Article Preview
Top1. Introduction
Service-oriented computing is an emerging paradigm where services are platform-independent computational entities. These services could be published, delivered, executed and discarded based on the quality of service (QoS) requirement (Phillips, 2013). It also supports the distributed, interoperable, and integration of services on common and shared platform called enterprise service bus (ESB). This platform aims at developing a novel approach for extending the features of service-oriented architecture (SOA) that support combining the “off-the-shelf” component-based software engineering. “QoS” may refer to extending the meaning of functional as well as non-functional support to service properties, protocol and business logic. This is well explained with building template to quality assurance as discussed in thesis (Reinecke et al, 2010; Kyusakov et al., 2013). Quality assurance methods typically inferred the service execution patterns which helps in better understanding the actual need for service integration in legacy application like sensor network. Whereas fundamental theories and methodologies associated with data acquisition for sensor network need to address issues associated with service invocation. Communication could be strengthened by well-define ‘link’ between correspondent ‘node’. Here, ‘node’ could be client or just sensor node which actually rise request for services and communication between them is governed by a ‘link’. The service execution ensures the quality of data which is being exchanged between multiple nodes with help of desired protocol (Fiadeiro et al., 2012).
Figure 1. Sensory data streaming over local and cloud server
The streaming of sensory data could be possible with both local as well as cloud server. Whenever, such data is stored at a repository through a local server, its access is limited with specific number of user. On the other hand, when such data is make available at cloud databases through global server, its access is vital and rapid through various distributed algorithms and techniques. This scenario is well depicted in Figure 1 in which sensory data is traffic through cloud infrastructure at global resources. This arrangement could be possible with different service modeling in SOA systems.
1.1. Service Modeling
The business conversation is important style in engagement of services between its clients. This could efficiently be organized through exchange of correlated messages (Myers et al., 2010; Klaus et al., 2015). For example, an insurance company client may request for status update to his/her insurance refund. This creates an ‘event’ that does the invocation and integration of multiple services. This service invocation could either commit the refund or cancel it on basis of certain predefined features. Such features are governed by service-oriented computing and it also models the service for which conversation is supported (Cámara et al., 2016). This concept can be conceived for sensory data acquisition in which following action is required to get an update:
- •
Interaction must be based on exchanging information within the system for which typically event must be occurred. This may include requests, reply, commit, cancel or invoke for each interaction;
- •
Interpretation must be done using predefined business patterns, usually defined in term of enterprise integration patterns (EIP). Such procedures enable the interaction between communicating party;
- •
Execution must be governed by ‘state’ of service and there has to be predefined path or route which justifies the transition of event from one ‘state’ to another;
- •
Invocation in business conversation must be done on ‘parity’ flag basis. This flag clears the interest of different ‘Node’ whether it want to interact with communicating other party.