From Knowledge Generation To Science-based Innovation Novel ns-3 Model Enabling Broadband Maritime Communications Simulation of Electromagnetic The Mare-Fi Project Wireless Underground Networks Rui Campos, Mário Lopes, Luciano Santos, Filipe Teixeira, Sérgio Conceição, Filipe Ribeiro, Rui Campos, Manuel Ricardo Jorge Mamede, Manuel Ricardo WNS3 2015, Barcelona, Spain 3rd Fórum do Mar, Porto 13 th May 2015 May 2013 Research and Technological Development | Technology Transfer and Valorisation | Advanced Training | Consulting Pre-incubation of Technology-based Companies
Outline • Introduction • Objectives • Underground propagation models • Work methodology • Results • Conclusions Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 2
Introduction - WUN • Wireless Underground Networks ( WUN) consist of – Nodes buried underground and aboveground – Wireless links – Two Propagation media • 4 types of links – Underground-to-Underground (U2U) – Aboveground-to-Aboveground (A2A) – Underground-to-Aboveground (U2A) – Aboveground-to-Underground (A2U) Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 3
Introduction - WUN • Playing fields, Agriculture – Monitor soil water content, temperature – Automatically control irrigation systems • Security – Border surveillance • Infrastructure monitoring – Pipeline monitoring Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 4
Introduction - ns-3 • No network simulators available for WUN • ns-3 characteristics – Open source – Experience in our research group in using ns-3 – Highly modular – Well documented – Allow easily integration of user implemented models – Well accepted by the research community Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 5
Objectives of the work • Study existing underground propagation models • Improve ns-3 towards WUN • Validate ns-3 models against results obtained in testbeds Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 6
Path loss in soils • Free Space Path Loss, Friis equation [dB] 2 4𝜌𝑒 𝑄 𝑠 = 𝑄 𝑢 + 𝐻 𝑢 + 𝐻 𝑠 − 𝑀 0 , 𝑀 0 = 10 log – 𝜇 0 • Path Loss in Soil – 𝑄 𝑠 = 𝑄 𝑢 +𝐻 𝑢 + 𝐻 𝑠 − 𝑀 𝑞 , 𝑀 𝑞 = 𝑀 0 + 𝑀 𝑡oil , 𝑴 𝐭𝐩𝐣𝐦 = 𝑴 𝜸 + 𝑴 𝜷 – Propagation constant (in soil) • 𝛿 = 𝛽 + 𝑘𝛾 • 𝛿 depends on soil dielectric properties type soil, water content – Attenuation constant 𝛽 [𝑛 −1 ] → 𝜇 = 2𝜌 – Phase constant 𝛾 𝑠𝑏𝑒. 𝑛 −1 → v = 𝜇𝑔 𝛾 2 𝜇 0 , 𝑀 𝛽 = 10 log 𝑓 2𝛽𝑒 – 𝑀 𝛾 = 10 log 𝜇 Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 7
Two-ray U2U model • Single direct ray – Lsl = = • Two-rays Δ𝜚 = 2𝜌Δ𝑠 Δ𝑠 = 𝑠 1 + 𝑠 2 𝑆: reflection coefficient soil−air 𝜇 Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 8
Three-ray U2U model 𝑄𝑒 = 𝑄𝑢 + 20 log 𝜇 𝑡 − 20 log 𝑠1 − 8.69𝛽𝑠1 − 45 𝑄𝑠 = 𝑄𝑢 + 20 log 𝜇 𝑡 − 20 log 𝑠2 − 8.69𝛽𝑠2 + 20𝑚𝑝Γ − 45 𝑄𝑚 = 𝑄𝑢 + 20 log 𝜇 𝑡 − 40 log 𝑒ℎ − 8.69𝛽 ℎ𝑢 + ℎ𝑠 + 20𝑚𝑝𝑈 − 30 𝑸𝒆 𝑸𝒔 𝑸𝒎 𝟐𝟏 + 𝟐𝟏 𝟐𝟏 + 𝟐𝟏 𝑸𝒔 = 𝟐𝟏 𝐦𝐩𝐡( 𝟐𝟏 𝟐𝟏 ) Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 9
A2U model 𝑀 𝑏 = 20 log 𝑔 + 20 log 𝑒1 − 147.56 𝑀 𝑣 = 6.4 + 20 log 𝑒2 + 20 log 𝛾 + 8.69𝛽𝑒2 (𝑑𝑝𝑡𝜄𝑗+ 𝜁 ′ −𝑡𝑗𝑜 2 𝜄𝑗) 2 𝑀 𝑏−𝑣 = 10𝑚𝑝 4𝑑𝑝𝑡𝜄𝑗∗ 𝜁 ′ −𝑡𝑗𝑜 2 𝜄𝑗 𝑀 𝑢𝑝𝑢𝑏𝑚 = 𝑀 𝑏 + 𝑀 𝑣 + 𝑀 𝑏−𝑣 − 10 log 𝜓 2 • Rayleigh distribution Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 10
U2A model 𝑀 𝑏 = 20 log 𝑔 + 20 log 𝑒2 − 147.56 𝑀 𝑣 = 6.4 + 20 log 𝑒1 + 20 log 𝛾 + 8.69𝛽𝑒1 𝑀 𝑣−𝑏 = 10𝑚𝑝 ( 𝜁 ′ − 1) 2 4 𝜁 ′ 𝑀 𝑢𝑝𝑢𝑏𝑚 = 𝑀 𝑣 + 𝑀 𝑏 + 𝑀 𝑣−𝑏 − 10 log 𝜓 2 • Rayleigh distribution Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 11
