Nokia rolls out AVA AI at the edge for 5G CX
Nokia has introduced an AI enhancement that sits at the network edge and promises to enhance the 5G customer experience around gaming and streaming.
AVA Quality of Experience (QoE) at the Edge enables AI to be deployed closer to the customer, and in turn facilitating real-time CX improvements. On the technicalities, Nokia said AVA QoE at the Edge brings “code to where the data is”, by deploying Machine Learning (ML) algorithms at the network edge to enable real-time automated actions.
The solution also eases the data burden on CSPs, with “an exponential reduction” in the volume of user plane data required to feed ML models.
The service will be available for multi-vendor Open RAN (O-RAN) networks, ensuring the different components work in unison with subscribers benefitting from ultra-low latency, reduced jitter and buffering across such services as YouTube, Netflix and cloud gaming.
Dennis Lorenzin, head of network cognitive services, global services, Nokia, said: “Today, many CSPs are keen to launch new low latency services to their customers. With Nokia’s AVA QoE at the Edge, we bring AI to the edge, so CSPs can deliver personalized 5G experiences and guaranteed performance.”
As many as 35% of telecom operators said their top 5G strategy objective is to improve customer experience. Nokia said that deployment of its AVA algorithms on traditional network architectures had achieved a 59% reduction in Netflix buffering and 15% fewer YouTube sessions that suffer from long playback.
Stefan Pongratz, VP of Dell’Oro Group, added: “The increased complexity with the various 5G technologies in combination with the shift towards Open RAN will potentially introduce new challenges to CSP operational teams tasked with managing end-to-end performance.
“AI will play an increasingly important role managing this complexity and deliver the Quality of Experience (QoE) that consumers and enterprises demand from mobile broadband applications and latency-sensitive services. Nokia’s approach combines centralised AI to generate network-wide insights and pre-trained models with distributed AI for real-time optimisation of the RAN.”