Wireless Coexistence of Cellular LBT Systems and BLE 5

Abstract

The 2.4 GHz spectrum is home to several Radio Access Technologies (RATs), including ZigBee, Bluetooth Low Energy (BLE), and Wi-Fi. Accordingly, the technologies' spectrum-sharing qualities have been extensively studied in literature. License-Assisted Access (LAA) Listen-Before-Talk (LBT) has been identified in technical reports as the foundation for the channel access mechanism for 5G New Radio-Unlicensed (NR-U) operating in the 2.4 GHz Industrial, Scientific, and Medical (ISM) band. The introduction of NR-U into this band raises new concerns regarding coexistence of the newcomer with traditional incumbents. This article reports an investigation of BLE 5 and cellular LBT coexisting systems by means of empirical evaluation. The importance of this study stems from that the studied LBT mechanism is indicative of how 5G NR-U would perform in the 2.4 GHz band. Tests were performed in conformity with the American National Standards Institute (ANSI) C63.27 standard for evaluation of wireless coexistence, and results were reported in terms of throughput and interframe delays. In accordance with the standard and under different BLE physical layers (PHYs) and LBT priority classes, three setups were investigated. These pertain to the three tiers of evaluation, which correspond to the criticality of the device under test. Results demonstrated how BLE throughput dropped as the intended-to-unintended signal ratio decreased, and LBT classes exhibited a diminishing effect as the class priority descended. Long Range BLE PHY was found to sustain longer gap times (i.e., delay) than the other two PHYs; however, it showed less susceptibility to interference. Results also demonstrated that low data rate BLE PHYs hindered the LBT throughput performance since they correspond to longer airtime durations.

Keywords: 5G NR-U; BLE 5; LBT; coexistence; empirical evaluation; wireless medical devices.

FIGURE 1.

A high-level flowchart demonstrates the LBT procedure for Frame Based Equipment, as stipulated in ETSI standard.

FIGURE 2.

Link layer packet format for BLE uncoded 1M and 2M PHYs.

FIGURE 3.

Link layer packet format for BLE Coded PHY (LR).

FIGURE 4.

Channel hopping pattern of BLE channel selection algorithms #1 (top) and #2 (bottom) over 100 connection events.

FIGURE 5.

Block diagram of channel selection algorithm #2 introduced in BLE version 5.

FIGURE 6.

The experimental setup of the coexistence test illustrating the arrangement of BLE nodes and the three LBT pairs with center frequencies 2412 MHz, 2437 MHz, and 2462 Mhz.

FIGURE 7.

Normalized BLE throughput under LBT interferers of class 1, 2, 3, and 4.

FIGURE 8.

Mean IFS durations of BLE PHYs in three tiers as a function of the I/U ratio.

FIGURE 9.

An example of the number of packets that can be sent during one connection event for 2M, 1M, and LR PHYs.

FIGURE 10.

Normalized LBT throughput in the evaluation of three tiers.

FIGURE 11.

Box plot of packet durations for BLE physical layers from tier 1 scenario.

FIGURE 12.

BLE channel histogram as a function of I/U ratio.