Research On Trusted Method Of Sensor Node Interaction Task In CPS Real-time Environment

Through Internet of Things(IoT),Cyber-Physical System(CPS) connects the embedded devices such as data acquisition unit, communication and routing device, controllers, sensors, etc., so as to achieve networked control. CPS integrates computing unit and physical object of highly heterogeneous and multi-dimensional network environment, and makes a complex hybrid system which allows the continuous-time and discrete-time interaction, so CPS has the comprehensive characteristics like real-time performance, robustness, autonomy, efficiency, high-performance and resource-constraint and so on. CPS is comprised of node layer, network layer, resource layer and service layer, and the core layer is the node layer. The node implements the functions of perception, recognition and control, which utilizes the radio frequency identifier, wireless sensors and controller, etc. in order to achieve real-time and interactive control feedback of tasks. In large-scale and long-term CPS space and time environment, under the premises which ensure the credibility of sensor node's interaction tasks, and keep the best balance between system performance and credible efficiency, it is difficult to research the load balancing by sensor node's task migration in CPS.In the environment of wireless sensor network, the interactive task's integrity, confidentiality and migration policies of nodes are the research focus of behavioral abstraction and depict credible interaction and fusion on CPS computing and physical tasks. Through the protection of security and privacy, it can effectively reduce or even keep nodes from being stolen, monitored or brute-force cracked, in order to ensure the high credibility of nodes and networks. On this basis, it can keep the load well-distributed by conducting real-time task migration, and it solves the problems of limited computing power and storage capacity, so the efficiency and stability of concurrent interactive tasks of the sensor node's resource constrained situation under CPS real-time environment are improved.This paper presents an integrity policy of trusted sensor networks, which bases on the Merkle Hash Tree--TS-MHT, though the use of natural synchronization method of hash tree for the distributed nodes, and safety performance, combining with their own characteristics of sensor nodes and network. This paper also proposes a method for targeted improvements, and designs two security control child policies about broadcast packets of base station and nodes. Finally, the integrity is demonstrated by theoretical analysis and simulation experiments, which guarantees the significant anti-tamper capacity and accuracy of information from the expected sources.The confidential policy of trusted sensor network is based on GCM mode--AES-GCM. Five kinds of block cipher mechanism are compared:Electronic Code-book mode(ECB), Cipher Block Chaining mode(CBC), Cipher Feedback mode(CFB), Output Feedback mode(OFB) and Counter mode(CTR), so as to find out their advantages and disadvantages, and the counter mode is selected out. Based on the selected counter mode, this paper proposes the AES-GCM algorithm, of which the high security and confidentiality are demonstrated by theory and simulation experiments.Finally, under the premise of ensuring the sensor nodes and network trustness and security, due to the limitations of node force and storage capacity, it leads to increased local node load and energy consumption. Therefore, this paper puts forward a kind of real-time migration strategy for nodes of trusted sensor network. According to the predetermined threshold value of node load and energy, it implements online autonomous migration within and across nodes. At last, the results prove that it can ensure the load balancing of physical nodes and sensor network under the premise of limited resources, and it can achieve confidentiality and integrity of sensor nodes' interactive tasks in CPS real-time environment.