Network and Computing-Aware Edge Datacenter Placement and Content Placement in Edge Compute First Networking

Edge caching has been proposed to reduce core network bandwidth consumption and content transmission delay, and it is being envisioned as one of the key components in edge compute first networking (CFN). Edge datacenter (DC) placement and content placement are coupling issues in practice, where the former affects the latter. In this paper, a joint edge DC placement and content placement algorithm is proposed in CFN to minimize delay and bandwidth consumption under the constraints of the relationship among edge DCs and base stations, the number of edge DCs, the storage capacity of each edge DC, bandwidth resources, and the popularity of content. The proposed algorithm combines network and computing resources, which includes an edge DC placement algorithm based on Kuhn-Munkres and content placement algorithm based on K-Means. The effectiveness of the proposed algorithm is verified by comparing with TOP-I- MPC, KMDP-RC, and RP- KMCC.