Personal tools

Pervasive Network Intelligence

RWTH Aachen University_Martin Braun_020722A
[RWTH Aachen University, Germany - Martin Braun]

- Pervasive Intelligence: Smart Machines Everywhere

Advances in artificial intelligence (AI) are giving rise to a multitude of smart devices that can recognize and react to sights, sounds, and other patterns—and do not need a persistent connection to the cloud. These devices could well unlock greater efficiency and effectiveness at organizations that adopt them. In some industries, they may even change how profits are apportioned.


- From Connected to Pervasive

The age of pervasive intelligence will be marked by a proliferation of AI-powered smart devices and machines that will learn from experiences, adapt to changing situations, and predict outcomes. Some will infer user needs and even collaborate with other devices. Moreover, with AI embedded rather than confined to the cloud, these smart devices will not depend on internet connectivity, and will not suffer from latency entailed in transmitting data to the cloud for analysis. This will enable applications that require instantaneous response and robust performance even when connectivity is poor or not available.


- Pervasive Network Intelligence for 6G Networks 

The sixth generation (6G) network is expected to support increasingly heterogeneous networking paradigms, adapt to dynamic network environments, and provide diverse intelligent services with strict quality of service (QoS) requirements. To this end, artificial intelligence (AI) will permeate and integrate into every aspect of the network, including the end user, network edge, and cloud, enabling ubiquitous network intelligence. 

Pervasive cyber intelligence can be enabled from two angles: AI for networks and networks for AI. The former is to leverage and customize AI-based approaches for complex 6G network management, while the latter is to design and optimize 6G networks to facilitate service-oriented AI applications (i.e., AI services). 

However, achieving ubiquitous network intelligence faces different challenges. It should support various AI services with different QoS requirements in terms of latency, reliability, accuracy, etc. 

Furthermore, service demand exhibits spatial and temporal dynamics due to traffic burstiness and user mobility. Improving the utilization of heterogeneous sensing, communication, computing, storage, and control resources is crucial to determine fine-grained user-centric network solutions.


[More to come ...]

Document Actions