Proportional Fair Resource Allocation on an Energy Harvesting Downlink - Part II: Algorithms

TL;DR

This paper develops scalable heuristic algorithms for proportional fair resource allocation in energy harvesting downlink systems, closely matching optimal solutions with reduced computational complexity.

Contribution

It introduces two new heuristics, PTF and ProNTO, based on structural insights, for efficient and near-optimal resource allocation in energy harvesting broadcast systems.

Findings

Heuristics closely match optimal BCD performance

Proposed algorithms are computationally scalable

Structural properties of optimal policies are characterized

Abstract

In this paper, the proportionally fair allocation of time slots in a frame, as well as power level to multiple receivers in an energy harvesting broadcast system, is considered. Energy harvest times in a frame are assumed to be known at the beginning of that frame. The goal is to solve an optimization problem designed to maximize a throughput-based utility function that provides proportional fairness among users. An optimal solution of the problem was obtained by using a Block Coordinate Descent (BCD) method in earlier work (presented in Part I of this study). However, finding the optimal allocation entails a computational complexity that increases sharply in terms of the number of users or slots. In this paper, certain structural characteristics of the optimal power-time allocation policy are derived. Building on those, two simple and computationally scalable heuristics, PTF and ProNTO are proposed. Simulation results suggest that PTF and ProNTO can closely track the performance of the BCD solution.