Study On Resource Management Scheme For Cognitive Radar

Cognitive radar is a new radar system.This system constantly scans the surrounding environment and establishs a dynamic database to save the environment historical information of the targets to enhance the performance of radar.The radar system is capable to estimate and feedback the tracking performance in real-time,intelligently adjust the resources for searching and tracking in order to save of resources and improve radar performance.Currently,cognitive radar has become a new trend in the development of modern radar technology,the main research directions include environment sensing,resources management and waveform optimization.In this article we focus on the resources management includes the prediction of RCS,single-beam cognitive radar resource management,multi-beam cognitive radar resource management,resource management for detecting radar,radar network time resource management.The main work is as follows.1.First,we describe the resource management process:estimating,forecasting and distribution.We discuss the accuracy of the prediction target state influence on the final effect of resource management and design an algorithm to predict the target's RCS at the next time-step.The method is based on the probability density function and hidden Markov process and it is possible to obtain the target's RCS with limited information,which enable the effective of resource management.2.For the single-beam cognitive radar,we present a resource management approach.First,differ from the traditional Cramer-Rao lower bound to be the evaluation function,we use the method of probability density function to model the target's tracking error,which considering the target's range accuracy,speed accuracy and the revisit interval.Next we use the method of Markov decision process to optimize the radar resources,on one hand we can be adjust the resources in real-time,on the other hand we can consider the long-term impact of the decision-making.3.For the multi-beam cognitive radar we designed two resources management algorithms.First we describe the advantage of the multi-beam cognitive radar to the single-beam radar.Then we present two different algorithms to fulfill the different performance requirements of radar.The first algorithm use the method of multidimensional dynamic programming to improve the tracking accuracy.While the second algorithm use the method of Knapsack problem to save the time resource.At the end,we designed a simulation scenario to simulate seven different algorithms to explain the advantage of cognitive radar resource management algorithm.4.In this part we designed a algorithm to manage the time resource for the radar searching.At first we verify that the dwell time of the radar beam is irradiated on a certain direction have a significant impact for the detection probability.Then we reference to the method of tracking before detection(TBD),processing all the data of radar echo,accumulating the signal by using the target's motion feature and calculate the probability target appears in one direction.Using the probability as the evaluation function,we manage the time resource to enhance the detection performance of radar.5.For the radar network we designed a method to manage the time resource.At first,we model the problem of radar tracking targets in the scene of radar network and derive the probability density function of the target tracking error.Considering the different type of radar in the radar network,we analysis the error generated by the Coordinate transformation.Then we describe the method of chance constrained and propose the constraint.In the end,we use the method of Markov Decision to manage the time and power resources of the radars in the radar network.