Network Slicing Based 5G and Future Mobile Networks: Mobility, Resource Management, and Challenges

Haijun Zhang, Haijun Zhang 0001, Na Liu, Xiaoli Chu, Keping Long, Abdol-Hamid Aghvami, Victor C. M. LeungIEEE Communications Magazine 2017 · DOI

TL;DR

This paper proposes a logical architecture for network slicing in 5G systems, focusing on efficient mobility management and resource allocation. It introduces a novel handover scheme and a joint power and subchannel allocation strategy for spectrum-sharing two-tier networks, demonstrating improved capacity for different service slices.

What Problem Does This Paper Address?

The advent of 5G mobile communication systems promises to meet diverse user Quality-of-Service (QoS) requirements across various application scenarios, from high-data-rate enhanced mobile broadband (eMBB) to ultra-reliable low-latency communication (uRLLC) for critical applications like autonomous driving, and massive connectivity for the Internet of Things (IoT). To achieve this, network slicing has emerged as a key paradigm, allowing a single physical network infrastructure to be partitioned into multiple virtual, isolated logical networks, each tailored to specific service demands. However, realizing the full potential of network slicing presents significant challenges, particularly in managing mobility across different network slices and efficiently allocating resources to meet their distinct QoS requirements, especially in complex, heterogeneous network environments with interference.

Methods

The paper introduces a logical architecture for network slicing based on Software-Defined Networking (SDN) and Network Function Virtualization (NFV). For mobility management, it proposes a handover management scheme designed for handovers between different access networks within a network slicing context. This scheme leverages SDN controllers to manage handover decisions, aiming for seamless user experience. For resource allocation, the authors formulate an uplink resource allocation problem for small cells in a two-tier spectrum-sharing network (small cells and macrocells). This problem is framed as maximizing uplink capacity on each subchannel, subject to constraints on transmit power, minimum data rate for uRLLC users, interference thresholds for the macrocell, and single-user allocation per subchannel per cell. The solution employs a Lagrangian dual decomposition method, relaxing binary allocation indicators to continuous variables and then solving a master problem with sub-problems. The Karush-Kuhn-Tucker (KKT) conditions are used for power allocation, and a sub-gradient method is used to update Lagrangian multipliers for subchannel allocation.

Datasets

Not discussed in the paper. The paper relies on simulation results to demonstrate the performance of the proposed schemes.

Key Insights and Ideas

  • Network Slicing as a Foundation for 5G Diversity: Network slicing is crucial for enabling 5G to cater to a wide spectrum of services with vastly different QoS needs, moving beyond a one-size-fits-all approach.
  • Mobility Management in Sliced Networks is Complex: The diverse requirements of different network slices (e.g., high mobility for railway communications vs. low mobility for IoT) necessitate new, flexible handover mechanisms that can adapt to slice-specific needs.
  • Interference is a Major Hurdle in Heterogeneous Networks: In spectrum-sharing environments with small cells and macrocells, both co-tier (within small cells) and cross-tier (between small cells and macrocells) interference significantly impact capacity, especially for slices like IoT that rely on macrocell coverage.
  • Resource Allocation Must Be Slice-Aware: Efficiently allocating resources requires considering the specific QoS demands and interference constraints of each slice, leading to tailored strategies rather than generic ones.
  • SDN/NFV Enable Dynamic and Flexible Network Slicing: The integration of SDN and NFV provides the necessary programmability and virtualization to dynamically create, manage, and allocate resources to network slices.

Key Results

The paper presents simulation results demonstrating the effectiveness of the proposed resource allocation scheme.

  • eMBB Slice Capacity: The total capacity of the eMBB slice increases nearly linearly with the density of small cells. However, this capacity is significantly reduced by inter-cell interference, particularly at high small cell densities.
  • uRLLC Slice Capacity: The total capacity of the uRLLC slice also increases with the number of small cells. Notably, its capacity is approximately 20 times less than that of the eMBB slice, attributed to eMBB's need for large bandwidth for data-intensive services, while uRLLC focuses on low-volume, low-latency control messages.
  • IoT Slice Capacity: The total capacity of the IoT slice, supported by the macrocell, decreases with an increasing number of small cells. This is due to cross-tier interference from small cells serving eMBB and uRLLC slices, and further degradation from co-tier interference.
  • Overall Efficiency: The proposed resource allocation scheme is shown to effectively allocate network resources for both latency-sensitive (uRLLC) and latency-tolerant (eMBB, IoT) scenarios, improving system capacity in dense heterogeneous networks.

Limitations and Future Work

The authors acknowledge that due to space limitations, they will discuss other metrics, such as latency, in future works. The paper focuses on the proposed resource allocation scheme and its impact on capacity. Further research could delve deeper into the performance of the proposed mobility management scheme under various realistic scenarios and explore other performance metrics like latency, energy efficiency, and security for different slices. The paper also highlights the need for further research into network reconstruction, slicing management, and cooperation with other 5G technologies as ongoing challenges.

Summary updated

BibTeX
@article{zhang2017network,
  author = {Haijun Zhang and Haijun Zhang and Na Liu and Xiaoli Chu and Keping Long and Abdol-Hamid Aghvami and Victor C. M. Leung},
  title = {{Network Slicing Based 5G and Future Mobile Networks: Mobility, Resource Management, and Challenges}},
  journal = {IEEE Communications Magazine},
  year = {2017},
  doi = {10.1109/mcom.2017.1600940},
}
Figure