The Analysis And Research On Transmitting End Delay Of GBN-ARQ Protocols With Random Feedback Delay

Since the emergence of the telecommunications network, data communication techniques has been used widely, and developed rapidly. Compared with other means of communication, the most characteristic feature of the data communication is the high reliability. Meanwhile, in a communication system, transmission delay is an important parameter of the network performance,which affects the flow control of the transmitter. In order to ensure the quality of the transmission and to improve the system throughput performance, it is an important researching direction to apply the error-control techniques which request a precise evaluation of the transmission delay.By adaptive modulating and coding techniques, which are imported for link adaptation, a transmitter can acquire the channel state through the feedback channel, thus it is able to adaptively change the modulating and coding scheme to maximize the spectrum efficiency. In this paper, a widely used error-control technique-Go-Back-N Automatic Retransmission Request (GBN-ARQ) protocol is researched. Because the feedback delay is a random variable, a markov model for the data packets transmission which applies the link adaptation techniques. This model concerns an important fact,the parameter of the wireless channel is time-varying,it can reduce the system performance,and with the development of the date services, spectrum resources are increasingly strained.The state transition model is presented while the link is adaptive. On this basis, the markov model of system states is established while the feedback delay is random variable.Then the transmitting end delay statistics are obtained by using the matrix-geometric methods,so the best flow control can be received,and the performance of the entire communication system is improved.We validate the transmitting end delay model by simulation. The calculated result of the packets delay at transmitter is same with the simulation result.In this paper, the calculating process for the transmitting end delay fully concerns the random time-varying feedback delay, meanwhile the results are precise. It provides a reliable way to evaluate the time-delay for a wireless network. Meanwhile, by comparing the simulated results, we analysis the impact on transmitting end delay in different system parameters. The conclusion is that there are no best solution to both the doppler frequency shift and the packet arrival rate of the transmitter.