A steady-state model for energy-efficient packet processing engines under mixed traffic

In this contribution, we focus on energy-aware devices able to reduce their energy requirements by adapting their performance. We propose an extension of an analytical model to accurately represent the impact of green networking technologies - namely, low power idle (LPI) and adaptive rate (AR) - on network- and energy-related performance indexes. We focus on a packet-processing engine within a switching node, where LPI and AR are applied in the presence of multiple services that require different processing times of the packet headers. Given a discrete set of processing times with their probabilities, we extend our previous modeling framework based on an M^x/D/1/SET queueing system to take into account the presence of multiple service times. We use the model to perform various numerical computations to evaluate energy consumption and packet delay. The results highlight how the proposed model can effectively be used to produce analytical estimates of performance indexes that can be the basis for the optimization of energy consumption-performance tradeoffs.