Press Releases

Technology Overview

In order for both the base station and the radio terminal to use narrow beams for multiuser transmission, it is necessary to determine who should direct the beam to each other, i.e., the combination of the distributed antenna and each terminal antenna. In order to determine this combination, we have devised a method that observes the reception level when the best analog beams are directed to each entity for each combination, and then selects the distributed antennas with the highest reception level from the combinations available, such that they do not overlap.

Since a narrow beam can suppress interference over a wide direction, this method, which enables the application of a narrow beam on both the distributed antenna side and the terminal antenna side, maintains the interference suppression effect even when the radio terminal moves, and is expected to stably hold the high capacity of multi-user transmission (Figure 1).

Experimental Contents and Results

A demonstration experiment of this technology was conducted in a laboratory with 4 pillars in an area of 29 m × 15 m (Figure 2-1). As for the frequency band used in the experiment, the 40 GHz band was used, which is higher than the 28 GHz band of 5G millimeter wave service in Japan*10. Other physical specifications are based on 5G NR (New Radio), and OFDM*11 with a 100 MHz bandwidth and 60 kHz subcarrier spacing are used. A coaxial cable (20 m long) connected multiple distributed antennas to the base station. We simulated the environment where a maximum of 14 distributed antennas were installed in positions #1 to #14, and the radio terminal was equipped with 4 antennas to cover the 4 directions of front, back, left, and right. The terminal was mounted on a trolley and moved along the route shown in the figure.

We experimentally evaluated the channel capacity characteristics of the downlink when Technology 1 was applied using narrow beams for both the distributed antenna and the terminal antenna and when pre-coding was applied using a wide analog beam (hereinafter referred to as "wide beam") at both sides (Figure 2-2). Specifically, a channel estimate between each distributed antenna and terminal antenna is acquired by using reference signals, and is used for calculation of the channel capacity. As reference data, we also evaluated the channel capacity characteristics when pre-coding was applied to this technology (narrow beam) and when there was only one radio terminal. The pre-coding update period was set to 40 ms, and the horizontal beam width of the narrow beam was about 16 degrees with the horizontal beam direction of -67 to +67 degrees. On the other hand, the width of the wide beam was set to about 60 degrees for the horizontal beam width, and the beam direction was fixed to the horizontal front direction. In order to fairly compare the effects of interference between radio terminals using narrow and wide beams, the beam gain of both beams was set to be equal. The horizontal axis of the graph shows the movement speed of the radio terminal, and the vertical axis shows the channel capacity per radio terminal (hereinafter referred to as "channel capacity"). The channel capacity was evaluated using the value of 50% cumulative distribution in the experiment area. The number of radio terminals was set to 4, and the number of simultaneous transmission streams per radio terminal was set to 2. The radio terminal selected 2 antennas with high reception level from the 4 antennas, and minimum mean square error (MMSE) reception processing*12 was performed using these 2 antennas.

According to the experimental results, when applying a wide beam and pre-coding, the interference suppression effect of pre-coding is local, and the channel capacity deteriorates by 90% at a moving speed of 3.3 km/h or more compared with that at rest. On the other hand, in the case of applying a narrow beam, the channel capacity at rest is maintained even at a moving speed of 3.3 km/h or more because the narrow beam maintains interference suppression over a wide range of direction. In addition, the degradation of the channel capacity is reduced to 27% compared to when only one radio terminal is active. Furthermore, when pre-coding is applied to a narrow beam, the channel capacity deteriorates compared to this technology without pre-coding at a moving speed of 3.3 km/h or higher. However, at rest, the channel capacity is improved compared to this technology, and the channel capacity is almost the same as when only one radio terminal is active. In the future, if the application/non-application of pre-coding can be controlled based on the movement of the radio terminal, we have confirmed the possibility that the application of narrow beams can achieve stable channel capacity regardless of the movement of the radio terminal.

Technology Overview

We continuously measure the radio communication quality of each distributed antenna and beam at each position in the area, and learn the optimal distributed antenna and beam combination at each position. During operation, we observe the radio communication quality of each distributed antenna and beam at any time, and estimate the position of the radio terminal by machine learning. Furthermore, we predict the subsequent position of the radio terminal from the estimated movement track of the radio terminal in the past, and predict the optimum distributed antenna and beam before the next radio communication quality information is obtained. As a result, even if there is a possibility that the transmission performance of the distributed antenna and beam selected based on the current radio communication quality information is drastically reduced or disconnection occurs due to the blocking triggered by the movement, this technology enables us to continue radio transmission by selecting the predetermined appropriate distributed antennas and beams based on the predicted subsequent position of the moving radio terminal.

Experimental Contents and Results

In an experimental environment similar to multiuser transmission technology utilizing narrow beams, we conducted a demonstration experiment to verify the effectiveness of this technology under an environment simulating a total of 14 distributed antennas. Specifically, we evaluated the relative received signal strength (relative to a predetermined standard in the device) of downlink transmission using Technology 2 (Figure 3) by using the channel estimate between each distributed antenna and terminal antenna. For example, if the acquisition interval of radio communication quality information is 40 ms, and the radio terminal moves at a jogging speed (6.6 km/h) along the route close to the pillars shown in Figure 2-1, the received signal strength of the conventional technology, which selects the distributed antenna and beam based on the current radio communication quality, decreased by about 13dB at the position blocked by the pillar. On the other hand, in Technology 2, which selects distributed antennas and beams based on terminal mobility prediction, the received signal strength at the same position is improved by about 9dB compared with the conventional technology, and the degradation in the received signal strength is restrained to about 4dB, confirming that it is possible to avoid disconnection, which is a concern in high-frequency band service. By using Technology 2 together with Technology 1, we can expect to realize stable, high-speed, high-capacity communication in the high-frequency band even when moving next to blockage.