Effective minimum free distance for convolutional codes

The minimum free distance (MFD) of convolutional error-correcting codes is a powerful and well-known parameter for indicating the behaviour of the communication system via an AWGN channel with medium and high SNR. But in many cases it happens that the MFD theory fails. The letter provides some examples where codes with a high MFD are being defeated by codes with lower MFD. Then a new method called the effective minimum free distance is introduced to achieve greater simplicity and to strengthen the MFD theory. Software simulation for some codes with constraint lengths 3 and 4 demonstrate this idea. The Viterbi decoding algorithm is used for the simulation program. The truncated generator function of a code is modified into a new, effective terms form that presents a very practical measure on the error correctability of the code.