使用基因演算法之強健式傳輸系統設計之研究

Robost Transmission Based on variable-Rate Error Control and Genetic Programming

陳逸洲(Yi-Chou Chen)

指導教授 ： 黃文吉

中原大學/工學院/電機工程研究所/碩士(2001年)

https://doi.org/10.6840/cycu200100084

摘要

在本篇論文中我們提出以疊代的方式結合通道最佳化來源編碼及來源最佳化通道編碼綜合設計系統，以降低通道雜訊對通訊系統的影響。
通道編碼之主要目的在於錯誤之保護。保護的方法通常可以分為兩類，相同錯誤保護與非相錯誤保護(unequal error protect, UEP)。因為來源編碼後的二進制碼內每一位元之重要性通常並非相同，我們希望對不同重要性的位元予以不同程度的保護。因此，使用UEP會比使用相同錯誤保護來的好。雖然全搜尋法常被使用來作為位元指派的工具，但在許多例證中可發現其結果需大量的計算複雜度。
根據以上所述，我們提出了一個新的使用基因演算法之UEP位元指派方式。它可以有效降低搜尋最佳保護之計算複雜度。模擬結果證實，使用基因演算法之搜尋最佳保護方法，其效能不但優於相同錯誤保護法更接近於全搜尋法，且計算複雜度比全搜尋法少許多。

關鍵字

基因演算法 ； 非相同錯誤保護

並列摘要

The objective of this thesis is to perform the joint design of source and channel encoder, so that the end-to-end average distortion of communication system over a noisy channel can be minimized.
The objective of channel coding is error correction. After source encoding, there are two methods used for protecting the bit streams. One is equal error protection and the other is unequal error protection (UEP). Since different locations of bit streams delivered by source encoders may have different degree of importance, unequal error protection is desired. The full search scheme is usually utilized for implementing the bit allocation, which results in high computational complexity in many cases.
In lights of these facts, we present a new bit allocation algorithm based on the genetic programming (GP) for UEP. It can reduce the computational complexities effectively to determine the degree of protection at each location. Numerical results show that, the technique outperforms the equal error protection method. Moreover, as compared with exhaustive search algorithm for optimal unequal error protection, the GP technique attains comparable performance with significantly lower computational complexities.

並列關鍵字

unequal error protection ； genetic programming

參考文獻

[1]H. Abut, Vector Quantization, IEEE Press, 1990.

[2]J. C. Bezdek, Pattern Recognition with Fuzzy Objective Function Algorithms, Plenum Press, 1981.

[3]D. Bhandari, S. K. Pal and M. K. Kundu, “Image Enhancement Incorporating Fuzzy Fitness Function in Genetic Algorithm, ” in Proc. IEEE. Int. Conf. Fuzzy Systems, pp. 1408-1413, 1993.

[4]C. W. Chao, C. C. Chiu, P. C. Lu and C. H. Hsieh, “Codebook Design for Vector Quantization of Image Based on Fuzzy C-Means Clustering Algorithm,” Optical Engeineering, Vol. 32, pp. 580-587, Feb. 1997.

[5]L. H. Charles Lee, Convolutional coding : fundamentals and applications, Artech House, 1997.