RouteNet: Leveraging Graph Neural Networks for Network Modeling and Optimizat...

杨利霞

RouteNet: Leveraging Graph Neural Networks for

- P; s, `# m, J0 Q) \' @8 C

Network Modeling and Optimization in SDN

- q7 P' d3 D2 e. q

8 I6 j; ?8 K* t) k& L; SNetwork modeling is a key enabler to achieve
! {/ i7 F) {* T3 Reffificient network operation in future self-driving Software
Defifined Networks. However, we still lack functional network
models able to produce accurate predictions of Key Performance
' H4 O" q- Z: l( oIndicators (KPI) such as delay, jitter or loss at limited cost.
In this paper we propose RouteNet, a novel network model based
" [% l( G8 Y/ [. M8 V2 q9 Ron Graph Neural Network (GNN) that is able to understand
7 e$ Q! P. L! _& Y- zthe complex relationship between topology, routing, and input
traffific to produce accurate estimates of the per-source/destination
per-packet delay distribution and loss. RouteNet leverages the
/ l! }1 [) f7 ^ability of GNNs to learn and model graph-structured information
% `1 e8 l* ~8 C, ?8 o3 O6 m7 @% band as a result, our model is able to generalize over arbitrary
5 n9 L& Y7 u+ _/ C! c) c) \topologies, routing schemes and traffific intensity. In our eval
uation, we show that RouteNet is able to predict accurately
the delay distribution (mean delay and jitter) and loss even in
3 Z0 Y0 V) _) N$ j: w$ s7 Y; _+ h  Xtopologies, routing and traffific unseen in the training (worst case
MRE = 15.4%). Also, we present several use cases where we
leverage the KPI predictions of our GNN model to achieve
effificient routing optimization and network planning.
7 r( l9 W0 u5 ^/ T! B9 R