RAS1 – Radio Access and Softwarisation
Thursday, 8 June 2023, 16:00-17:30, Room R22-R23
Session Chair: Catalina Stan (Eindhoven University of Technology, The Netherlands)
Designing Medium Access Control Protocol Sequences Through Deep Reinforcement Learning
Cedric Adjih (INRIA, France); Chung Shue Chen (Nokia Bell Labs, France); Chetanveer Sharma Gobin (INSA Lyon, France); Iman Hmedoush (Nokia Bell Labs, France)
This work aims to design protocol sequences through deep reinforcement learning (DRL). Protocol sequences are periodic binary sequences that define multiple access control among users, introduced for systems considering collision channel without feedback (CCw/oFB). In this paper, we leverage the recent advancement of DRL methods to design protocol sequences with desirable new properties, namely Throughput Maximizing User-Irrepressible (TMUI) sequences. TMUI has two specific properties: (i) user-irrepressibility (UI), and (ii) maximizing the minimum individual throughput among the users. We assumed that the transmission channel is divided into time slots and the starting time of each user in joining the system is arbitrary such that there exist random relative time offsets. We use a DRL approach to find TMUI sequences. We report the obtained TMUI protocol sequences and conduct numerical studies comparing TMUI against slotted ALOHA. Simulation results also show that the new medium access control (MAC) protocol does hold the UI property and can achieve substantially higher minimum individual user throughput, under the same system parameters.
Hybrid Radio Resource Management Based on Multi-Agent Reinforcement Learning
Lam Ngoc Dinh (CEA-LETI, France); Mickael Maman (CEA-Leti, France); Emilio Calvanese Strinati (CEA-LETI, France)
In this paper, we propose a novel hybrid grant-based and grant-free radio access scheme. We provide two multi-agent reinforcement learning algorithms to optimize a global network objective in terms of latency, reliability and network throughput: Multi-Agent Deep Q-Learning (MADQL) and Multi-Agent Deep Deterministic Policy Gradient (MADDPG). In MADQL, each user (agent) learns its optimal action-value function, which is based only on its local observation, and performs an optimal opportunistic action using the shared spectrum. MADDPG involves the attached gNB function as a global observer (critic), which criticizes the action of each associated agent (actor) in the network. By leveraging centralised training and decentralised execution, we achieve a shared goal better than the first algorithm. Then, through a system level simulation where the full protocol stack is considered, we show the gain of our approach to efficiently manage radio resources and guarantee latency.
Hybrid Multiple Access Scheme Employing NOMA and OMA Simultaneously Under Non-Uniform User Distribution and Multiple Transmit Antennas
Nozomi Sasaki, Fuga Tanaka, Shuhei Saito, Hirofumi Suganuma and Fumiaki Maehara (Waseda University, Japan)
The fifth generation (5G) mobile communication systems provide increased functionality and performance, and can be implemented through orthogonal multiple access (OMA) and non-orthogonal multiple access (NOMA). Both methods present advantages and disadvantages under certain conditions. In this study, we analyze the performance of a hybrid multiple access scheme using NOMA and OMA simultaneously in a practical scenario under non-uniform user distribution and multiple transmit antennas through computer simulations to verify its effectiveness. The proposed hybrid multiple access scheme avoids the NOMA-specific performance deterioration caused due to small channel gain differences between users by applying OMA within the same bandwidth. Additionally, we implement resource allocation to account for both the channel gain and desired user traffic volume considering the diversified wireless service for Beyond 5G (B5G). We apply the Thomas cluster process as user distribution and maximum ratio transmission (MRT) as transmit diversity, and elucidate the effects of the non-uniformity of user distribution and an increase in the number of transmit antennas on the performance of the proposed hybrid multiple access scheme.
5G Radio Resource Allocation for Communication and Computation Offloading
Catalina Stan and Simon Rommel (Eindhoven University of Technology, The Netherlands); Ignacio de Miguel (Universidad de Valladolid, Spain); Juan José Vegas Olmos (NVIDIA Corporation, Israel); Ramón J. Durán Barroso (Universidad de Valladolid, Spain); Idelfonso Tafur Monroy (Eindhoven University of Technology, The Netherlands)
Edge computing is envisioned as a key enabler in future cellular networks by bringing the computing, networking and storage resources closer to the end users and enabling offloading for computation-intensive or latency-critical tasks coming from the emerging 5G/6G applications. Such technology also introduces additional challenges when it comes to deciding when to offload or not since the dynamic wireless environment plays a significant role in the overall communication and computation costs when offloading workload to the nearby edge nodes. In this paper, we focus on the communication cost in computation offloading via wireless channels, by formulating an α-fair utility-based radio resource allocation (RRA) problem tailored for offloading in a multi-user urban scenario where the uplink connection is the main focus. We begin by modeling the wireless channel with large- and small-scale fading at both lower and millimetre-wave frequencies, followed by data rate calculation based on 3GPP for a more realistic approach. Then, while assessing the fairness of the RRA, we simulate the resource allocation framework while taking into account both users who need to offload and users who are only interested in high downlink data rates. Simulation results show that the weighted proportional fairness method adapted for computation offloading can provide a good trade-off between fairness and performance compared to other benchmark schemes.
5G in the 3.8-4.2 GHz Band: Coexistence with Fixed Satellite Service Earth Stations In-Band and IMT-2020 in Adjacent Band
Theodoros Spathopoulos (Nokia, United Kingdom (Great Britain)); Jianhua Liu (Nokia Shanghai Bell, China); Fabiano Chaves (Nokia, USA)
This paper analyses the coexistence of low and medium power 5G Base Stations (BSs) operating in the 3.8- 4.2 GHz to provide local area network connectivity with Fixed Satellite Service Earth stations in the same band as well as IMT2020/5G services in the adjacent band below 3.8 GHz. This work follows the developments of the current regulatory discussion in Europe, providing an analysis of how different configurations for 5G BSs operating in the 3.8-4.2 GHz band, including Active Antenna System (AAS) BSs with increased radiated power and non-AAS BSs, can impact the resulting separation distances required to coexist with incumbent services. It is observed that AAS BSs operating with considerably higher power than nonAAS BSs lead to similar coexistence conditions.