最近幾年，網路的普及率越來越高，透過網路溝通的人也相對的越來越多，但也因為網路傳送的不安全性，有心人可以擷取到訊號，進而知道溝通的內容，為了個人或公司的保密，已經有好多語音加密的演算法被提出，本論文也是其中一。
論文中以盲處理(BSS)來做為加密的方式，透過矩陣的相乘-- ，矩陣A以不同比例把矩陣 裡各段訊號混合，產生混合訊號 ，而且混合訊號就像是雜訊一樣，另外透過獨立成份分析(ICA)的方法找到最適合的解混合矩陣W來還原訊號-- ，這個方法的好處是除了加密訊號外，並不需要太多資料就可以還原訊號，而且也不需要會造成使用不方便的碼本(codebook)。
在改良的地方，首先在加密的部份改變Lin對矩陣A和key訊號的限制，目的在增加key space，讓所有可能發生的情況變多，增加破解需要的時間，另外在解密的部份，為了不讓變大的key space造成運算上的負擔，利用主要成份分析(PCA)的特性--white與分群，做為ICA的前處理，目的在減少運算的資料量，讓運算的時間縮短，在最後的模擬結果可以證實理論上的推導是正確的。

In the recent year, according to the network is more and more popular, people communicate by the network are also increase. But the network is not always safe for the signal transmission. Some people can intercept the signal and get the content of the signal. Therefore, many speech encryption algorithms have been proposed for protecting the secrets of a person or the company. The thesis is also one of them.
The thesis uses blind source separation (BSS) as the encrypted way. By the multiplication between matrices-- , the signal in the matrix is mixed in different proportion because of the matrix A, and generate the mixed signal . The mixed signal is like the noise. Besides, BSS uses independent component analysis (ICA) to get a suitable de-mixed matrix W to reduce the signal-- . The advantage of this method does not need too many data to reduce the signal besides the encrypted signal. And no codebook causes the inconvenience of use.
In the improvement, at first the encrypted part changes the restriction on the matrix A and the key signal of Lin. The purpose is to increase key space and let the decrypted time be more. Then in the decrypted part, to avoid the great key space cause the load of computation, use principal component analysis (PCA) be the prior step. The purpose is to decrease the data of computation and let the computational time be reduced. The simulation results can verify the theory is correct.

[1] H. J. Beker and F. C. Piper, Secure Speech Communications. London, U.K.:
Academic, 1985.
[2] Wikipedia—RC4，Data Encryption Standard, 01/25/2008
[3] Q. H. Lin, F. L. Yin, T. M. Mei, and H. L. Liang, “A speech encryption
algorithm based on blind source separation,” in Proc. Int. Conf. Commun., Circuits
Syst.: Signal Process., Circuits Syst., 2004, vol. II, pp. 1013–1017.
[4] A. Cichocki and S. Amari, Adaptive Blind Signal and Image Processing:
Learning Algorithms and Applications. New York: Wiley, 2003.
[5] A. Hyvärinen, J. Karhunen, and E. Oja, Independent Component Analysis. New
York: Wiley, 2001.
[6] J. F. Cardoso, “Blind signal separation: Statistical principles,” Proc. IEEE , vol.
86, no. 10, pp. 2009–2025, Oct. 1998.
[7] S. Amari and A. Cichocki, “Adaptive blind signal processing—Neural network
approaches,” Proc. IEEE , vol. 86, no. 10, pp. 2026–2048, Oct. 1998.
[8] T. Ristaniemi, K. Raju, J. Karhunen, and E. Oja, “Inter-cell interference
cancellation in CDMA array systems by independent component analysis,” in Proc.
4th Int. Symp. Independent Compon. Anal. Blind Signal Separ., 2003, pp. 739–744.
[9] V. D. Calhoun, T. Adali, L. K. Hansen, J. Larsen, and J. J. Pekar, “ICA of
functional MRI data: an overview,” in Proc. 4th Int. Symp. Independent Compon.
Anal. Blind Signal Separ., 2003, pp. 281–288.
[10] M. Kawakatsu, “Application of ICA to MEG noise reduction,” in Proc. 4th Int.
Symp. Independent Compon. Anal. Blind Signal Separ., 2003, pp. 535–541.
[11] J. F. Cardoso, J. Delabrouille, and G. Patanchon, “Independent component
analysis of the cosmic microwave background,” in Proc. 4th Int. Symp. Independent
Compon. Anal. Blind Signal Separ., 2003, pp. 1111–1116.
[12] W. Kasprzak and A. Cichocki, “Hidden image separation from incomplete
image mixtures by independent component analysis,” in Proc. 13th Int. Conf. Pattern
Recogn., 1996, vol. II, pp. 394–398.
[13] Q. H. Lin and F. L. Yin, “Blind source separation applied to image
cryptosystems with dual encryption,” Electron. Lett., vol. 38, no. 19, pp. 1092–1094,
2002.
[14] Q. H. Lin and F. L. Yin, “Image cryptosystems based on blind source
separation,” in Proc. IEEE Int. Conf. Neural Netw. Signal Process., 2003, vol. II, pp.
1366–1369.
[15] James V. Stone, “Independent Component Analysis: a tutorial introduction,”
The MIT Press 2004.
[16] Aapo Hyvarinen, Juha Karhunen and Erkki Oja, “Independent Component
Analysis,” John Wiley & Sons 2001.
[17] Qia-Hua Lin, Fu-Liang Yin, Tie-Min Mei, and Hualou Liang, “A blind source
separation based method for speech encryption,” IEEE Transactions on circuits and
systems—I: Regular papers, VOL. 53, NO. 6, JUNE 2006.
[18] Shujun Li, Chengqing Li, Kwok-Tung Lo, Guanrong Chen, “Cryptanalysis of
an Encryption Scheme Based on Blind Source Separation,” Computer Science Aug
2006.
[19] S. Sridharan, E. Dawson, B. Goldburg, “Fast Fourier transform based speech
encryption system,” IEE PROCEEDINGS-I, Vol. 138, No. 3, JUNE 1991.
[20] Fulong Ma, Jun Cheng and Yumin Wang, “Wavelet transform-based analogue
speech scrambling scheme,” ELECTRONICS LETTERS 11th April 1996 Vol.32
No.8
[21] Matt J. B. Robshaw, “Stream Ciphers Technical Report TR-701, version 2.0,”
RSA Laboratories, 1995
[22] M. Liskov, R. Rivest, and D. Wagner, “Tweakable Block Ciphers,” Crypto 2002