WATM operation optimisation based on effect of FEC code rate and ARQ retransmission

This paper applies a new methodology to different coding schemes for WATM efficiency evaluation. This methodology is based on the separation of error based efficiency (EBE), which is the ratio between the mean raw data rate and the raw data rate plus all overheads associated with error corrections and retransmission of erroneous blocks of data, from queuing based efficiency (QBE), which is the ratio between the error free data rate and the overall channel capacity. We compare between the efficiency of convolutional coding (in the case of binary transmission and high level modulation transmission) and turbo coding. We show that under most value of signal to noise ratio (SNR), turbo coding is the most efficient. We also provide means to support optimal adaptive setting of the code rate and demonstrate the improvement gained by implementing adaptive forward error correction code rate. We further demonstrate that, for certain realistic SNR conditions, the wastage due to retransmissions can be very high.