004). An integration subsystem is an open, modular, parameterized system of linking and integration of two or more independent systems into the so-called conjugated systems model. The integration subsystem has its own database for the needs of parameterization and tracking of message exchange between linked systems. It is installed on an external server, because its work is autonomous, and the linked systems can be on the same or separate device system, with compatible or a different kind of operating systems. The integration subsystem relies on middleware, which ensures quality communication and guarantees safe data transfer between the linked systems. Secondly, it directly supervises the interfaces to the linked systems, and thirdly its interface is exposed toward the user who controls and manages the integration subsystem.

3.1. Functions of Integration Subsystem The concept of loosely-coupled interfaces basically relies on a communication channel that can carry messages as self-contained units of information. They may be in either binary or text form. By the usage of messages for information exchange systems are abstractly decoupled. Therefore systems do not need to know details of each other. Message-oriented middleware provides the abstraction of a messaging queue that is reachable from the network (Bakken, 2003). Message-oriented middleware does not model messages as method calls; it is rather based on sets of messages. Message-oriented middleware provides set of Application Programming Interfaces (APIs) so that clients can send and receive messages. Using the message-oriented middleware and the messages which they generate, applications communicate directly with each other. Message-oriented middleware performs func-



tion of a message router and in some cases of a queuing system. Message-oriented middleware supports synchronous and asynchronous messaging; although it is mostly associated with asynchronous messaging using queuing. Enterprise messaging is at the core of enterprise service bus (ESB) architecture. A messageoriented middleware is a key part of the ESB architecture, as it provides the underpinnings of the network of virtual channels that an ESB uses to route messages throughout an extended enterprise and beyond (Chappel, 2005). Integration capability enables integration of people, processes, information, and systems throughout the enterprise (IBM Corporation, 2003a). Integration subsystem functions are divided into 4 segments, as follows: 1. InterfaceA: With two functionality subgroups: a. AIn: Function set is in charge of retrieving the requested changes and applying, that is, propagating them in the information system “A”. It also keeps track of exchange for the needs of control and analysis; b. AOut: Function set is in charge of retrieving the requested changes from the information system “A”, and their corresponding record and sending to the Dispatcher Module; 2. InterfaceB: With two analogous functionality subgroups: a. Bin: Function group is in charge of retrieving the requested changes and applying, propagating them into the information system “B”. At the same time, it keeps track of changes for the purpose of control; b. Bout: Function group for retrieving changes initiated in the information system “B” and sending them on to the Dispatcher Module; 3. Dispatcher: Module, in charge of: a. Message manipulation, their distribution, forwarding, enrichment and control; b. Data mapping;

c. Authorization of requested changes; d. Keeping records of changes; 4. Admin: Functions allow: a. Parameterization of the integration subsystem; b. Administration of mapping charts; c. Administration of exchange rules and condition; d. Analytical review of synchronized changes (who, when, what); e. Synthetic, statistical review of synchronized changes; f. Control and management of the integration subsystem (audit, alert, manual synchronization etc.) An integration solution model, with the above characteristics, is shown in Figure 1. Communication by questions and answers goes as follows: 1. 2. 3. 4. 5. 6. 7. 8. 9.

Application generates message; Outgoing Queue; Outgoing Channel - Incoming Channel; Incoming Queue; Application generates reply message; Outgoing Queue; Outgoing Channel - Incoming Channel; Incoming Queue; Application records reply message.

3.2. Integration Implementation and System Inter-Communication The integration is implemented through message exchange between the linked systems. They are uniquely specified with a message type and operation direction. Every message type can have its own structure, or message types can be grouped into a uniform structure, and it is also possible to define a unique structure for all message types. The concept used depends primarily on the nature of the data exchanged. The number of different message types is limited and countable, but it is always possible to add new messages, if needed.



