[Paper Review] A Centralized Metropolitan-Scale Radio Resource Management Scheme

Zhiyi Zhou, Dongning Guo|arXiv (Cornell University)|Aug 7, 2018

Advanced MIMO Systems Optimization24 references4 citations

TL;DR

This paper proposes a centralized, optimization-based radio resource management framework for metropolitan-scale wireless networks with thousands of access points and users. By formulating joint spectrum allocation, user association, and power control as a network utility maximization problem, it proves that optimal resource allocation can be achieved with a sparse, piecewise-constant power profile and develops a fast, convergent algorithm that computes solutions in closed form, significantly improving throughput and reducing average packet delay compared to existing schemes.

ABSTRACT

This work studies centralized radio resource management in metropolitan area networks with a very large number of access points and user devices. A central controller collects time-averaged traffic and channel conditions from all access points and coordinates spectrum allocation, user association, and power control throughout the network on an appropriate timescale. The timescale is conceived to be seconds in today's networks, and it is likely to become faster in the future. The coordination problem in each time epoch is formulated as a network utility maximization problem, where any subset of access points may use any parts of the spectrum to serve any subsets of devices. It is proved that the network utility can be maximized by an extremely sparse spectrum allocation. By exploiting this sparsity, an efficient iterative algorithm with guaranteed convergence is developed, each iteration of which is performed in closed form. The proposed centralized optimization framework can incorporate a broad class of utility functions that account for weighted sum rates, average packet delay, and/or energy consumption, along with very general constraints on transmission powers. Numerical results demonstrate the feasibility of the algorithm for networks with up to 1,000 access points and several thousand devices. Moreover, the proposed scheme yields significantly improved throughput region and average packet delay comparing with several well-known competing schemes.

Motivation & Objective

• To address the limitations of distributed, non-adaptive spectrum allocation in large-scale metropolitan wireless networks with dense access points and high user loads.
• To design a centralized framework that jointly optimizes spectrum allocation, user association, and power control for improved network utility.
• To develop a low-complexity, convergent algorithm that computes optimal or near-optimal solutions in real time for large-scale deployments.
• To demonstrate the feasibility and performance gains of centralized optimization in networks with up to 1,000 APs and thousands of devices.

Proposed method

• Formulates the joint resource allocation problem as a network utility maximization (NUM) problem with continuous spectrum allocation, dynamic user association, and variable power control.
• Proves that the optimal solution requires only a sparse, (k+n+1)-piecewise constant power allocation, where k is the number of subbands and n is the number of users.
• Applies Carathéodory’s theorem to show that any feasible solution can be represented as a convex combination of at most k+n+1 extreme points, enabling efficient computation.
• Develops an iterative algorithm with closed-form updates per iteration, ensuring fast convergence and low computational complexity.
• Supports a broad class of utility functions, including weighted sum rates, delay, and energy efficiency, with general power constraints.
• Employs a central controller that collects time-averaged traffic and channel state information from all APs and computes global allocation plans on a timescale of seconds.

Experimental results

Research questions

• RQ1 Can centralized optimization achieve superior performance in large-scale metropolitan wireless networks compared to distributed or heuristic schemes?
• RQ2 What is the structural sparsity of the optimal power allocation in joint spectrum, user, and power allocation problems?
• RQ3 Can a low-complexity, convergent algorithm be designed to compute near-optimal solutions in real time for networks with thousands of APs and users?
• RQ4 How does the proposed scheme compare in terms of throughput region and average packet delay to existing schemes like full reuse or fractional reuse?

Key findings

• The optimal solution to the joint resource allocation problem requires only a (k+n+1)-piecewise constant power allocation, enabling significant sparsity and computational efficiency.
• The proposed iterative algorithm converges to a locally optimal solution with closed-form updates per iteration, ensuring low computational complexity.
• Numerical results confirm feasibility in networks with up to 1,000 access points and several thousand user devices.
• The scheme achieves a significantly larger throughput region and lower average packet delay compared to conventional schemes like full reuse and fractional frequency reuse.
• The framework supports diverse utility functions, including weighted sum rates, delay, and energy efficiency, under general power constraints.
• The timescale of centralized coordination is feasible at the second timescale today and is expected to scale down to sub-frame levels with future technology advances.