Multi-User MAC Protocol for WLANs in MmWave Massive MIMO Systems With Mobile Edge Computing

Yifeng Zhao, Xueting Xu, Yuhan Su, Lianfen Huang, Xiaojiang Du, Nadra GuizaniIEEE Access 2019 · DOI

Introduction

The rapid evolution of wireless applications, such as 4K video streaming, virtual reality (VR), and remote-controlled robotics, has placed unprecedented demands on Wireless Local Area Networks (WLANs). These "heavy" applications require both high computational power and high-speed, low-latency data transmission. Mobile Edge Computing (MEC) has emerged as a solution to the computational bottleneck by offloading tasks from resource-constrained mobile devices to the network edge. However, the communication link between the user and the MEC server remains a critical bottleneck.

Millimeter-wave (mmWave) frequencies and Massive Multiple-Input Multiple-Output (MIMO) technologies offer the bandwidth and spatial degrees of freedom necessary to support these demands. Yet, traditional Media Access Control (MAC) protocols, such as IEEE 802.11ad, are not optimized to fully exploit the spatial multiplexing capabilities of Massive MIMO. This paper introduces two novel MAC protocols designed to bridge this gap by treating spatial resources as a primary dimension for scheduling.

System Model and Architecture

The proposed system considers a WLAN environment where a Massive MIMO Access Point (AP) is equipped with a large antenna array and integrated with an MEC server. This setup allows the AP to serve multiple users simultaneously using narrow mmWave beams, effectively implementing Space Division Multiple Access (SDMA) alongside traditional Frequency and Time Division Multiple Access (FDMA/TDMA).

System architecture of the mmWave Massive MIMO WLAN with MEC integration.

As illustrated in Figure 1, the AP uses beamforming to create dedicated spatial channels for different users. The MEC server, co-located or directly connected to the AP, handles the offloaded computational tasks, significantly reducing the latency perceived by the end-user.

Proposed MAC Protocols: LSMWN-MAC and MLSMWN-MAC

To manage the complex task of multi-user resource allocation, the researchers developed two protocols:

1. LSMWN-MAC (Large-scale MIMO Based Wireless Network MAC)

LSMWN-MAC adopts a hybrid resource allocation strategy. It divides the available frequency spectrum and time into discrete "resource blocks," similar to the structure used in LTE. In this protocol, the AP estimates the Channel State Information (CSI) based on "Clear to Send" (CTS) frames received from users. This CSI is then used for beamforming and scheduling. The allocation is relatively static, providing a stable framework for multi-user communication but with limited flexibility in rapidly changing environments.

2. MLSMWN-MAC (Modified LSMWN-MAC)

MLSMWN-MAC is a dynamic evolution of the base protocol. In this variant, user nodes take a more active role by estimating the channel state across all available sub-channels and feeding this information back to the AP. This global view allows the AP to perform real-time optimization, dynamically grouping users and assigning them to the most favorable spatial and frequency slots. While this increases control overhead, it significantly improves the overall network throughput by adapting to instantaneous channel conditions.

Both protocols utilize Matched Filtering (MF) for precoding. In Massive MIMO systems, MF is highly effective due to "channel hardening," a phenomenon where the effects of small-scale fading are averaged out, and inter-user interference naturally diminishes as the number of antennas increases.

Performance Evaluation

The protocols were evaluated using numerical simulations based on the 3GPP Urban Micro (UMi) path-loss model and Rayleigh fading channels. The simulations compared the proposed protocols against the IEEE 802.11ad standard under various conditions.

Throughput Gains

The results demonstrate significant improvements in saturated throughput. LSMWN-MAC achieved an increase of approximately 4.7 Gb/s over 802.11ad, while the dynamic MLSMWN-MAC reached a gain of 5.4 Gb/s. These results indicate that the proposed MAC layers can effectively unlock the multi-gigabit potential of mmWave hardware.

Scalability and User Density

One of the most critical findings relates to how the protocols handle high user density. Traditional WLAN protocols often suffer from a "collision collapse" as the number of users increases. In contrast, the proposed protocols maintain high performance even in "hotspot" scenarios with 30 to 50 active users.

Throughput performance as a function of the number of users, comparing the proposed protocols with standard benchmarks.

As shown in Figure 5, MLSMWN-MAC generally outperforms LSMWN-MAC due to its dynamic optimization. However, an interesting "threshold" effect was observed: at extremely high user counts, the overhead associated with collecting CSI for the dynamic protocol begins to outweigh its optimization benefits, making the simpler LSMWN-MAC more efficient in those specific regimes.

Conclusion

This research highlights that the transition to 5G-grade WLANs requires more than just better physical layer hardware; it necessitates a fundamental redesign of the MAC layer to incorporate spatial multiplexing as a core resource. By integrating MEC with mmWave Massive MIMO through the LSMWN-MAC and MLSMWN-MAC protocols, the authors have demonstrated a path toward supporting the next generation of bandwidth-intensive, low-latency applications. Future work will focus on optimizing uplink interactions and reducing the feedback overhead to further enhance energy efficiency and real-world applicability.

Summary updated

BibTeX
@article{zhao2019multi,
  author = {Yifeng Zhao and Xueting Xu and Yuhan Su and Lianfen Huang and Xiaojiang Du and Nadra Guizani},
  title = {{Multi-User MAC Protocol for WLANs in MmWave Massive MIMO Systems With Mobile Edge Computing}},
  journal = {IEEE Access},
  year = {2019},
  doi = {10.1109/access.2019.2952174},
}
Figure