Figure 1. Logical scheme of integration subsystem

3.2.1. Synchronization of Change Initiated within Information System Called «A» If the change (action) is not from the segment in which actions for data transfer are, nothing happens, the status remains the same. If the change has come from the transfer request segment, the AOut module will recognize this information, prepare it for sending, and send it to the Dispatcher. The Dispatcher will verify the message, authorize it, edit it, do some mapping if needed (depending on the parameters and rules of exchange, which are set through the Admin module), and route it to system “B”. Automatically, the BIn module accepts the message and carries it through in system “B”, according to all the business rules of this system. All relevant information is recorded in the integration system database for quality control, management and analysis.

3.2.2. Synchronization of Change Initiated within Information System Called «B» If the change is not from the segment of required transfer changes, nothing happens, the status remains the same.

If the change is requested for transfer, the BOut module will recognize this information, prepare it for sending, and send to the Dispatcher. The Dispatcher will verify the message, authorize it, edit it, do some mapping if needed (depending on the parameters and rules of exchange), and direct it to system “A”. Automatically, the AIn module receives the message and carries it through all the business rules of system “A” through this system. All relevant information are recorded in the integration system database, for the purpose of quality control, management and analysis. If the transfer actions have not been successful, the messages remain in the system, accessible through the Admin module, and can be manually processed at the target system.

3.2.3. Case Study The example (shown at Figure 2) with contributing model of the running information system illustrates the work principle and communication methods within complex military system. The major problem to be solved is integration and achievement of reliable messaging between the two application systems:



Figure 2. Integration of OPT and FMAS by means of message-oriented middleware

1. Application system of the electro-mechanical system Outdoor Payment Terminal (OPT); 2. Fuel Management Application System (FMAS). Fuelling station (one or more) and computing center are situated at physically distant locations. OPT application system and Websphere Message Queue Series™ (MQ) client (at communication server) are at fuelling station location. Customer Activated Terminal application system and communication server are connected through File Transfer Protocol (FTP). OPT performs the following functions: 1. By concentrator of Multi Pump Dispenser (MPD), it links MPDs used for fuel pouring out and records fuel consumption; 2. Through sound concentrator, it connects sounds in fuel tanks and enables fuel level recording; 3. Data from previously mentioned processes, by the use of OPT application system, are forwarded to FMAS, which is at a remote location; 4. In case of many fuelling stations, OPTs and communication server are connected through router. WebSphere MQ Series™ server and WebSphere Application Server™ are situ-

ated at computing center. Communication between aforementioned application systems is performed by means of inter-application communication software with asynchronous message-passing. OPT application system access to the WebShere MQ Series™ client has been made possible by MQ application programming interface (MQ API). Activating business logic of OPT application system and creation of sender thread (task) within the process of operating system, data from application system at fuelling station location (for ex.) in comma-separated values (CSV) format are put into outgoing message queue (WebSphere MQ Series™ client) and through MQ protocol transfer to incoming message queue (WebSphere MQ Series™ server in computing center). From message queue (over JMS protocol) Java Message Service (JMS) listener of the WebSphere Application Server™ “listens” changes, and sends received data to the business logic of FMAS. In the case study, data format and semantics are the same for both application systems, i.e. thanks to business logic of FMAS it is known which processes should follow on the data which are received in a certain format and with a certain value. The whole procedure is performed in a similar way the other way around. FMAS by means of JMS manager and through JMS protocol puts data into outgoing message queue (WebSphere Application Server™). Thence



through MQ protocol data reach incoming message queue (WebSphere MQ Series™ client). Through Websphere MQ Series™ interface and listener thread of execution, data from message queue are given to business logic of OPT application system to be further processed. As in previously described communication procedure, data syntax and semantics are defined and accepted in each application system which participates in mutual data exchange. In the example given, because of characteristics of the described process and appertaining data (data syntax and semantics are defined once and are not changed for a longer period), function of mapping charts creation and administration, data mapping as well as administration of exchange rules and conditions are not necessary.

4. INTEGRATION WITH EXTERNAL SYSTEMS All organizations must communicate reliably and seamlessly with the Enterprise Resource Planning system as a backbone (IntelliCorp, Inc., 2003). As for each complex system, there is constant need for data exchange not only between its corresponding information systems, but also with the elements from its environment, expanding this research gives the option of using the planned inner system processes integration solution for all the relevant outer system processes. Considering that by using this concept we get further openness of the organization, and it is necessary to pay extra attention to the appearance and content of the integration safety infrastructure, in order to fulfill the conditions required.

