As in the GSM environment, GPRS measures the radio environment by analyzing the channel for carrier strength, bit error rate, etc. Performing these measurements takes time for a mobile station, which is of no concern in the speech world as the same coding is used all the time. In a packet-switched environment, it is essential to analyze the radio link quickly in order to adapt the coding toward the new environment. The channel analysis procedure that is used for GPRS makes the selection of the right coding scheme difficult since measurements for interference are performed only during idle bursts. As a result, measurements can only be performed twice during a 240-millisecond period. For EGPRS, the standard does not rely on the same “slow” measurement mechanism. Measurements are taken on each and every burst within the equalizer of the terminal, resulting in an estimate of the bit error probability (BEP). Estimated for every burst, the BEP is a reflection of the current C/I, the time dispersion of the signal and the velocity of the terminal. The variation of the BEP value over several bursts will also provide additional information regarding velocity and frequency hopping. A very accurate estimation of the BEP is then possible to achieve. A mean BEP is calculated per radio block (four bursts) as well as the variation (standard deviation of the BEP estimation divided by the mean BEP) over the four bursts. These results are then filtered for all radio blocks sent within the measurement period. This results in highly accurate measurements even during short measurement periods. Short measurement periods, in turn, enable quick reaction to changes in the radio environment. It is therefore possible to achieve a better and more flexible link adaptation for EGPRS.