Millimeter-Wave Feasibility in 5G Backhaul: A Cross-Layer Analysis of Blockage Impact

Abstract

The wireless backhaul has become a key enabler for 5G technology by presenting a costeffective and scalable alternative to the typical fiber backhaul. WiGig protocols, such as IEEE 802.11ad and later IEEE 802.11ay, have been considered for backhaul connectivity of 5G mobile networks, thanks to the availability of high bandwidths capable of achieving fiber-like data rates. However, this band suffers from high propagation loss that can only be compensated using highly directional antennas, making mmWave links more susceptible to blockage and errors. Thus, to effectively evaluate the viability of WiGig-based technologies in wireless backhaul scenarios, it is crucial to characterize the impact of obstruction across the different network layers. This article presents an extensive measurement campaign and cross-layer analysis of physical (PHY), medium access control (MAC), and transport layers metrics measured for outdoor WiGig-based hardware submitted to short-term and long-term blockage. This study found that maintaining constant and higher modulation and coding schemes (MCSs) in long-term blocked channels may induce packet errors as high as 100%, round-trip-time (RTTs) that can be in the order of a few seconds, and packet losses as high as 90%. Even dynamically adjusting the MCS, the performance can be highly degraded. This effect was exacerbated in short-term links, as they suffered from more extreme MCS changes upon sudden obstructions. Temporary line of sight (LOS) obstruction was shown to cause maximum delays of half a second and a PER of around 20%; in more extreme cases, it has even led to temporary link failures.

Experimental mmWave WiGig-based backhaul network dataset

“…

Fig. 1 Nodes and topology of the deployed outdoor testbed, presenting the obstacle used for the long-term blockage shortage [2] .

…”

“…After the data collection for each scenario was complete, the transport, network, and physical metrics were retrieved from the database using the recorded timestamps that marked the beginning of each experiment. The data was later analysed to study the consequences of long-term and short-term blockage on the multiple network layers [2] .…”

“…Note that this dataset includes a broader set of metrics (particularly from a lower perspective) and a wider range of blockage scenarios compared to other related works. Future use: The dataset provided allows defining the impact of both long-term and short-term obstructions on mmWave-based WiGig networks, as explored in [2] . It has also supported the development of a proactive machine learning framework [3] in an outdoor setting [ 4 ], leveraging real-life data to enhance the reliability and resilience of WiGig-based networks that are prone to blockages.…”

A Flexible mmWave Layer 2 Protocol Implementation for Integrated Access and Backhaul Architecture

“…The millimeter wave (mmWave) frequency band is enabling technology in 5G communication networks, being a topic of interest in recent years by the academia, standardization organizations, regulatory committees, and industry, as well as it is a key candidate for addressing the challenges of bandwidth shortage for 5G systems and beyond [11], [12], [13]. However, signals at mmWave bands undergo severe path loss and are highly sensitive to blockage as compared to frequency range 1 (FR1) of 5G [14], [15], [16], [17].…”