Combined Latency-Aware and Resource-Effective Virtual Network Function Placement

Simple and Scalable Approach for Virtualized Network Function Placement in Wireless Multi-hop Networks

“…Resource Usage Delay Load Balancing Energy Deployment Cost Objective [13] Combines a heuristic with mathematical model Minimizes the maximum link and CPU core utilization and maximum delay of flows [14] Works based on centrality matrix Minimizes the number of activated nodes [15] Proposes 3 heuristic each place NFs based on one the objectives Minimizes the number of active nodes and balance the load between physical nodes [17] Works based on centrality Minimizes the overall cost of deployment [18] A combination of installation, migrations, and reassignments is applied to optimize the placement Minimizes the number of active nodes to reduce Energy and deployment cost [20] Places VNFs increasing BW close to destination and VNFs decreasing BW close to source Minimizes the maximum link load ratio on the flow path [21] Breaks the problem of placement into 3 parts Minimizes the resources usage [29] Dynamic Programming Minimizes the resources usage [30] Optimizes placement of each NF Minimizes the resources usage [31] Reduces the search space and then applies the mathematical model Minimizes the resources usage [32] Graph matching theory Minimizes the resources usage [33] Two phase algorithm that reuse already deployed VNF or a adds a new VNF Minimizes end-to-end delay [34] Stable Matching algorithm Minimizes delay between user and the deployed VNF [35] Breaks the problem of an SG placement into placing each NF and its link Minimizes cost, delay, Energy [36] Uses graph partitioning game Minimizes the nodal cost and balance the load between physical nodes.…”

“…Some services such as ultra-Reliable Low-Latency Communication (uRLLC) services require low latency access between the remote server and the clients while running in a highly reliable environment to guarantee service continuity. The authors of [33] propose a two stage heuristic. The first phase considers reusing VNFs already placed in the cloud since hosting a new one incurs additional cost.…”