?
Methodology for Experimental Comparison of Redundancy for QUIC and TCP Protocols
The paper presents a comparison of the redundancy introduced when transmitting data using the QUIC and TCP protocols. Issues concerning the impact of redundancy—including that introduced by error-corretcing codes—on protocol efficiency are discussed. To quantitatively assess protocol performance, the protocol efficiency coefficient is introduced, defined as the ratio of the volume of payload data to the total volume of data transmitted over the network during protocol operation. Experiments are described in which a file containing pseudorandom data is transmitted over the network, the resulting network traffic is analyzed, and an estimate of the protocol efficiency coefficient is computed based on the analysis. It is noted that using a file with pseudorandom data eliminates the influence of file content on protocol behavior, while comparing protocol efficiency based on the estimated efficiency coefficient helps reduce the impact of network conditions on the comparison results. Experimental results are presented for scenarios in which no packet losses occur. Conclusions are drawn regarding which protocol demonstrates higher efficiency in such cases, and assumptions are made about performance in other scenarios.
