Access Point Density and Bandwidth Partitioning in Ultra Dense Wireless Networks

TL;DR

This paper analyzes how access point density and bandwidth partitioning affect user rates in dense wireless networks, providing analytical insights into optimal configurations and minimum density requirements.

Contribution

It introduces an analytical model for user rate distribution considering multiple access and SIR outage, addressing limitations of prior models in dense networks.

Findings

User rate outage depends on bandwidth partitions and multiple access scheme.

Lower bounds on optimal number of partitions are established.

Closed-form upper bounds for access point density are derived.

Abstract

This paper examines the impact of system parameters such as access point density and bandwidth partitioning on the performance of randomly deployed, interference-limited, dense wireless networks. While much progress has been achieved in analyzing randomly deployed networks via tools from stochastic geometry, most existing works either assume a very large user density compared to that of access points which does not hold in a dense network, and/or consider only the user signal-to-interference-ratio as the system figure of merit which provides only partial insight on user rate, as the effect of multiple access is ignored. In this paper, the user rate distribution is obtained analytically, taking into account the effects of multiple access as well as the SIR outage. It is shown that user rate outage probability is dependent on the number of bandwidth partitions (subchannels) and the way they are utilized by the multiple access scheme. The optimal number of partitions is lower bounded for the case of large access point density. In addition, an upper bound of the minimum access point density required to provide an asymptotically small rate outage probability is provided in closed form.