Abstract: frequency-hopping (FH) is one of the commonly used spread spectrum techniques that finds wide applications in communications and radar systems because of its inherent capability of low interception, good confidentiality, and strong anti- interference. However, non-cooperation FH transmitter classification is a significant and challenging issue for FH transmitter fingerprint feature recognition, since it not only is sensitive to noise but also has non-linear, non-Gaussian, and non-stability characteristics, which make it difficult to guarantee the classification in the original signal space. Some existing classifiers, such as the sparse representation classifier (SRC), generally use an individual representation rather than all the samples to classify the test data, which over-emphasizes sparsity but ignores the collaborative relationship among the given set of samples. To address these problems, we propose a novel classifier, called the kernel joint representation classifier (KJRC), for FH transmitter fingerprint feature recognition, by integrating kernel projection, collaborative feature representation, and classifier learning into a joint framework. Extensive experiments on real-world FH signals demonstrate the effectiveness of the proposed method in comparison with several state-of-the-art recognition methods.

基于核空间投影和联合表征的跳频信号辐射源指纹特征识别

[1]Angelosante D, Giannakis GB, Sidiropoulos ND, 2010. Estimating multiple frequency-hopping signal parameters via sparse linear regression. IEEE Trans Signal Process, 58(10):5044-5056.

[2]Bellman R, 1997. Introduction to Matrix Analysis (2^nd Ed.). McGraw-Hill, New York.

[3]Boser BE, Guyon IM, Vapnik VN, 1992. A training algorithm for optimal margin classifiers. Proc 5^th Annual Workshop on Computational Learning Theory, p.144-152.

[4]Chandran V, Elgar SL, 1993. Pattern recognition using invariants defined from higher order spectra one-dimensional inputs. IEEE Trans Signal Process, 41(1):205.

[5]Cherkassky V, 1997. The nature of statistical learning theory. IEEE Trans Neur Netw, 8(6):1564.

[6]Cherkassky V, Mulier F, 1998. Learning from Data: Concepts, Theory, and Methods. Wiley, New York, NY, USA.

[7]Cover T, Hart P, 1967. Nearest neighbor pattern classification. IEEE Trans Inform Theory, 13(1):21-27.

[8]Friedl MA, Brodley CE, 1997. Decision tree classification of land cover from remotely sensed data. Remote Sens Environ, 61(3):399-409.

[9]Jameson A, 1968. Solution of the equation AX + XB = C by inversion of an M × M or N × N matrix. SIAM J Appl Math, 16(5):1020-1023.

[10]Lawrence RL, Wright A, 2001. Rule-based classification systems using classification and regression tree (CART) analysis. Photogramm Eng Remote Sens, 67(10):1137- 1142.

[11]Liao XJ, Bao Z, 1998. Circularly integrated bispectra: novel shift invariant features for high-resolution radar target recognition. Electron Lett, 34(19):1879-1880.

[12]Liu SH, Zhang YD, Shan T, et al., 2018. Structure-aware Bayesian compressive sensing for frequency-hopping spectrum estimation with missing observations. IEEE Trans Signal Process, 66(8):2153-2166.

[13]Liu W, He JF, Chang SF, 2010. Large graph construction for scalable semi-supervised learning. Pros 27^th Int Conf on Machine Learning, p.679-686.

[14]Mercer J, 1909. Functions of positive and negative type, and their connection with the theory of integral equations. Phil Trans R Soc A, 209(441-458):415-446.

[15]Muller KR, Mika S, Ratsch G, et al., 2001. An introduction to kernel-based learning algorithms. IEEE Trans Neur Netw, 12(2):181-201.

[16]Nie FP, Xu D, Tsang IWH, et al., 2010. Flexible manifold embedding: a framework for semi-supervised and unsupervised dimension reduction. IEEE Trans Image Process, 19(7):1921-1932.

[17]Quinlan JR, 1993. C4.5: Programs for Machine Learning. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA.

[18]Song TC, Cai JF, Zhang TQ, et al., 2017. Semi-supervised manifold-embedded hashing with joint feature representation and classifier learning. Patt Recogn, 68:99-110.

[19]Tadjudin S, Landgrebe DA, 1996. A decision tree classifier design for high-dimensional data with limited training samples. Int Geoscience and Remote Sensing Symp, p.790-792.

[20]Tang Z, Lei YK, 2017. Radio transmitter identification based on collaborative representation. Wirel Pers Commun, 96(1):1377-1391.

[21]Tugnait JK, 1994. Detection of non-Gaussian signals using integrated polyspectrum. IEEE Trans Signal Process, 42(11):3137-3149.

[22]Wang B, Li WF, Poh N, et al., 2013. Kernel collaborative representation-based classifier for face recognition. IEEE Int Conf on Acoustics, Speech and Signal Processing, p.2877-2881.

[23]Wang LP, Chen SC, 2017. Joint representation classification for collective face recognition. Patt Recogn, 63:182-192.

[24]Wright J, Yang AY, Ganesh A, et al., 2009. Robust face recognition via sparse representation. IEEE Trans Patt Anal Mach Intell, 31(2):210-227.

[25]Wu XD, Kumar V, Quinlan JR, et al., 2008. Top 10 algorithms in data mining. Knowl Inform Syst, 14(1):1-37.

[26]Yang M, Zhang L, Zhang D, et al., 2012. Relaxed collaborative representation for pattern classification. IEEE Conf on Computer Vision and Pattern Recognition, p.2224-2231.

[27]Yang MH, 2002. Kernel Eigenfaces vs. Kernel Fisherfaces: face recognition using kernel methods. Pros 5^th IEEE Int Conf on Automatic Face Gesture Recognition, p.215-220.

[28]Yang Y, Wu F, Nie FP, et al., 2012. Web and personal image annotation by mining label correlation with relaxed visual graph embedding. IEEE Trans Image Process, 21(3): 1339-1351.

[29]Yoshikawa M, Shindo H, Nishii R, et al., 1995. A fully automated design of binary decision tree for land cover classification. Int Geoscience and Remote Sensing Symp on Quantitative Remote Sensing for Science and Applications, p.1921-1923.

[30]Zhang L, Yang M, Feng XC, 2011. Sparse representation or collaborative representation: which helps face recognition? Int Conf on Computer Vision, p.471-478.

[31]Zhao LF, Wang L, Bi GA, et al., 2015. Robust frequency- hopping spectrum estimation based on sparse Bayesian method. IEEE Trans Wirel Commun, 14(2):781-793.