4.1. Integration Infrastructure Complex ICT environment nowadays are heterogeneous; heterogeneity of hardware, operation systems, applications, development tools, communication channels, lots of computer paradigms (messaging, objects, transactions, data, processes, networking and communication

- synchronous and asynchronous), an integrated business process passes fully through a system parts of which are heterogeneous, through islands within the system. This introduces the need for developing new methodologies to simplify and accelerate the development of new applications, as well as integration of the existing systems into a unified information flow, in a unified way, with flexibility concerning business processes. Prompted by these needs, the information-communication technology industry has developed solutions for information systems embodied in a strong integration and application infrastructure. The representative of the application infrastructure is the application server, which represents a unified operative system for heterogeneous platforms, which integrates and standardizes the approach to system resources, meta-data and uses the safety infrastructure services. Also, it simplifies system maintenance and administration, and supports system openness through the openness of the technologies/ standards used. Figure 3 shows the integration infrastructure for the interaction between information systems and its applications and the complex system environment. In the authentication process the usage of safety infrastructure is fully standardized. If we look at the public key infrastructure (PKI) as the basic safety infrastructure, then by its standards it generates a unique identification of an individual (a certificate). The certificate is stored separately from the public part of Lightweight Directory Access Protocol (LDAP); it is stored on the real person’s property, most often a smart card (private part). The LDAP server represents a repository which gets data from PKI, and is read by the application infrastructure for the need of authentication of registered users. The authenticated user is then authorized, given permissions and functions. Authentication and authorization processes take place in the background of the applications themselves. All applications use the same service for authentication and authorization of users through the application infrastructure.



Figure 3. Integration infrastructure environment

Application servers also have messaging services so there is no more a clear line between integration and application infrastructure in the sense of commercial products. The most robust integration infrastructure has its representatives in separate message servers. It is for critical integration processes, where great reliability of data exchange is required, that asynchronous messaging communication gains advantage over classical RPC (Remote Procedure Call) communication or the more recent technology of communication through web services. The integration is conducted through message exchange between linked systems. They are uniquely specified with a message type and operation direction. Every message type can have its own structure, or message types can be grouped into a uniform structure, and it is also possible to define a unique structure for all message types. The concept used depends primarily on the nature of the

data exchanged. It is always possible to add new messages, if needed. Message content is transparent for the integration infrastructure and presents a sequence of bytes, and the information content of the message is retrieved by using applications. Messaging does not technologically exclude XML (EXtensible Markup Language) either, XML messages can be tunneled through the messaging infrastructure. Authentication and authorization processes take place transparently from the applications themselves; all applications use the same user authentication and authorization service through the integration infrastructure. If the communication between two or more systems requires utmost secrecy of data, it can be additionally protected by hardware cryptographic devices, which execute linear protection of such communication by the highest safety standards.



4.2. Generic Integration Elements The example of the integration of two information systems (application system “A” and application system “B”) shows generic elements, which can be expanded to other information systems in a organization, but also to outside systems, which belong to the environment. Information systems of an organization present isolated islands where the processed information usually remains, without the option of its further distribution and use. Isolation is usually a consequence of a series of heterogeneities; heterogeneities of the hardware platform, operation systems, systemic software and development tools. That is the reason why it is necessary to establish data flow, open the application systems, so they would become source and destination of information (IN and OUT processes) toward other application system. This approach uses the integration structure as a lever for successful overcoming of the information systems heterogeneity, enabling the establishment of a unified integration methodology. An application adapter (shown in Figure 3), is defined as a standardized adjustment hardware for the access point of the complex system to the integration infrastructure, and as a group of functions it is construed and becomes a constituent part of an information system within complex system. The administration module unites the transformation and processing rules in connecting two information systems within business subsystems. The structure described represents horizontal integration of the existing systems and also implicates transparent construction of new horizontal systems, as a condition for establishing vertical integration processes. An example of accomplishing a vertical integration process is the information portal, which in a new way evaluates the existing information, and it is the example of establishing new services and system values.

