Optimal Radio Resource Allocation for Hybrid Traffic in Cellular Networks: Traffic Analysis and Implementation

TL;DR

This paper analyzes and compares the transmission overhead and sensitivity of centralized and distributed radio resource allocation architectures for hybrid traffic in cellular networks, supported by simulations and real-world implementation.

Contribution

It provides a detailed transmission overhead analysis, derives lower bounds, and evaluates sensitivity to user and application changes for both architectures, including real-world implementation.

Findings

Centralized architecture has significantly lower transmission overhead than distributed.

Lower bounds for transmission overhead are derived for both architectures.

Simulation results verify the analysis and implementation effectiveness.

Abstract

In part I of this paper, a distributed and a centralized architecture for an optimal radio resource allocation aware of the traffic delay-tolerance nature, user subscription type, and application usage variations were developed. In the current article, a transmission overhead analysis of the aforementioned distributed and a centralized architectures is investigated and it is proved that the centralized scheme endures a significantly lower transmission overhead than does the distributed approach. Furthermore, the lower bounds of the transmission overhead for both the centralized and the distributed architectures are derived. Moreover, a sensitivity analysis of the resource allocation procedures of the aforesaid centralized and distributed architectures to the changes in the number of users in the system is presented. Besides, a sensitivity analysis of the centralized and distributed approaches to the temporal changes in application usages are investigated. Ultimately, the transmission overhead and sensitivity relevant statements are verified through appropriate simulations. And last but not the least, a real-world implementation of the resource allocation methods developed in Part I is provided.