Performance analysis of Compound TCP with AQM

We study Compound TCP (C-TCP), the transport protocol in the Windows operating system, in different buffer sizing regimes along with Drop-Tail and Random Exponential Marking (REM). The buffer sizing regimes we focus on are the widely deployed bandwidth-delay rule and a small buffer regime. The performance metrics we consider are stability of the queue size, queuing delay, link utilisation and packet loss. We analyse the following models: (i) a non-linear model for C-TCP with Drop-Tail and small buffers, (ii) a stochastic variant of REM along with C-TCP, and (iii) the original REM proposal as a continuous time non-linear model with delayed feedback. We develop conditions to ensure local stability and show that variations in system parameters can induce a Hopf bifurcation, leading to the emergence of limit cycles. With Drop-Tail, and small buffers, the Compound parameters and the buffer size play a key role in maintaining stability. With the stochastic variant of REM, larger thresholds for marking/dropping packets can destabilise the system. With REM, an increase in the feedback delay, or variations in the queue management parameters, can induce a Hopf bifurcation. Design guidelines for Compound parameters, based on analysis, to ensure stability are provided. Packet-level simulations corroborate some of the analysis.