Analyzing the Synergy between Broadcast Polling and Piggybacking in WiMAX Networks

TL;DR

This paper develops a comprehensive model to analyze the combined effects of broadcast polling and piggybacking in WiMAX networks, considering realistic bandwidth and queueing constraints, validated through simulations.

Contribution

It introduces a generic, parameterizable model that accurately captures the interaction between polling and piggybacking in WiMAX, applicable to various network conditions.

Findings

Model accurately predicts system performance in different load conditions

Piggybacking offers benefits and drawbacks depending on network state

Validation confirms the model's effectiveness across scenarios

Abstract

In this work we present a model for analyzing the combined use of broadcast polling and piggybacking in Worldwide Interoperability for Microwave Access (WiMAX) networks. For an accurate analysis of piggybacking, the model focuses on the realistic case of limited up-link bandwidth and non-trivial queueing capability at the subscriber stations. We first model the activity of a subscriber station using a Markov chain and its queue as an M/G/1 system with vacations in order to facilitate the analysis of the piggyback mechanism. We then derive a set of fixed point equations that describe not only the contention process at the network level but also bandwidth allocation to contending and piggybacked requests. Our model uses a minimal set of assumptions and is generic in the sense that it is customizable through a set of parameters. It can also reproduce the system performance in both saturated and non-saturated conditions. After validating our analysis through extensive simulations, we shed light on the aspects of the synergy between broadcast polling and piggybacking and unveil the pros and cons of using the latter.