The Natural Attribute Modeling Of Physical Object In Information-physical Space Convergence

The development of the Internet of Things(IoT)has built an interactive bridge between the diverse physical world and the efficient information world.This has been bringing network into daily lives,and promotes the constant convergence between physical space and information space.The fundamental task of information-physical space convergence is to establish the mapping between physical space and information space,and to realize the comprehensive perception and interconnection of ubiquitous things(i.e.,physical objects).In order to provide precise and efficient intelligent services for physical space,the information space needs to have an overall understanding of the appearance,rules and phenomena for physical objects.Therefore,researching on natural attributes modeling of physical objects has become the key point in the area of information-physical space convergence.This paper focuses on natural attributes modeling of physical objects,including the natural attributes based identification modeling of physical objects,the identity resolution of physical objects based on the natural attributes identification model,the semantic description modeling of natural attributes of physical objects and the search of physical objects based on natural attributes description model.It addresses the issues of identification,identity resolution,object representation,and search for physical objects in the information-physical mapping process.The main research works include:(1)The natural attributes based identification modeling of physical objects.By researching on the identification modeling with unique identity features of natural attributes,the identification method of physical objects based on uniquely identified natural attributes is proposed.Taking facial feature as an example,two face identification system models based on cloud and fog computing are proposed separately.It can resolve the problems of computation,communication and storage in the process of face identification,and improve the efficiency of identification systems.(2)The identity resolution of physical objects based on the natural attributes identification model.In view of the characteristics of the uniquely identified natural attributes,the physical object identity resolution framework based on the identification with uniquely identified natural attributes is presented.Taking face identification model as an example,the parallel face identifier matching mechanism based on MapReduce and the face resolution scheme based on cloud computing are proposed to improve the efficiency of face resolution,and it provides an unified face resolution service for different cross-industry and cross-platform IoT applications.Meanwhile,the face resolution scheme based on fog computing is proposed,which greatly reduces the network transmission and resolution time of the system.Moreover,towards the security and privacy issues brought by the properties of fog computing-based identification and resolution framework,some security and privacy schemes are proposed,including the session key agreement and authentication schemes based on Diffie-Hellman(DH)key agreement algorithm,the public key encryption mechanism based on elliptic curve and the hash data integrity checking algorithm based on secure hash algorithm-1(SHA-1).They resolve the problems of confidentiality,integrity and availability in the processes of face identification and resolution.(3)The semantic description modeling of natural attributes of physical objects.By researching on semantic expression methods for the natural attribute elements of physical objects and their relations,a semantic description model based on non-uniquely identified natural attributes of physical objects is proposed.Take the IoT devices as an example,the concept and modeling method of"Software-Definition Device(SDD)" is proposed from the view of information-physical space mapping.It separates the software management and control function from the hardware,eliminates the heterogeneity of devices,and realizes the unified management,sharing,reusing and recombination of device resources in IoT.For smart home scenes,the ontology-based semantic description modeling method of the natural attributes of the SDD device resources is proposed.It realizes the abstraction and natural attributes description modeling of device resources in smart home.(4)The the search of physical objects based on natural attributes description modeling.In view of the characteristics of the semantic description models and the physical objects search,the logical framework of semantic-based physical objects search system is proposed.Taking the devices in smart home as examples,the semantic based search method of SDD devices in smart home is proposed by combining the description model of SDD.It achieves the fast and efficient search for physical devices.By researching on the natural attributes-based identification modeling of physical objects,the identity resolution of physical objects based on the natural attributes identification model,the semantic description modeling of natural attributes of physical objects and the physical objects search and discovery based on natural attributes description model,the natural attributes is defined,organized,associated and stored in the manner which can be comprehended and processed by machine.Then the natural attributes model can be analyzed,processed and excavated by the various relevant applications,and act on physical object or physical space to realize the bidirectional mapping and convergence of physical space and information space.