Samsung tackles 4G/5G antenna optimization in 15 minutes with drone, AI

Samsung is touting speed, improved safety for tower climbers, and more efficient network maintenance as main benefits of the product. (Samsung)

Samsung is taking the task of 4G/5G antenna configuration off the tower and into a smartphone with a new drone-based measurement tool that leans on AI to collect visual data and deliver it to engineers on the ground for faster analysis.

Samsung just completed a successful demo of the product at its campus this week, which involved a camera-equipped drone transmitting antenna angle data to a mobile device. A global launch is planned for later this year. The tool works on both 4G and 5G smartphones and tablets, and can analyze 4G/5G antennas and radios made by other vendors.

According to Samsung, out of all the antennas in deployment, about 10-15% are out of position by up to 10 degrees, usually due to reliance on the human eye or because of environmental conditions like natural disasters.

Sponsored by VoltDB

Webinar: The Hidden Inflection Point in 5G: When the Changing Definition of Real-Time Breaks Your Existing Tech Stack

Rethink your definition of real-time to match the changing reality brought to the forefront by 5G. Your users expect milliseconds, in-event decision making. Is your tech stack ready?

While performance can be affected by other external factors, including spectrum, deployment location, type of radio, even slight deviations from antennas’ predefined optimal angles can impact user experience. For example, Samsung told Fierce via email that when antenna tilt is off by just one degree from its intended configuration in a dense urban location, it can create a gap between an operator’s expectation and the actual coverage delivered.

Samsung is touting speed, improved safety for tower climbers, and more efficient network maintenance and cell site management for optimized performance, as the product’s main benefits.

With the tool, an engineer can use a smartphone to remotely fly a camera-equipped drone that takes pictures of installed antennas on rooftops or towers. Visual data is transmitted to a cloud server in less than a minute and then viewable on the mobile device on the ground.

AI is meant to verify the rotation and tilt of the antennas so that engineers can check that they have been installed correctly. A Samsung spokesperson didn’t provide specifics about the accuracy observed. 

“As more data is gathered through deep-learning, accuracy continues to improve,” the spokesperson said. “One way AI is employed is by automatically detecting antennas (whether it is the font, back, or side face of the antenna) based on the learnings of statistical features of the antennas.”  

RELATED: Carr: 5G Upgrade Order will make antenna swaps faster, more predictable

In the demo, Samsung finished the task within 15 minutes, starting from when the drone took flight to when the measurement results were delivered back to the device. That compares to antenna configuration measurement usually requiring a tower climb, with engineers carrying heavy equipment, which can take several hours from start to finish, according to the news release.  

Safety benefits could be particularly useful in the U.S., according to Samsung, for site audit and maintenance, which requires climbs that involve more advanced training and often mean operators have to dispatch two field technicians. The company plans to add more feature that will enable technicians to adjust antenna tilts to the optimal position remotely via a mobile device and PC. 

“As the number of 5G network sites grows, there has been a heightened focus on network performance by operators, and we are seeing an increased market demand for intelligent solutions for site maintenance,” said Sohyong Chong, vice president and head of Network Automation, Networks Business at Samsung Electronics, in a statement. “Once this solution launches globally later this year, it will offer a safer, more cost-effective and convenient way to satisfy market demands, leveraging our unique capabilities in combining the latest technologies—drones, AI and 5G.”

5G Americas President Chris Pearson, in a blog post Wednesday, wrote that despite some COVID-19 related obstacles, “5G is still on track to become the quickest generation of wireless cellular technology to be widely adopted.”

RELATED: FCC’s 5G Upgrade order passes 3-2

There are currently 82 commercial 5G networks, which is expected to grow to 206 by the end of the year, according TeleGeography.  

Citing data from Omdia, Pearson noted that in North America, there were 1.18 million 5G connections at the end of Q1, representing 100% growth over the prior quarter. By the end of 2020, Omdia expects North America will account for 10 million of the 238 million global 5G connections.

“In North America, we are seeing consistent, strong uptake of new 5G subscribers as new devices have been released that can take advantage of low-band and millimeter wave frequencies. At the same time, new network capabilities are being added,” Pearson stated.

In the U.S. the FCC this month adopted a 5G upgrade order meant to streamline wireless infrastructure deployments and allow mobile operators to upgrade existing sites with next-gen equipment. Earlier, FCC Commissioner Brendan Carr told Fierce Wireless the declaratory ruling would help technicians switch out 3G and 4G antennas for new 5G ones much more quickly and on a predictable scale.

Suggested Articles

Vodafone U.K.’s head of networks said if the government decides to strictly ban Huawei it would cost Vodafone “low-single-figure billions."

A new NTIA report determined 100 MHz in the 3.1-3.55 GHz band looks most viable for near-term sharing with commercial wireless services.

Instead of running on virtualized machines, Verizon is changing its underlying software architecture to run using cloud-native container-based tech.