Assessment of data and interfacing of service-oriented system with sensors

Abstract

Data is a key factor in delivering service across the networks. It is an important aspect in handling service interaction which governed by instructions to facilitate legacy application like wireless sensor network. Such data is generally produced in passes and it is used as interaction parameters for the communication between two parties. This integration between multiple service applications could be strengthened by adopting efficient and effective mechanism for these web services. The Enterprise Service Bus (ESB) satisfies all the rules for handling these web services and it also extends the feature for service-oriented architecture (SOA) like routing, protocol transformation, mediation support for existing IT assets and many more. ESB is middleware framework that helps in designing and developing web services through which software intermediary could be possible. This could also be best solution for handling data from sensors and provide underlying support for protocol conversion between gateway and cloud. ESB is a kind of depletion layer that efficiently handles various overheads during communication. This thesis investigates the detailed issues about ESB and its implication with different mediation solution. This also includes essential design principles, architectural framework, routing functionalities, and survey for existing ESBs available in the market. This thesis also presents the detailed mathematical formulation that derives the relationship between multiple services (also termed as service components) for handling data through common and shared platform of ESB. Data from various sensors is collected through gateway and it need to check for noise/error before storing such data to global repository. Such noise/error in data may arise due to sudden voltage fluctuation, lose connection, short circuit and link breakage problems. This fluctuation in the data value creates unstable, discontinuity and noisy and need to be stabilize before storing. This thesis advises an “Intelligent Sensor Network” (IntSeN) methodology which is multi-level architecture for handling sensor data and filter data on the basis of delay, density, type and value threshold. Sensed data may contain noise or error due to environmental hindrances or voltage fluctuations that must be reduced for stabilizing an overall system. Such methodology reduces the relatively error by more than 50 % that occurred during sensing process.