retken 6G Sebeke Mimarisi ve Servisi: Yerli Yapay Zek dan Yerli Sebeke retken Yapay Zek ya

Establishing endogenous intelligence capabilities within telecommunication networks in the form of native artificial intelligence (AI) has emerged as a significant topic of research for promoting the prospects of 6th Generation (6G) networks. In the upcoming 6G era, the network is expected to provide services for building and deploying distributed AI applications in the network infrastructure, and such applications will be used solely by the network or offered to operators or third parties as new services. This paper begins by analysing the current status of standardisation in the network AI domain, and next elaborates on the major shortcomings of the current approaches for natively setting up network AI services. On this basis, we present the concept of Native Network Generative AI (NNGAI), which builds on and further advances the concept of native AI for telecom networks. To formulate the paradigm of NNGAI, we first devise the network architecture, and then finally the underlying logical flow of network services to realize NNGAI by carefully inserting generative AI technologies into both the architecture and the service flow of 6G networks.