Learning Autonomy in Management of Wireless Random Networks

TL;DR

This paper introduces a novel distributed message-passing neural network (DMPNN) for managing wireless networks with random topologies, enabling robust and universal optimization through iterative message sharing.

Contribution

The paper develops a flexible DNN formalism called DMPNN that is topology-independent and capable of learning optimal distributed coordination strategies in wireless networks.

Findings

DMPNN achieves convergence in various network configurations.

Numerical results show DMPNN outperforms traditional optimization methods.

DMPNN demonstrates universality and robustness in random topologies.

Abstract

This paper presents a machine learning strategy that tackles a distributed optimization task in a wireless network with an arbitrary number of randomly interconnected nodes. Individual nodes decide their optimal states with distributed coordination among other nodes through randomly varying backhaul links. This poses a technical challenge in distributed universal optimization policy robust to a random topology of the wireless network, which has not been properly addressed by conventional deep neural networks (DNNs) with rigid structural configurations. We develop a flexible DNN formalism termed distributed message-passing neural network (DMPNN) with forward and backward computations independent of the network topology. A key enabler of this approach is an iterative message-sharing strategy through arbitrarily connected backhaul links. The DMPNN provides a convergent solution for iterative coordination by learning numerous random backhaul interactions. The DMPNN is investigated for various configurations of the power control in wireless networks, and intensive numerical results prove its universality and viability over conventional optimization and DNN approaches.