Topology generators are a key asset for researchers in computer science and telecommunications that often need to test network protocols or distributed systems in simulated environments that resemble real scenarios. Despite that, in the research area of distributed wireless networks still many works use very simplistic models that do not have the characteristics of the currently existing large-scale wireless mesh networks. The only topology generator that tries to produce synthetic graphs that look like real networks is NPART [1].In this work we test the characteristics of NPART against another, completely different approach: TrueNets [2]. TrueNets uses accurate data representing land surface of a real world location to create topologies of networks that could actually exist. The downside of TrueNets is twofold: it can be used only when data-sets are available and generating topologies is computationally intensive. We show that using aggregate data from TrueNets we are able to improve NPART. We call the new generator NPART+ and we show that compared to topologies generated with TrueNets, NPART+ (or its variants) improves NPART in several metrics, but still it can not match the accuracy of TrueNets.

NPART+: Improving Wireless Network Topology Generators with Data from the Real World

Gemmi G.;Maccari L.
2020-01-01

Abstract

Topology generators are a key asset for researchers in computer science and telecommunications that often need to test network protocols or distributed systems in simulated environments that resemble real scenarios. Despite that, in the research area of distributed wireless networks still many works use very simplistic models that do not have the characteristics of the currently existing large-scale wireless mesh networks. The only topology generator that tries to produce synthetic graphs that look like real networks is NPART [1].In this work we test the characteristics of NPART against another, completely different approach: TrueNets [2]. TrueNets uses accurate data representing land surface of a real world location to create topologies of networks that could actually exist. The downside of TrueNets is twofold: it can be used only when data-sets are available and generating topologies is computationally intensive. We show that using aggregate data from TrueNets we are able to improve NPART. We call the new generator NPART+ and we show that compared to topologies generated with TrueNets, NPART+ (or its variants) improves NPART in several metrics, but still it can not match the accuracy of TrueNets.
2020
2020 Mediterranean Communication and Computer Networking Conference, MedComNet 2020
File in questo prodotto:
File Dimensione Formato  
camera-ready-final.pdf

non disponibili

Tipologia: Documento in Post-print
Licenza: Accesso chiuso-personale
Dimensione 884.32 kB
Formato Adobe PDF
884.32 kB Adobe PDF   Visualizza/Apri

I documenti in ARCA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10278/3743146
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? ND
social impact