- Title
- An enhanced radio resource management scheme for machine-to-machine communications over LTE networks
- Creator
- Afrin, Nusrat
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2022
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- The importance of Machine-to-Machine (M2M) communications has increased manifold in the recent years due to being a key enabler of the Internet of Things (IoT) comprising of billions of devices exchanging information with different levels of urgency, security and reliability requirements. Although the M2M communications encompass a wide spectrum of services with different overlapping technologies, one of the major challenges lies in ensuring ubiquitous network access to machine-type devices while matching their density, economic and energy constraints. The Long Term Evolution (LTE) standard developed by the 3rd Generation Partnership Project (3GPP) is a widely deployed cellular technology aiming to fulfill the demands of the connected society. Consequently, the unique service requirements of the M2M communications drive significant system enhancements led by the 3GPP groups as well as academic and industrial research initiatives. After a close review of the recent developments in this area, this thesis addresses key issues underlying the performance limitations of the LTE standards for massive M2M communications and proposes a pragmatic radio resource management scheme by developing key algorithms to support large-scale M2M communications with diverse traffic characteristics, Quality of Service (QoS) requirements and power constraints while warranting efficient utilization of the LTE system resources for human and machine-based users’ co-existence. First of all, this thesis presents an in-depth analysis of the uplink delay components of an LTE system and identifies the hidden delays which could hinder the timely delivery of ultra delay sensitive M2M packets waiting in the device buffers. To communicate the true age of the critical M2M data to the eNodeB scheduler, a novel Packet Age (PA) MAC control element is proposed and a delay-sensitive uplink scheduler is designed leveraging on the PA information to maximize the probability of satisfying the delay budgets of event-based, bursty M2M traffic. Besides, a new performance metric is proposed to measure the efficiency of radio resource utilization for different scheduling policies. Most of the other research endeavors focused on alleviating the random access channel congestion caused by simultaneous access attempts from thousands of M2M devices. However, this thesis emphasizes the issue of downlink control channel overload as a major system bottleneck to realize the full potential of LTE data capacities for the distinctive M2M traffic profiles. The challenge of serving the concurrent uplink scheduling requests from massive number of connected M2M devices within prescribed QoS boundaries and minimal control overhead is resolved in this thesis. To cater for the non-deterministic arrival patterns and variable burst sizes associated with potential M2M applications, this thesis proposes an adaptive Semi-persistent Scheduling (SPS) scheme which adjusts the allocated radio resources to the instantaneous queue sizes of the M2M devices/gateways yet is free from the requirement of any explicit control signaling except for the initial grant. This ingenious algorithm combines the benefits of both LTE dynamic scheduling and static semi-persistent scheduling by facilitating flexible resource allocations with nominal control channel consumption. A step-by-step selection mechanism is developed for the appropriate scheduling policy (dynamic or adaptive SPS) according to the M2M traffic patterns and specific QoS requirements. To compensate for the performance degradation due to channel fading around the M2M terminals, an enhancement of the adaptive SPS algorithm is also presented to adapt to variable channel conditions but without expansive signaling. Another novelty of this thesis is building a new system model to support multi-service M2M traffic classes with wide range of QoS requirements within the LTE system framework while ensuring fair resource sharing among human-centric and M2M users. A control mechanism is designed to offer variable service capacities and levels of QoS satisfactions to different classes of M2M traffic based on their subscription profiles. Instead of rigid allocation priorities to certain classes, a pre-emptive feature is incorporated to prioritize comparatively delay sensitive traffic that join the network later by freeing up resources allocated to delay tolerant traffic and update the requests with appropriate scheduling parameters. Moreover, to realize the low-power M2M architecture, the standard Discontinuous Reception (DRX) mechanism is enhanced by synchronizing with the adaptive SPS transmissions to ensure uninterrupted and extended device sleep cycles, thereby saving battery power of M2M terminals. The proposed algorithms in this thesis have been implemented using the Riverbed Modeler simulation software which facilitated the design and performance analysis of the novel algorithms in a realistic LTE network environment. Extensive simulations and statistical analysis of the presented algorithms demonstrated significant improvement over the conventional LTE schedulers in terms of resource efficiency and QoS satisfaction of M2M users with diverse delay budgets, channel conditions and low power requirements. As a whole, this thesis makes significant contributions to solve the key challenges related to the radio resource management policies of the LTE networks to support mass M2M communications in an efficient way.
- Subject
- LTE; M2M; scheduling; radio resource management; QoS; IoT
- Identifier
- http://hdl.handle.net/1959.13/1513697
- Identifier
- uon:56759
- Rights
- Copyright 2022 Nusrat Afrin
- Language
- eng
- Hits: 21
- Visitors: 21
- Downloads: 0