Complex Virtual Instruments Cloud Service Model Design for Ocean Observation
Chao LIU,Nai-Jun HU,Zhi-Jin QIU,Zhong-Wen GUO
DOI: https://doi.org/10.16441/j.cnki.hdxb.20150393
2018-01-01
Periodical of Ocean University of China
Abstract:As the great growth of the population and usage of modern technologies,the terrestrial resource has become highly demanded.The development of ocean technologies draws great attention by researchers from different institutes of different countries.The current ocean observation systems are usually developed and deployed by different platforms and institutions,which brings problems of sensor accession,architecture difference,un-standard data format,and system integration.Virtual instrument technologyhas been widely used to solve this kinds of problems.However,it has lots of limitations.First,all the virtual instrument system are dedicated system which is designed to solve specific problem.They cannot adapt to different situations.Second,due to non-standard technologies,the systems are hard to integrate.Third,integrating large scale systems will increase the response time.To solve above problems,this paper proposes a new complex virtual instruments cloud servicemodel forocean observation.This model includes four main model,item description model,sensor model,complex virtual instrument model and mapping model.Item description model provide a new data structure that can describe all kinds of monitoring object.The sensor model provides standard data format to transfer the data which make it possible to use standard data interface to share the data.The complex virtual instrument model is designed to develop standard component which could be applied to all the system.The mapping model is provided to query the relationship of all the item.Based on this model,the developers can develop reconstructed ocean observation system.The system can automatically access different kinds of sensors and seamless integrate with heterogeneousocean observationsystem.To solve the slow responseproblemoflargescaleintegration,we propose self-adaption multi-component algorithm.The algorithm uses request queue's length to judge when to activate the algorithm.The algorithm can automatically replicate or reduce the components to deal with lots of requests based on different situation.To validate the algorithm,we randomly generate the component request to access the components.We use the response time to measure the performance under different situation.The result shows the algorithm is effective.We also validate the model through development,integration,maintenanceand application of the real systems.We rebuild our system by using the new model.The system uses 20 nodes to monitoring the surface of ocean.Two of them are near shore nodes which responsible to transfer the data to the complex virtual instrument system.We use the configuration system to manage all the components.It use drag-and-drop operations and parameters configuration to develop the system.The whole process of system development and integration shows that the model is effective and efficient.It can seamless integrate of heterogeneous system,improve development efficiency,and makes great contribution on large scale ocean observation system integration.