5G Routing Interfered Environment

TL;DR

This paper introduces 5GRIE, a Python-based simulation environment for testing 5G routing algorithms under interference, supporting reinforcement learning and heuristic methods to optimize packet routing in high-density networks.

Contribution

It presents a novel environment for simulating 5G routing with interference, enabling comparison of reinforcement learning and heuristic algorithms.

Findings

Profitable algorithm demonstrates effective routing performance.

Environment supports diverse algorithm testing including Deep RL and heuristics.

Facilitates research on interference-aware 5G routing strategies.

Abstract

5G is the next-generation cellular network technology, with the goal of meeting the critical demand for bandwidth required to accommodate a high density of users. It employs flexible architectures to accommodate the high density. 5G is enabled by mmWave communication, which operates at frequencies ranging from 30 to 300 GHz. This paper describes the design of the 5G Routing Interfered Environment (5GRIE), a python-based environment based on Gym's methods. The environment can run different algorithms to route packets with source and destination pairs using a formulated interference model. Deep Reinforcement Learning algorithms that use Stable-Baselines 3, as well as heuristic-based algorithms like random or greedy, can be run on it. Profitable is an algorithm that is provided.