Artificial Intelligence Assisted Softwarization, Virtualization, and Intelligentization of Everything for Improving Network Access

Future warfare will require commanding and controlling resources across multiple warfighting domains (i.e., sea, land, air, space, and cyber). Successfully performing command and control relies on reliable and accessible information aggregated from varied sensors. Connecting the sensors in contested environments needs a reliable communications network. This paper investigates the potential of AI-assisted softwarization, virtualization, and intelligentization of everything to improve network access in contested environments significantly. Using multiclass queueing theory, we optimize the allocation of resources to multiple services in a network-slicing environment. Our approach considers the limitations imposed by the dynamic spectrum allocation and the need for reliability and low latency in military communications. Some of the limitations imposed by the dynamic spectrum allocation in military communications include spectrum jamming and radio-triggered roadside bombs. We demonstrate that AI-assisted softwarization, virtualization and intelligentization of everything can significantly improve network access by increasing the carried traffic load at base stations. The carried traffic load is increased due to the enhanced number of physical links available at those base stations. The analysis results indicate that the carried traffic load or the number of physical links can be increased by 100,000. As our model and analysis are limited to the base stations and do not consider the network as a whole, future work will address these limitations by modelling the entire network and exploring the impact of jamming.