To summarize, there are two kinds of systems. We recognize the first one by looking at the existing system (current status). It is closed, single-layered, non-standardized and needs upgrading with an application adapter. Characteristic for the other type of systems (status as it should be) is that they are multilayered; when approaching the system resources and data application infrastructure is used, with the transparency of authentication and authorization.

5. CONCLUSION AND FUTURE WORK The processes within the complex system which are to lead to the integration of the corresponding information systems, based on the standard business functions model independently of one another, are long-lasting and expensive (cost-intensive in both time and money). Such processes demand many of organizational efforts and adjustments, with the constant danger of possible conflicts between separate decision makers (representatives of the organizational sections whose information systems are being integrated within complex system) and of execution interruptions. The classical way of integrating information systems, through the usage of database technology, cannot in this sense give satisfactory results because it needs significant cuts into the model and database structure. The integration procedure and the solution wipes out this need for core changes in the databases and the integrating information systems data, introducing adjustment hardware and interfaces between systems instead. With the implementing of a message transfer system the problem is solved on a higher level of organization and technology, without the need for direct changes in the data structure and organization. In this way the unchanged data model still, in a simplified way, through



a set of entities, attributes and connections, shows the complex systems characteristics, while for the data copying a new mechanism has been introduced, with full “inner” control by the complex systems authorized expert staff. Despite the need for additional modeling and programming in the project development stages in order to make an interface, and the need for educating the expert and user staff for the future use of the programmatic solution, after the system has reached a stable level one only needs to keep the data up-to-date and supervise the process (which is characteristic for all approaches). This paper investigates possibility of integration of heterogeneous application systems within complex military system. In this connection, the specific contribution of the authors’ research to the EAI discipline is in design of the new EAI model. Based upon the model, the integration solution within complex military system is implemented (example given in the case study), allowing communication between application subsystem of electro-mechanical system Outdoor Payment Terminal (OPT) and Fuel Management Application system (FMAS) by means of inter-application communication software with asynchronous message-passing (e.g. message-oriented middleware). The integration solution with reliable messaging between information systems, concerning its characteristics, can be expanded to all system elements in the inner and outer environment which in regard to business processes interconnection and data exchange, need integration. This approach also leads to the restructuring of business processes as a positive effect, because the introduction of interfaces among application systems reduces the number of participants and connections in data exchange, and increases the data flow speed. Limitations of implemented integration solution are related to semantic interoperability achievement between information systems within which data syntax and semantics change more frequently, including the master data. In this case function of mapping charts creation

and administration, data mapping as well as administration of exchange rules and conditions are mandatory. Further investigations are directed towards implementation of solutions conducive for data mapping and administration between information systems that have previously described properties.

REFERENCES Bakken, D. E. (2003). Middleware. In J. Dasgupta (Ed.), Encyclopedia of Distributed Computing. Kluwer Academic Publishers. Chappell, A. D. (2005). Enterprise Service Bus. O’Reilly Media, Inc. Cognizant Technology Solutions. (2005). EAI methodology. Retrieved March 27th, 2007, from http:// www.cognizant.com/services/advancedsolutions/ eai/eai_method.htm Corporation, I. B. M. (2003a). Retrieved February 20th, 2007, from http://www.software.ibm.com/ mqseries Corporation, I. B. M. (2003b). Retrieved February 20th, 2007, from http:// www-306.ibm.com/software/ integration/wbiserver/anncv421/ Dang Van Mien, A. (2003). Five Security Strategies for IBM WebSphere MQ series. Business Issues Research COM-19-7492. Gartner, Inc. Fowler, M. (2002). Patterns of Enterprise Application Architecture. Addison-Wesley Professional. Galinec, D., & Vidović, S. (2008). Methodological Framework and Application Integration Services in Enterprise Information Systems. In M. Pagani (Ed.), Encyclopedia of Multimedia Technology and Networking (2nd ed., pp. 901–911). Information Science Reference. doi:10.4018/978-1-60566-014-1.ch123 IntelliCorp, Inc. (2003). Retrieved January 23rd, 2006, from http://www.intellicorp.com/application_integration.htm, Mountain View. Kovačić, A., Bosilj Vukšić, V. & Lončar, A. (2006). A process-based approach to knowledge management. Ekonomska istraživanja. 36(2), 53-66. Magic Software Enterprises. (2004). Retrieved September 18th, 2006, from http:// http://www. magicsoftware.com/bin



