In-service link quality estimation for link adaptation algorithms, applied to GSM

This paper investigates link quality estimates for GSM, in order to use them as an input for link adaptation algorithms. Two channel heuristics are extracted from the GSM transmission chain. Their suitability as triggering criteria is assessed by mapping them onto a bit error rate estimate. Then a real bit error rate characterisation is outlined, and the prediction error obtained is compared to the prediction error observed for the bit error rate estimates extracted from the heuristics. Results show that the accuracy of these estimates is similar to what could be achieved if real bit error rate measurements were available, being suitable for application to link adaptation algorithms