雷達信號處理領域其中一個重要的研究方向為自動目標識別，隨著雷達設備及訊號接收器的發展，自動目標識別的應用越來越廣泛，例如：軍事作戰、航線管理，領海領空警戒管理。
傳統辨識方法需要透過演算法找出特徵，再以特徵為基礎進行目標識別；而深度神經網路主要透過大量資料訓練，自動獲得目標特徵，再由選好之分類器進行目標識別。近年的研究，神經網路應用於雷達目標自動識別效果優於傳統識別方法。雷達距離輪廓圖是一種簡潔快速的雷達描述特徵方法，它是由沿著目標視線將目標切割成相對應的散射中心距離所組成。因此本論文研究使用深度長短期記憶網路和真實高解析度雷達距離輪廓圖，進行船艦的自動目標識別。由實驗結果得知，長短期記憶網路測試準確率皆可高達97%以上，實驗證明本論文所提方法可有效自動識別雷達目標。

One of the important research directions of radar signal processing is automatic target recognition. With the development of radar equipment and signal receivers, the application of automatic target recognition is becoming more and more widely used, such as: military operations, route management, and territorial airspace warning management.
Traditional identification methods need to find features through algorithms, and then target recognition based on the features; and deep neural networks mainly through large amounts of data training to automatically obtain target features, and then selected classifier for target recognition. In recent years, the neural network applied to the automatic recognition of radar targets is better than traditional recognition methods. Radar High Resolution Range Profile(HRRP) is a simple and rapid method for describing characteristics of radar. It consists of cutting the target into corresponding scattering center distance along the line of sight of the target. Therefore, this paper use Long Short-Term Memory(LSTM) networks and real Radar High Resolution Range Profile s for automatic target recognition of ships. According to the experimental results, using LSTM network can be as high as 97% or higher.The experiment proves that the method proposed in this paper can effectively identify radar targets automatically.

[1] P. Tait, Introduction to radar target recognition, 2005.
[2] Patricia López-Rodríguez , Raúl Fernández-Recio, Ignacio Bravo, Alfredo Gardel, José L. Lázaro and Elena Rufo, "Computational Burden Resulting from Image Recognition of High Resolution Radar Sensors," Sensors, 2013.
[3] H.-J. Li and S.-H. Yang, "Using range profiles as feature vectors to identify aerospace objects," IEEE Transactions on Antennas and Propagation, vol. 41, no. 3, pp. 261-268, 1993.
[4] Luo, S., Li, S. Automatic target recognition of radar HRRP based on high order central moments features. J. Electron.(China) 26, 184–190 (2009)
[5] Du, Lan & Liu, Hongwei & Bao, Zheng. Radar HRRP target recognition based on higher order spectra. Signal Processing, IEEE Transactions,2005
[6] Zhou, Daiying & Shen, Xiaofeng & Liu, Yangyang. Nonlinear subprofile space for radar HRRP recognition. Progress In Electromagnetics Research Letters. 2012
[7] B. Feng, L. Du, C. Shao, P. Wang and H. Liu, "Radar HRRP target recognition based on robust dictionary learning with small training data size," 2013 IEEE Radar Conference (RadarCon13) ,2013
[8] Ajorloo, Abdollah & Hadavi, Mandi & Bastani, Mohammad & Nayebi, Mohammad. Radar HRRP Modeling using Dynamic System for Radar Target Recognition. Radioengineering, 2014
[9] Zhou, D. Orthogonal maximum margin projection subspace for radar target HRRP recognition. J Wireless Com Network 2016, 72 (2016).
[10] Yuan, Lele. A time-frequency feature fusion algorithm based on neural network for HRRP. Progress In Electromagnetics Research M, 2017.
[11] K. Liao, J. Si, F. Zhu and X. He, "Radar HRRP Target Recognition Based on Concatenated Deep Neural Networks," in IEEE Access, vol. 6, pp. 29211-29218, 2018
[12] Jiang Y, Li Y, Cai J, Wang Y, Xu J. Robust Automatic Target Recognition via HRRP Sequence Based on Scatterer Matching. Sensors (Basel). 2018
[13] Jianbin Lu, Zemin Xi, Xianghui Yuan, Guishui Yu, Mingmin Zhang, "Ship target recognition using high resolution range profiles based on FMT and SVM," IEEE CIE International Conference on Radar, 2011.
[14] Lu Jianbin, Tian Shusen, Xi Zemin, "Frame segmentation and recognition algorithm for ship's HRRPs based on hypothesis testing," CIE International Conference on Radar (RADAR), 2016.

[15] Osman Karabayır, Okan Mert Yücedağ, Mehmet Zahid Kartal, Hüseyin Avni Serim, "Convolutional neural networks-based ship target recognition using high resolution range profiles," International Radar Symposium, 2017.
[16] Jinwei Wan, BoChen, BinXu, HongweiLiu, Lin Jin, "Convolutional neural networks for radarHRRP target recognition and rejection," EURASIP Journal on Advances in Signal Processing, 2019.
[17] Afshine Amidi and Shervine Amidi , "Recurrent Neural Networks cheatsheet" https://stanford.edu/~shervine/teaching/cs-230/cheatsheet-recurrent-neural-networks
[18] Sepp Hochreiter and Jürgen Schmidhuber,"Long Short-Term Memory",Neural Computation 1997 9:8, 1735-1780
[19] Dishashree Gupta ,"Fundamentals of Deep Learning – Activation Functions and When to Use Them? "https://www.analyticsvidhya.com/blog/2020/01/fundamentals-deep-learning-activation-functions-when-to-use-them/
[20] vikashraj luhaniwal, "Analyzing different types of activation functions in neural networks — which one to prefer?"https://towardsdatascience.com/analyzing-different-types-of-activation-functions-in-neural-networks-which-one-to-prefer-e11649256209
[21] Ravindra Parmar,"Common Loss functions in machine learning"
https://towardsdatascience.com/common-loss-functions-in-machine-learning-46af0ffc4d23
[22] Tommy Huang,"機器/深度學習: 基礎介紹-損失函數(loss function)"
https://medium.com/@chih.sheng.huang821/%E6%A9%9F%E5%99%A8-%E6%B7%B1%E5%BA%A6%E5%AD%B8%E7%BF%92-%E5%9F%BA%E7%A4%8E%E4%BB%8B%E7%B4%B9-%E6%90%8D%E5%A4%B1%E5%87%BD%E6%95%B8-loss-function-2dcac5ebb6cb
[23] Diederik P. Kingma and Jimmy Ba, "Adam: A Method for Stochastic Optimization," International Conference for Learning Representations, 2015.
[24] Lin,Wan-Yu,Applying Convolution Neural Network to Automatic Recognition of Radar Targets