Joint Fronthaul Load Balancing and Computation Resource Allocation in Cell-Free User-Centric Massive MIMO Networks

TL;DR

The work addresses scalable cell-free MIMO with constrained fronthaul by proposing a joint fronthaul load balancing and cluster-processor placement framework, cast as MILPs for both full- and half-duplex fronthaul and augmented by rate-distortion modeling of UL fronthaul quantization. It introduces a comprehensive CF-mMIMO model with user-centric RU clusters, decentralized cluster processors in DUs, and an explicit fronthaul topology, enabling exact optimization of routing and computation placement. A Bussgang-based UL quantization model links fronthaul rate to achievable UL/DL PHY rates, revealing an optimal distortion level under balanced UL/DL traffic and the practical advantage of coarse quantization. Numerical results under 3GPP urban micro pathloss illustrate the tradeoffs between fronthaul load and PHY SE, showing that UL fronthaul is typically the bottleneck and that half-duplex fronthaul can reduce overall load, offering actionable insights for implementing scalable CF-mMIMO with O-RAN principles.

Abstract

We consider scalable cell-free massive multiple-input multiple-output networks under an open radio access network paradigm comprising user equipments (UEs), radio units (RUs), and decentralized processing units (DUs). UEs are served by dynamically allocated user-centric clusters of RUs. The corresponding cluster processors (implementing the physical layer for each user) are hosted by the DUs as software-defined virtual network functions. Unlike the current literature, mainly focused on the characterization of the user rates under unrestricted fronthaul communication and computation, in this work we explicitly take into account the fronthaul topology, the limited fronthaul communication capacity, and computation constraints at the DUs. In particular, we systematically address the new problem of joint fronthaul load balancing and allocation of the computation resource. As a consequence of our new optimization framework, we present representative numerical results highlighting the existence of an optimal number of quantization bits in the analog-to-digital conversion at the RUs.

Figures (7)

• Figure 1: An example network with $K = 5$ UEs, $L = 4$ RUs, $Q = 3$ routers, and $N = 2$ DUs.
• Figure 2: Example of the data in a resource block with $T$ signal dimensions, where blue and white boxes represent UL pilot and data signals, respectively.
• Figure 3: Left: The fronthaul links between RUs and routers. Right: The fronthaul links between routers and DUs.
• Figure 4: UL/DL PHY SE vs. distortion level (a) and corresponding fronthaul UL/DL load (b).
• Figure 5: Total fronthaul load (a) and optimization objective function (b) vs. fronthaul quantization distortion level.

Theorems & Definitions (11)

• Example 1
• Definition 1
• Lemma 1
• proof
• Lemma 2
• proof
• Remark 1
• Remark 2
• Remark 3
• Remark 4