An Efficient Searching Method for Best Convolutional Codes Using Iterative Calculation of Weight Spectrum

摘要

To find the best convolutional code, we calculate weight spectrum of all code candidates. Since the number of codes grows exponentially with constraint length, it becomes very difficult to obtain the best code with relatively large constraint length. We show a new method of searching for the best convolutional code efficiently. The method searches first adequately restricted range of the code tree and iterates the search with expanding the range gradually to obtain weight spectrum of the code. In this search it discards codes with bad performance without calculating weight spectrum completely. We show that the proposed method can find the best code more efficiently than those presented before do.