Bit error probability (BEP) and word error probability (WEP) are used as reliability-based retransmission criteria in conjunction with hybrid ARQ (HARQ) protocols. Instead of exploiting an error-detecting code, the decision for a retransmission is based on the error probability of the decoded word, which can be calculated in or after the decoding process.

In this paper, we introduce a new method for calculating word error probabilities by means of a modified Viterbi decoder. The proposed method can be seen as an approximation of the reliability output Viterbi algorithm (ROVA). The results provided by the proposed algorithm are only slightly worse than that of the ROVA, given a significantly lower computational complexity especially for codes with large memory lengths. As a potential application, both algorithms are compared in reliability- based HARQ.

Physical layer constrained routing takes the conditions of the physical layers of a network into account when finding an appropriate route. The constraints are usually given by a routing metric, which describes the cost arising when transmitting via a link between two nodes. In nearly all wireless networks, on the physical layer channel coding is used to correct errors and to establish a reliable transmission in terms of error probabilities. In this paper, we introduce new routing metrics that are based on reliability information provided by a soft-output channel decoder. This is different from other approaches calculating error probabilities or link states by using channel statistics. In this paper, the actual error probability is calculated for a decoded packet rather than using a look-up table for a certain SNR or averages for certain channel statistics. The calculation and utilization of bit error probability, packet error probability, and the expected number of hops for a certain path or link are explained. Monte Carlo simulations indicate that using these routing metrics in wireless communications result in a significantly lower average packet delivery time than shortest path routing.