Upper and Lower Estimates of Frame Error Rate for Convolutional Codes

In this paper, we suggest a new approach for frame error rate (FER) evaluation for the convolutional codes. We consider binary symmetric channel and Viterbi decoding of convolutional codes. Convolutional codes we studied here have code rate one half and recursive encoder. We precisely define active distances and their distance spectrum for the code. Unique distance properties allow us to construct estimates for error bursts probabilities and for the probability of erroneous decoding of the convolutional code. We derive upper and lower estimates for error burst probabilities. Based on these expressions we suggest upper and lower bounds for FER performance.