Research On Modulation Signal Recognition Of Underwater Acoustic Communication Based On Random Forest

Military powers around the world attach great importance to research in the field of underwater communication signal countermeasures,of which acoustic communication signal reconnaissance is of great importance.It will be a great information countermeasure advantage for the reconnaissance side,no matter which side first reconnoiters the enemy’s important information and interferes with it or even decoys it.The recognition of acoustic communication modulation signal is one of the important parts of acoustic communication signal reconnaissance.Moreover,due to the increasing number of acoustic communication signal modulation methods,the acoustic communication channel is also more complex compared with the terrestrial communication channel.As a result,it is more difficult to identify the acoustic communication signal modulation methods.The main research contents and contributions of this paper are as follows.Firstly,the characteristics of acoustic channel and the theory of acoustic modeling are introduced,and the acoustic channel model based on BELLHOP model is established as the simulation experimental environment of communication modulation recognition algorithm.Next,each type of communication modulation and the mechanism are analyzed,and the signal set in this paper is selected from {2ASK,4ASK,2FSK,4FSK,2PSK,4PSK,DSSS/BPSK } 7modulation signals,and then 11 feature parameters are extracted based on the instantaneous characteristics,Spectral line characteristics,high-order cumulants,cyclic spectral functions and autocorrelation functions of each communication modulation signal.The 11 feature parameters are used to provide training and testing samples for the subsequent random forest algorithm.Finally,this paper proposes a modulation recognition method based on the random forest algorithm and optimizes the random forest parameters using the AOA algorithm.In this paper,a comparative analysis study of recognition rate based on instantaneous features,based on high-order cumulants,default parameter random forest and AOA algorithm optimized random forest algorithm(AOA-RF)will be conducted.The results of the simulation study show that the recognition algorithm using AOA-RF has a correct recognition rate of over 80% for 2ASK,4ASK,2FSK,4FSK,2PSK,4PSK and DSSS/BPSK at SNR ≥-7d B,and its optimized random forest algorithm has a significantly higher recognition rate than the other three algorithms.Finally,the AOA-RF algorithm is used for modulation recognition in the simulated acoustic channel,and the simulation study shows that its algorithm has good modulation recognition performance in the underwater acoustic channel.