Research On Special Radar Signal Deinterleaving Algorithm In Complex Environment

In recent years,with the continuous emergence of new system radars,the electromagnetic environment is becoming more and more complex,and the pulse signal density is getting higher and higher,and reconnaissance receivers need to receive and process multiple types of radar pulse signals that overlap each other in a short time.In order to analyze the radiation sources in the electromagnetic environment,these overlapping pulses must first be sorted,and the results of signal sorting directly affect the subsequent intrapulse analysis and target positioning results.The types of pulse-to-pulse modulation of radar signals can generally be divided into five types: fixed,jagged,jitter,resident conversion,and slippery radar signals.Among them,the resident conversion and slip radar signals greatly increase the difficulty of radar signal sorting by virtue of the randomness of their pulse repetition interval(PRI),which is a novel special radar signal.In this thesis,the single-parameter radar signal sorting algorithm in complex electromagnetic environment is studied to cope with the problem of radar signal sorting of novel special PRI modulation types.1.The principle of traditional radar signal single-parameter sorting algorithm is expounded,such as the commonly used cumulative difference histogram algorithm,sequence difference histogram algorithm,PRI transformation method and modified PRI transformation method,etc.,and the algorithm process and advantages and disadvantages are analyzed in detail.2.Resident conversion radar signal is a type of radar signal with both PRI residency and PRI conversion characteristics.When sorting and residently converting radar signals,the signal sub-period is difficult to detect due to its complex and changeable forming rules.The type of pulse repetition period modulation is unknown,and the pulse train cannot be successfully extracted;The extracted multiple sub-periodic pulse sequences cannot be combined into one radar signal,resulting in false alarms.To solve the above problems,a signal sorting algorithm based on PRI resident conversion radar is proposed,which uses double thresholds to improve the probability of sub-period detection,determines whether the signal belongs to resident conversion radar by Hough transform combined with iterative self-organizing clustering,improves the search algorithm to extract pulse trains,and merges multiple pulse trains into one resident conversion radar signal according to the permutation entropy index.Simulation experiments show that the proposed algorithm can successfully sort the resident conversion radar signals in a complex electromagnetic environment.3.The interpulse slip radar signal is a special modulation style radar signal.In the complex electromagnetic environment,it is difficult to accurately estimate the PRI and its variation range due to the interpulse slip radar signal.Multiple slip forms make it difficult to identify PRI modulation types;Even if the modulation type is identified,the pulse signal cannot be accurately extracted,resulting in sorting failure.In order to solve these problems,this thesis uses the multi-level box PRI transform method to estimate the slip center and slip boundary of the slip signal,identifies the modulation type by the attention-based bidirectional long short-term memory network(BiLSTM-Attention),and extracts the slip radar pulse signal by selecting the corresponding mode sequence search method according to the recognition results.The simulation results show that the algorithm has high PRI estimation accuracy and modulation recognition accuracy,and can effectively sort slip radar signals in complex electromagnetic environments.