Font Size: a A A

Research On Resource Index Construction And Query Optimization In Cyber Physical System

Posted on:2015-05-17Degree:DoctorType:Dissertation
Country:ChinaCandidate:W B MaFull Text:PDF
GTID:1108330509961071Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
With the integrated development of micro-sensor technology, embedded system technology, pervasive computing technology and network communication technology, also, accompanied with cloud computing, networking and other new emerged concepts, a close connection between the physical and information domain is increasingly required by people, which correspondingly generates a Cyber-Physical System(CPS). CPS is a cosmically complicated system which integrates control technology, information processing technology and network transmission technology. It has been considered as a core technology of the next industrial revolution.Military Cyber-Physical System is a CPS that is specifically used in military domain. Unlike a general CPS, the MCPS emphasizes on military physical space(e.g., the large-scale military struggle environment, local conflicting military operation environment) and military information space(e..g., cyberspace battlefield, battlefield information space, electromagnetic space, etc.), consisting of maintenance resources, transportation tools and the like. In order to use these resources the first thing is enable a query of these resources, that is, being able to find these resources. However, due to the real-time, dynamic and massive resistance features of these resources, it is often not easy to do a enervate query within a manageable time. Thus, in this dissertation, a method of resource index and query optimization(resource index construction and query optimization, RICQO) is proposed for solving the issue of the environmental resources query under CPS. Overall, the main work is as follows.First, the architecture of query indexing feature in CPS is proposed. Although the classic CPS has mentioned that a query indexing function is important, no detailed architecture for the query indexing function in CPS has been introduced. Based on the classic architecture of the CPS, a new query indexing architecture which includes a query indexing function is proposed. An analysis of the sub-modules(i.e., road network maintenance sub-module, resource trajectory index construction sub-module, index traversal and maintenance sub-module and knowledge index construction and maintain sub-module) of this CPS architecture is conducted. Also a detailed description of the function of each sub-model is provided.Second, a RICQO method using semantic features under a static environment is proposed. Due to fact that resource index having semantic relationships has a high query efficiency, when constructing a resource index, the semantic relationships between two resources is considered, from which a IR-Tree index is constructed. In addition, the top-k query optimization algorithm is presented to support a user to do the semantic query for resource. Last, the effectiveness and the efficiency of the proposed method is verified in three aspects, i.e., resource index construction time cost, query time cost, query correctness. Experiment results show that the proposed method is valid date and correct.The third Chapter proposes a RICQO method for handling resource index construction and query under static environment given the consideration of the real road network. Based on the semantic classification index construction, a real road network model is established. Correspondingly, the RICQO method that is for static resources in a real road network is elaborated, that is, an algorithm of Directed By Edge, an algorithm of Directed By Point and an algorithm using the IR-tree method. A real data based experiment is conducted to examine the effect and efficiency of the proposed method, and the results are show to be good.The fourth Chapter proposes a belief based RICQO method for handling resource index construction and query under an uncertain military environment. In a traditional spatial resource retrieving algorithm, the important feature of the uncertain factor of resources is often ignored. This indicates that the traditional spatial resource retrieving algorithm cannot be extended to the military resource network. This Chapter proposes an index structure based on the belief method, and its corresponding top-k retrieving algorithm. Based on the traditional IR-tree, a belief based resource index structure is developed. The calculation method for the resource belief and region belief is also provided. The top-k query optimization that is specially designed for the belief based resource index structure is proposed as well. Results of our empirical studies with an implementation of the proposed index structure and algorithm show that our proposed method is effective.The fifth Chapter proposes a history trajectory RICQO method for handling resource index construction and query under dynamic environments. The method is proposed to enable an query of dynamic resources. Specifically, first a model of the trajectory of dynamic resources, called the trajectory two-state separating model(TTSM) is constructed. Next, a index structure named the STIR-tree is proposed which can integrate the trajectory information into resources locations and so enable an construction of dynamic resources. At last, a STIR-Tree based neighboring query method is proposed. Experiment results show that the proposed algorithm is effective.The sixth chapter proposes a knowledge based RICQO method for handling resource index construction and query under the dynamic environment. For the dynamic resources, first three different methods of knowledge discovery approaches are proposed, aiming to obtain the regular knowledge of state changes. According these obtained knowledge, some index construction methods and query optimization approaches are proposed which supports a user to do future resource query. A real data based experiment shows that the proposed methods are effective and efficiency..In conclusion, this dissertation focuses on index construction and query optimization of the CPS resource. Given a consideration of the massive, real-time and dynamic features of CPS resources, different RICQO research methods for different resources are proposed. Research results have both theoretical and practical significance on improving the query performance and CPS resource management technology.
Keywords/Search Tags:Cyber Physical System, Internet of Things, Real Road network, RICQO, Knowledge discovery
PDF Full Text Request
Related items