AI Helps 5G Broadcasts Reach Homes Without Fiber Wiring
New research shows how artificial intelligence can stabilize wireless TV signals for buildings that cannot afford expensive fiber optic upgrades, ensuring reliable access to emergency broadcasts.
Many older apartment buildings lack the internal wiring needed for modern cable television, leaving residents without access to national broadcasts or emergency alerts. While installing fiber optics is the standard solution, the cost often makes it impractical for landlords and tenants alike. This infrastructure gap creates a significant information deficit, particularly in urban areas where digital access is assumed to be universal.
A new approach leverages 5G multicast broadcasting to send television signals wirelessly to multiple devices at once. By using shared radio spectrum efficiently, this method offers a viable alternative to physical cabling. However, wireless signals are prone to interference, and unlike standard smartphone data, broadcast streams often cannot request retransmission if a packet is lost, leading to frozen video and missed critical information.
Solving the One-Way Signal Problem
Standard 5G connections rely on a two-way dialogue where devices ask for missing data. Multicast broadcasting lacks this return channel, meaning lost data is permanently gone. This limitation makes conventional speed-focused protocols unreliable for television delivery, as they are not designed to prioritize consistency over raw throughput in one-way transmissions.
Researchers from Waseda University in Japan developed a lightweight AI model to address this specific challenge. The system predicts changes in wireless conditions before they cause signal degradation and adjusts transmission settings in real time. This proactive approach ensures that the video stream remains stable even when the radio environment fluctuates rapidly.
Practical Performance on Commercial Networks
The model was trained on approximately 26 million data points collected from a commercial 5G network. It uses information already gathered by smartphones during normal operation, requiring no specialized hardware. In real-world tests, the AI-selected settings resulted in error-free video for 87% of segments, a significant improvement over the 32% success rate of conventional methods.
Performance metrics indicate the solution is ready for consumer devices. The model operates in less than 0.07 milliseconds on chipsets released since 2020, introducing no perceptible delay for viewers. This efficiency allows the technology to run on existing smartphones and tablets, making it a scalable option for widespread adoption without requiring new infrastructure.
Expanding Access Beyond Television Services
According to the report from GN technics/ai (en-US), this technology could bridge the information gap in underserved communities by providing equal access to broadcast services. The researchers note that the same principles apply to other one-way communication systems, including satellite links, autonomous vehicle networks, and industrial control systems where retransmission is impossible.
The findings suggest that local 5G infrastructure, currently used primarily by large organizations, can serve broader public needs. By enabling reliable one-way communication, this approach supports the convergence of broadcasting and broadband on a single platform. This shift could transform how public services are delivered, ensuring that critical information reaches everyone regardless of their building's age or wiring capabilities.
