Performance evaluation of GPRS/EDGE networks: A novel discrete-time Markov model

摘要

In this paper, we focus on a simple yet important objective: to provide simple and accurate models for GPRS/EDGE radio engineering. In [2], we have developed an Erlang-like law based on a discrete-time Markovian model that captures the detailed behavior of the radio allocator, when a finite number of users are performing infinite ON/OFF sessions in the cell. In this paper, we extend this work by developing new computational formulas that aim at deriving more than average performance parameters. More precisely, we propose in this paper an original recursive formula that can accurately calculate transfer time and throughput distributions. These distributions are quite interesting for network engineering as they provide a view on the full statistics of the end-user quality of experience. Validation of the method and preliminary conclusions on EDGE/GPRS dimensioning are finally discussed.