Average Packet Delay Analysis for a Mobile User in a Two-Tier Heterogeneous Cellular Network

In this paper, we analyze the average packet queueing delay of a slow-moving user in a macro-femto two-tier heterogeneous cellular network. We model the user as a queue that experiences vertical handover between macro and femto base stations. We define the queue states progression at each handover interval as a 2-D discrete time Markov chain. We determine the Markov chain's steady-state probability distribution so that the average packet delay can be calculated using the Little's Theorem. With the model, we can analyze the effects on packet delay as caused by variations in coverage area dwell time, packet service rate, and probability of entering a femtocell. We have verified the analytical model against results from random-event simulations. We find that the average packet delay can be lowered by deploying more femtocells only if the time a user stays in a femtocell is longer than a certain value. The model can help us in avoiding unnecessary vertical handovers, as well as in deciding the number and size of femtocells to achieve a desired packet delay.