Papavassiliou, G., Ntioudis, S., Abecker, A., & Mentzas, G. (2003). Supporting knowledge-intensive work in public administration processes. Knowledge and Process Management, 10(3), 164–174. doi:10.1002/ kpm.176 Pezzini, M., & Thompson, J. (2003). IBM WebSphere Integration Offer Signals Long-Term Plan. Products Research P-17-6956. Gartner, Inc.

Rocha, L. (2010). Complex Systems Modeling: Using Metaphors From Nature in Simulation and Scientific Models. Retrieved April 21st, 2011, from http:// informatics.indiana.edu/rocha/complex/csm.html Todea, L. (2005). US Army Completes Testing of New Sensor Solution. Retrieved April 21st, 2011, from http://news.soft32.com/us-army-completes-testingof-new-sensor-solution_31.html, Soft32.com.

Darko Galinec is an College Professor of Application Integration at Zagreb Polytechnic for Technical Sciences, Croatia. He obtained his Ph.D. degree in Information and Communication Science and master’s degree in Information Science (Program and information engineering) both from University of Zagreb. His research interests include fuzzy logic appliance in project managament, enterprise information systems strategic planning, architecture and integration, command and control information system (C2IS) interoperability. Currently his research is focused on cyber security. He presented his research at several international conferences, including Institute of Behavioral and Applied Management Conference (IBAM) - USA, Central European Conference of Information and Intelligent Systems (CECIIS) – Croatia, International Conference on Informatics - Slovakia. His articles appeared in the International Journal of Applied Strategic Management, Journal of Behavioral and Applied Management, Journal of Information and Organizational Sciences, International Journal of Information Systems in the Service Sector, International Journal of Human Capital and Information Technology Professionals, Central European Journal of Computer Science, Journal of Applied Mathematics and Computational Mechanics and Acta Electrotechnica et Informatica. He also published four book chapters. He taught courses at University of Split Computing studies of University Center for Applied Sciences and University College for Applied Computer Engineering in Zagreb. He teaches at Zagreb Polytechnic for Technical Sciences. Also, he serves as reviewer for International Journal of Human Capital and Information Technology Professionals and International Journal of Society Systems Science. Ljerka Luić is an Full College Professor of Information Systems at Karlovac University of Applied Sciences. Since the academic year 2001/2002 to date, she has continuously worked as a guest teacher for the Telemedicine course within the Biomedicine and Healthcare University Postgraduate Doctoral Study Program at the School of Medicine of the University of Zagreb. She has also lectured for the Information and Communication Science University Postgraduate Doctoral Study Program at the Faculty of Humanities and Social Sciences of the University of Zagreb on the subject of ‘Designing a Integrated Information System Strategic Planning Model’. As a sceintific researcher, she has actively participated in 3 national scientific projects and 2 international scientific projects, acting as Lead Researcher/Manager on one of them. She was appointed Senior Scientific Associate on January 2011, Associated Professor and Full College Professor on December 2013. All her degrees were obtained in social sciences, field: Information and Communication Sciences, branch: Information Systems and Informatology. Ljerka Luić has published a total of 42 papers (30 scientific papers and 8 professional papers). She has taken part in 20 international scientific conferences, at 12 of which she personally presented papers authored and/or co-authored by her. She has been mentor for 82 graduate theses and co-authored 10 papers with her students. She has autonomously published 1 textbook and co-authored 1 manual and 2 chapters of a book. Copyright © 2014, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.