Reconfigurable Multiple Access Schemes for URLLC-mMTC Coexisting Networks

Since the fifth generation (5G) wireless networks, the focus of machine-type communication (MTC) has shifted to emerging applications such as the Industrial Internet of Things (IIoT) and smart cities. Massive MTC (mMTC) and ultra-reliable low-latency communication (URLLC), two distinct MTC service classes targeting scalability and reliability with low latency, are expected to be merged to support diversified and more stringent quality of service (QoS) requirements in MTC beyond 5G.With limited radio resources, supporting these contrasting requirements for URLLC-mMTC coexisting networks is challenging.Multiple access schemes should consider traffic characteristics and QoS requirements to adaptively prioritize devices with urgent traffic demands while providing massive connections.Additionally, as autonomous robots are increasingly utilized in various verticals, devices with mobility are likely to cooperate as clusters or operate in out-of-coverage areas.As a result, multi-hop relay connections should be enabled to address the coverage problem and support information exchanges in different network structures.In this thesis, we propose access schemes that combine grant-based (GB) and grantfree (GF) access techniques and equip massive multiple-input-multiple-output (mMIMO) antenna array at the access point (AP) to provide reliable and low-latency transmissions while increasing device connections, targeting different cases of URLLC-mMTC coexisting networks under manifold traffic and network configurations.First, we develop an access scheme to support scalable URLLC networks that contain a large number of devices with various latency requirements.The optimization problem is formulated to maximize the system throughput while minimizing packet delay violations.Considering that the objective function is hard to derive from unknown event-based traffic statistics and different delay constraints, both single-and multi-agent reinforcement-learning (RL)-based algorithms are designed to exploit the traffic correlation and delay status to grant resources for devices with high traffic demands and stringent delay requirements and let other devices contend for transmissions.Also, an action correction mechanism is introduced to ensure better convergence and safe operation of the RL agents.Second, we propose an access scheme for critical mMTC networks, which supports URLLC services for a varying fraction of massively connected devices, and each device has regular and delay-sensitive alarm traffic simultaneously instead of a homogenous traffic type.Additionally, unlike the previous single-hop network, we consider device clusters where devices transmit to the AP by two-hop relay connections through cluster leaders, and we allocate resources dynamically to mitigate inter-cluster interference.Then, the First and foremost, I would like to express my sincere gratitude to my supervisor, Professor Tho Le-Ngoc, for his tireless guidance,

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.