A Waterfilling Algorithm for Multiple Access Point Connectivity with Constrained Backhaul Network

TL;DR

This paper introduces a waterfilling power allocation algorithm for UEs with multiple AP connections over orthogonal channels, optimizing data rates despite constrained backhaul links without requiring detailed network topology knowledge.

Contribution

It presents a novel low-overhead, dynamic power allocation scheme that approaches optimal rates in multi-AP connectivity scenarios with constrained backhaul networks.

Findings

Achieves near-optimal data rates with minimal backhaul feedback.

Operates effectively without detailed backhaul network topology information.

Adapts dynamically to uplink conditions for improved performance.

Abstract

The 4G/5G paradigm offers User Equipment (UE) simultaneous connectivity to a plurality of wireless Access Points (APs). We consider a UE communicating with its destination through multiple uplinks operating on orthogonal wireless channels with unequal bandwidth. The APs are connected via a tree-like backhaul network with destination at the root and non-ideal link capacities. We develop a power allocation scheme that achieves a near-optimal rate without explicit knowledge of the backhaul network topology at transmitter side. The proposed algorithm waterfills a dynamic subset of uplinks using a low-overhead backhaul load feedback.