Performances of convolutional encoders trellises termination methods, components of MNBTCs

The closure/termination of convolutional encoders (CE) trellises can be achieved through several strategies, each offering advantages and disadvantages compared with the others. Because multi-non-binary turbo codes (MNBTCs) operate with shorter lengths of blocks of data than other families of turbo codes (TC), the strategy of terminating component CE trellises has a stronger influence on the encoding rate, and, implicitly, on the bit/frame error rate (B/FER) performance. This paper compares the performance of B/FER versus signal to noise ratio (SNR) for the main strategies for terminating the CE trellises, components of MNBTCs.