- Queen Mary University of London researchers have developed an AI-assisted communications system designed to improve data transmission when bandwidth is limited.
- Tests showed the approach can reduce communication errors by more than 40 per cent under severely bandwidth-constrained conditions.
- Researchers say the technology could eventually support satellite links, aircraft, spacecraft and next-generation wireless networks.
Researchers at Queen Mary University of London have developed an artificial intelligence-assisted communications system that could make future wireless and optical networks more reliable when bandwidth is limited.
The system uses machine learning to help optical receivers recover information that might otherwise become distorted or lost when communication channels are congested or bandwidth is restricted.
The research, led by Dr Paul Anthony Haigh, takes a different approach from AI systems designed to replace conventional communications technology.
Instead, the researchers have integrated machine learning into an existing receiver design, allowing AI to make targeted adjustments when necessary while retaining established communications engineering as the foundation of the system.
The researchers say the approach could help address growing pressure on communications networks as demand for data increases across satellite internet, airborne networks and next-generation wireless systems.
Errors reduced under difficult conditions
Experimental testing showed that the AI-assisted system significantly outperformed conventional receivers under challenging operating conditions.
According to the researchers, communication errors were reduced by more than 40 per cent in severely bandwidth-limited scenarios. Under some operating conditions, estimated bit error rates were reduced twofold.
The system is also designed to determine when AI intervention is unnecessary.
When communication conditions are favourable, it automatically returns to the conventional receiver instead of continuing to use machine learning.
Researchers say this could avoid unnecessary computing and energy consumption compared with AI-based communication systems that continuously rely on computationally demanding models.
The development could prove particularly important as engineers seek to transmit increasing amounts of information through limited bandwidth without substantially increasing the complexity and cost of communications infrastructure.
Haigh said many current AI approaches seek to replace conventional communication systems with computationally demanding “black box” models that can be difficult to interpret.
“Instead of asking AI to do everything, it lets decades of communications engineering do most of the work, while AI makes small corrections only when they’re needed,” Haigh said.
“It’s a partnership between human engineering knowledge and machine learning rather than a replacement for either.”
The approach demonstrates how artificial intelligence could complement established communications theory instead of replacing proven engineering techniques.
Potential for satellites and spacecraft
Although the technology remains at the research stage, the researchers say it could eventually have applications across several communications environments.
These include satellite communications, laser-based links between aircraft, drones and spacecraft, data-centre networks and future wireless systems where available bandwidth is restricted or conditions can change rapidly.
Such applications could become increasingly important as satellite internet services, airborne communication networks and other data-intensive technologies expand.
READ ALSO: Be careful who you give your loyalty to at the workplace
The researchers also see potential sustainability benefits, particularly if the technology can improve network efficiency while avoiding unnecessary AI processing when conventional receivers are already performing adequately.
The study has been published in npj Wireless Technology.
By Joseph Mambili
You can also follow our social media pages on Twitter: Education News KE and Facebook: Education News Newspaper for timely updates.
>>> Click here to stay up-to-date with trending regional stories
>>> Click here to read more informed opinions on the country’s education landscape
