Intelligent Network Control Mechanism And Function Verification Based On Semantic Interactions

The rapid development of various autonomous unmanned systems has increased the endogenous intelligence of machines,which has enhanced functional collaborations among intelligent machines and has brought new challenges to the autonomous control of intelligent networks.Traditional network control protocol mechanisms have limitations regarding functionality,efficiency,and scalability,and existing research on intelligent network control has not yet resulted in a unified paradigm to facilitate efficient interactions among machines and between machines and humans.Therefore,it is important to explore a new paradigm of intelligent network control mechanism for building functional collaboration of intelligent machines.Natural language,as a concentrated expression of human intelligence,is highly robust and efficient,and these advantages are well suited for enabling intelligent network control mechanisms.Currently,with the rapid development of semantic communication,natural language processing and other technologies,the semantic information in natural language can give full play to its application value.Based on this,this thesis proposes the intelligent network control mechanism and functional verification based on semantic interactions,and conducts an in-depth study on the new intelligent network control mechanism paradigm from two aspects:the design of intelligent network control mechanism based on semantic interactions and the functional verification of intelligent network control based on semi-physical simulation,respectively.Firstly,to address the problem of a new intelligent network control mechanism paradigm,this thesis proposes an intelligent network control mechanism based on semantic interactions,which improves the robustness and efficiency of network control while semantically empowering network control.In order to realize the intelligent network control mechanism,the intelligent network control mechanism architecture is outlined,a corpus of natural language-based control instructions is constructed,a natural language-based control instruction interaction architecture is built,and a natural language-based interaction protocol is designed,so as to integrate semantic interactions into the network control mechanism.On this basis,the intelligent network control mechanism is implemented in software simulation,and its network control performance is evaluated and analyzed from two perspectives of accuracy and efficiency.The experimental results show that the intelligent network control mechanism proposed in this thesis has higher robustness and efficiency compared with the control scheme.Secondly,for the functional verification of the intelligent network control under the electromagnetic interference avoidance scenario,a semiphysical simulation system carrying the intelligent network control mechanism is proposed in this thesis,which uses a Universal Software Radio Peripheral(USRP)to functionally verify the proposed intelligent network control mechanism.Specifically,a semi-physical simulation system architecture for electromagnetic interference avoidance scenario is proposed,and USRP-based wireless communication frameworks are built and GNU Radio-based electromagnetic interference avoidance function modules are developed.Finally,a semi-physical simulation system with intelligent network control functions such as automatic interference identification,autonomous frequency band switching,and restoration of normal communication is realized.The simulation experiments not only prove the feasibility and effectiveness of the semi-physical simulation system,but also complete the functional verification of the proposed network control mechanism.