Vehicular Network Spectrum Allocation Using Hybrid NOMA and Multi-agent Reinforcement Learning
The recent years have seen a proven impact of the reinforcement learning use in many applications which showed tremendous success in solving many decision-making paradigms in machine learning. Most of the successful applications involves the existence of more than one agent, which makes it fall into...
Saved in:
Main Authors: | , , , , , |
---|---|
Other Authors: | |
Format: | Conference Paper |
Published: |
Springer Nature
2024
|
Subjects: | |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Universiti Tenaga Nasional |
Summary: | The recent years have seen a proven impact of the reinforcement learning use in many applications which showed tremendous success in solving many decision-making paradigms in machine learning. Most of the successful applications involves the existence of more than one agent, which makes it fall into the multi-agent category, taking autonomous driving as an example of these applications. We know that today�s Internet of Vehicles (IoVs) consists of multi-communication patterns which work efficiently in keeping all the IoV network components connected. With regards to sharing the frequency spectrum, applying Non-Orthogonal Multiple Access (NOMA) communication built over deep deterministic policies gradients (DDPG) scheme to cope with the rabid erratic channels conditions due to fast mobility nature of vehicles network has proven promising results. In this paper the framework of NOMA communication-based DDPG and multiple agent reinforcement learning approach (MARL) are discussed in brief, and then, the performance evaluation of DDPG scheme compared with MARL and random spectrum allocation approaches for vehicular network spectrum and resources allocation is analysed. � 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG. |
---|