Software defined network inference with evolutionary optimal observation matrices

Abstract

A key requirement for network management is the accurate and reliable monitoring of relevant network characteristics. In today's large-scale networks, this is a challenging task due to the scarcity of network measurement resources and the hard constraints that this imposes. This paper proposes a new framework, called SNIPER, which leverages the flexibility provided by Software-Defined Networking (SDN) to design the optimal observation or measurement matrix that can leads to the best achievable estimation accuracy using Matrix Completion (MC) techniques. To cope with the complexity of designing largescale optimal observation matrices, we use the Evolutionary Optimization Algorithms (EOA) which directly target the ultimate estimation accuracy as the optimization objective function. We evaluate the performance of SNIPER using both synthetic and real network measurement traces from different network topologies and by considering two main applications for per-flow size and delay estimations. Our results show that SNIPER can be applied to a variety of network performance measurements under hard resource constraints. For example, by measuring only 8.8% of all per-flow path delays in Harvard network [1], congested paths can be detected with probability of 0.94. To demonstrate the feasibility of our framework, we also have implemented a prototype of SNIPER in Mininet.

Artificial intelligence enabled software‐defined networking: a comprehensive overview

“…A large variety of meta-heuristic algorithms such as ACO [183][184][185][186][187][188][189][190][191], EAs [192][193][194], GAs [195][196][197][198][199][200][201][202][203], PSO [204][205][206][207][208][209], SA [210][211][212], bee colony optimisation-based [213,214], whale optimisation [215,216], FFO [217], BA [85], TLBO [87] and GWO [207] were used in SDN.…”

“…Malboubi et al [201,202] presented, SNIPER, a framework for designing the optimal observation matrix that can lead to the best estimation accuracy using matrix completion techniques. The authors used GA and PSO to deal with large-scale optimal observation matrices.…”

“…GAs in SDN: GA were mainly used for routing [195], virtual network planning [196], optimising the cost of deploying DPI functions in SDN [197], load balancing [198–200], designing optimal observation matrices [201], resource reallocation for SDN‐based data centres [202] and congestion avoidance [203].…”

Artificial Intelligence Enabled Software Defined Networking: A Comprehensive Overview

“…A large variety of meta-heuristic algorithms such as ant colony optimization [187][188][189][190][191][192][193][194][195], evolutionary algorithms [196][197][198], genetic algorithms [199][200][201][202][203][204][205][206][207], particle swarm optimization [208][209][210][211][212][213], simulated annealing [214][215][216], bee colony optimizationbased [217,218], whale optimization [219,220], firefly optimization [221], bat algorithm [88], teaching-learning-based optimization [90] and grey wolf optimization [222] were used in SDN.…”

“…Genetic algorithm were mainly used for routing [199], virtual network planning [200], optimizing the cost of deploying deep packet inspection functions in SDN [201], load balancing [202][203][204], designing optimal observation matrices [205], resource reallocation for SDN-based data centres [206], and congestion avoidance [207].…”