​AI-based joint source and channel coding

A complementary proof-of-concept, developed through a collaboration between Qualcomm Technologies and Nokia Bell Labs, applies AI-based joint source‑channel coding to HARQ feedback. By learning real-world network conditions and adapting feedback signaling in real time, the system reduces errors and unnecessary retransmissions. This improves reliability and efficiency today while laying the groundwork for intelligent, AI-native air interfaces in future 6G networks.

​AI-based joint source and channel coding

AI‑enabled context‑aware communications for 6G — where on‑device AI helps connectivity adapt to what matters in the moment. What happens when wireless stops treating every app the same and starts responding to intent, context, and changing conditions? This demo points to the AI‑native direction of 6G, exploring how future devices and networks could collaborate to deliver smoother, more responsive experiences for applications such as video calling and cloud gaming — without relying solely on traditional QoS. Instead of rigid, one‑size‑fits‑all behavior, AI‑native protocols open the door to more adaptive performance, tuned to the experience you’re actually trying to have. The result is connectivity that feels less like a utility and more like an intelligent system working alongside you.

Agentic AI and AR experiences​

Agentic AI and augmented reality (AR) are enabling new experiences that depend on continuous perception, memory, and recall. Supporting these capabilities at scale requires rethinking how devices compute and connect. We demonstrate an AI recall use case enabled by distributed compute — a key 6G enabler — that allows AI workloads, such as creating and retrieving visual memories, to dynamically shift between devices and the network edge based on link conditions and power constraints.  

​Network compute and inference services at scale​

Network compute and inferencing for 6G — where intelligence can move beyond the device to meet the growing demands of AI‑driven systems. As models scale in capability, future connectivity must do more than transport data; it must help intelligence flow to where it is needed. This demo provides insight into how 6G could enable devices to seamlessly access greater AI capabilities while balancing energy, latency, and reliability. By allowing intelligence to be distributed across device and edge network, 6G opens new possibilities for physical AI and robotics operating in wide-area dynamic, real‑world environments. The result is a foundation where 6G connectivity actively supports scalable intelligence, creating new monetization opportunities for operators while preserving privacy and security.