Development Of Intelligent Communication System Test Platform

A new paradigm will be emerged in transportation,manufacturing,medical and other industries due to the introduction of artificial intelligence to wireless communications.Under this circumstance,the intelligent communication systems will be developed rapidly.Intelligent communication systems refer to the wireless communication systems with the capabilities of scenario perception,autonomous learning and decision-making by applying artificial intelligence to wireless communication systems.The efficient and reliable operation of intelligent communication systems in various fields can be ensured by comprehensive,objective and accurate performance evaluation of intelligent communication systems.However,the intelligent communication systems are only tested from the spectrum perception accuracy,spectrum access efficiency,interruption probability and other aspects in the current research of testing of intelligent communication systems.To address the above-mentioned problems,the intelligent communication system test platform is developed.In addition,how to design of test scenarios and how to evaluate intelligence entropy are proposed in this thesis.Different existing test methods of intelligent communication systems are compared in this thesis.The scenario generation system based on scenario feature layer description and intelligence entropy evaluation method of intelligent communication systems based on scenario are proposed.Aiming at the problems of low reusability,high redundancy and poor scalability of building wireless simulation scenario,a scenario generation system based on the hierarchical description of scenario features is designed.Firstly,the scenario components are divided into environment feature class,system device class and intelligent device class.Then,the statistical parameter values of various scenario components are calculated.In addition,the performance parameters of scenario components in the environment feature layer,system device layer and intelligent device layer are determined in turn based on the statistical parameter values of scenario components.Finally,the components of each layer are combined into the simulation scenario.The MATLAB is used to verify the scenario generation system.The redundancy of simulation scenario can be effectively reduced by the proposed method.It is an open issue how to define the “intelligence level” of intelligent communication systems,and intelligent communication systems can hardly be tested using the performance globally.The model of intelligence entropy evaluation of intelligent communication system based on scenario level is designed.In this model,the performance gain,resource saving rate and scenario-challenge level are proposed as the influence factors of intelligent entropy of intelligence evaluation.The relations between the affecting factors of the intelligence entropy are determined.Furthermore,the evaluation model of intelligent communication systems is established.The MATLAB is used to verify the model of intelligent entropy evaluation of intelligent communication system.The systematic performance of intelligent communication system can be effectively tested by the proposed model.In addition,it is difficult to determine suitable test scenarios for intelligent communication systems due to its high complexity.An evaluation feedback mechanism is proposed to achieve accurate test with as few test scenarios as possible.The feasibility of the proposed method is verified by evaluating the intelligence entropy value of the method to manage spectrum based on piecewise pricing.