Markov-modulated queueing network for mobile traffic aggregation with threshold-controlled buffers

Source abstract

We study the problem of data transmission from mobile platforms operating in high-speed transit between cellular base stations, under conditions of unstable and intermittent connectivity. Conventional queueing and connectivity models often fail to capture the combined effects of rapidly changing signal conditions, finite buffer capacity, and dynamic topology. We aim to develop a tractable yet expressive model that integrates stochastic link availability, queue dynamics, and buffer control. We propose a mathematical framework based on queues whose service intensities are modulated by a continuous-time Markov chain (CTMC) representing signal conditions along a high-speed trajectory. The core subsystem is a two-stage (aggregation $\to$ transmission) tandem queue in which transmission can be slowed down (weak coverage) or fully interrupted (outage). We construct a three-zone coverage CTMC calibrated to the \emph{geometric mean} residence times, and obtain a stability condition $\lambda