Researchers Split mmWave Signals Into Multiple Beams to Offer the Best 5G Connectivity

Engineers at the University of California San Diego have announced a successful project to boost the throughput and reliability of mmWave radio signals — including those used in 5G cellular networks.

Radio networks are all about balance: It's possible to offer extremely high throughput at the cost of coverage, or wide coverage at the cost of throughput. The UC San Diego team, however, claims it can offer both — by allowing mmWave signals to travel further.

"Relying on a single beam creates a single point of failure," Dinesh Bharadia, professor of computer engineering and senior author of the paper, explains of the issue his team sought to solve: The tendency for traditional single-beam mmWave signals to hit an obstacle and fail to reach the receiving device.

The solution: Splitting the signal up into multiple beams instead. "You would think that splitting the beam would reduce the throughput or quality of the signal," Bharadia claims. "But with the way that we've designed our algorithms, it turns out mathematically that our multi-beam system gives you a higher throughput while transmitting the same amount of power overall as a single-beam system."

In testing, both in an office environment and out-of-doors, the team's multi-beam implementation was able to offer 800Mbps throughput with a 100 per cent reliability rate— reaching up to 262 feet in the outdoor test.

The team's system relies on two algorithms, which can be implemented on existing hardware: One instructs the base station to split the beam into multiple paths, including indirect routes to the receiver, and figures out which paths work in order to combine all beams into a strong signal at the target receiver; the second algorithm tracks the movement of the target receiver, compensating for moving obstacles and repositioning.

"You don’t need any new hardware to do this," says Ish Jain, the paper's first author. "Our algorithms are all compliant with current 5G protocols."

The only catch: At present, the system supports just a single user. The team has confirmed that its next step will be scaling it to support multiple simultaneous users, to better simulate real-world conditions of a cellular network.

The team's work was presented at the ACM Special Interest Group on Data Communication conference (ACM SIGCOMM 21) this week, with more information and a link to the paper available on the project website.