Benchmarking Multi-agent Reinforcement Learning-based Access Control using Real-world IoT Traffic

Tien-Thanh Le, Yusheng Ji, Linh Truong, John C.S Lui

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussaPosterScientificvertaisarvioitu

Abstrakti

In recent years, the proliferation of Internet of Things (IoT) devices and applications has given rise to massive Machine Type Communication (mMTC), a critical domain within next-generation wireless communication. mMTC is characterized by a massive number of low-power devices generating data sporadically, posing a formidable challenge in efficiently allocating the limited physical wireless channel resources among them. Conventional centralized allocation methods prove infeasible in mMTC scenarios due to the overwhelming volume of control messages relative to the actual data transmissions required by mMTC devices. To address this challenge, there has been a notable upsurge in research endeavors dedicated to enhancing fully distributed channel access protocols for mMTC, leveraging the power of cooperative multi-agent deep reinforcement learning. However, an existing research gap becomes evident when we observe that prior studies predominantly tested their algorithms under conditions of either saturated traffic or simulated traffic models. This limited scope fails to accurately capture the dynamics of real-world mMTC networks. In this study, we bridge this gap by utilizing real-world IoT traffic data from a network operator in Vietnam. We conduct a comparative analysis of the delay and fairness of four fully distributed and dynamic channel access strategies: (1) the classical multi-channel ALOHA, (2) a heuristic enhancement of mult-channel ALOHA to improve fairness, and (3) the centralized training decentralized execution MADRL. This research aims to provide valuable insights into the practicality and effectiveness of these strategies in addressing the unique challenges posed by mMTC scenarios in real-world wireless communication environments.
AlkuperäiskieliEnglanti
TilaJulkaistu - 2023
OKM-julkaisutyyppiEi oikeutettu
TapahtumaInternational Symposium on Computing and Networking - Matsue, Japani
Kesto: 28 marrask. 20231 jouluk. 2023
Konferenssinumero: 11

Conference

ConferenceInternational Symposium on Computing and Networking
Maa/AlueJapani
KaupunkiMatsue
Ajanjakso28/11/202301/12/2023

Sormenjälki

Sukella tutkimusaiheisiin 'Benchmarking Multi-agent Reinforcement Learning-based Access Control using Real-world IoT Traffic'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä