摘要 |
PURPOSE:To apply real time processing by means of a microprocessor without any complicated hardware by forming a transmission packet in response to the output of the 3rd conversion table converted into an optimum data size. CONSTITUTION:An index (x) of a line bit error rate epsilon=10<-x> is obtained from a number of times of transmission L by using a conversion table 5 and a correction coefficient DELTA depending on the data size N is obtained by using a conversion table 6, the results are added by an adder 7 and a coefficient alpha<0<alpha<1) is multiplied by a multiplier 8. On the other hand, the index of error rate in a packet stored in a memory 9 is multiplied with a coefficient (1-alpha) by a multiplier 10. Outputs of the multipliers 8, 10 are added by an adder 11 and given to a conversion table 12 as an averaged error rate index x' and stored in the memory 9. The conversion table 12 outputs a data size N' optimum to the error rate from the averaged error rate index x'. A protocol section 3 determines the size of the packet sent next based on the said data size N'.
|