Methodology • Include propagation models into ns-3 • Develop a new Wi-Fi channel with two propagation media – Soil propagation medium – Air propagation medium • Carry-out network simulations using the new models • Compare simulation results against testbed results previously obtained at INESC TEC • Conclude about validity of models Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 12
New models in ns-3 • Estimate soil dielectric constant – estimateSoilDielectricConstantSMDM – estimateSoilDielectricConstantMBSDM – Based on type of soil and water volume contents • Estimate path loss between two nodes – ns3::UndergroundPathLossModel – U2U: 2 and 3 ray models – Hybrid: U2A e A2U – A2A • Estimate propagation delay between two nodes – ns3::UndergroundConstantSpeedPropagationDelayModel – Using velocity of EM wave in the soil, t = 𝑒 𝑤 = 𝑒 𝛾 𝜇𝑔 = 2𝜌𝑔 𝑒 Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 13
New models in ns-3 • New Wi-Fi channel – ns3::UndergroundWifiChannel – Supports two different propagation media – Use underground path loss model for underground links – Reuse ns-3 propagation models for over the air links • New Wi-Fi phy – ns3::UndergroundWifiPhy – Uses the ns3::UndergroundWifiChannel – Similar to the ns3::YansWifiPhy • New Wi-Fi helper – ns3::UndergroundWifiPhyHelper – ns3::UndergroundWifiChannelHelper Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 14
New models in ns-3 Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 15
Simulations – network topologies • 2 nodes, single wireless link • Nodes running UDP/IP/802.11g • Traffic source: ns-3 OnOff (CBR) • Traffic sink: ns-3 DataSink • Bands: 2.4 GHz | 433 MHz Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 16
Simulations – network topologies • Transmission power: 20 dBm • Antenna gain: 2 dBi (transmitter) • Antenna gain: 3 dBi (receiver) • U2U: 2 nodes buried at 20 | 30 cm • U2A, A2U: node buried at 35 cm • Air node at 2.5 m height • Soil: loam Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 17
Simulations – metrics • Performance metrics – RSSI – Throughput – Packet Loss Ratio (PLR) – Delay – Delay Jitter • Measured using ns-3 Flow monitor Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 18
Simulation results - U2U, 2.4 GHz, RSSi • RSSi difference 2 ray: 11 dBm @ 20 cm | 14 dBm @ 30 cm • RSSi difference 3 ray: 5 dBm @ 20 cm | 8 dBm @ 30 cm • Distance difference 3 ray : 21% @ 20 cm | 21% @ 30 cm • 2-ray model not adequate for high horizontal distances – Lateral wave is the dominant component (d > 1m) – 3 ray model should be used Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 19
Simulation results - U2U, 2.4 GHz, Throughput • Difference 2 ray: 4.5 Mbit/s @ 20 cm | 5 Mbit/s @ 30 cm • Difference 3 ray: 7 Mbit/s @ 20 cm | 4 Mbit/s @ 30 cm • Higher precision for high depths • 2 ray model with results only until 1.1 m Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 20
Simulation results - U2U, 2.4 GHz, Delay RSS sand RSS loam 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 -20 -20 -40 -40 RSS (dBm) RSS (dBm) -60 -60 -80 -80 -100 -100 -120 -120 Horizontal distance (m) Horizontal distance (m) A2U U2A A2U Experimental U2A Experimental A2U U2A A2U Experimental U2A Experimental • Experimental results with ping – Round-trip time (RTT) • Simulation results measure packet delay Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 21
Simulation results - U2U, 2.4 GHz, Jitter RSS sand RSS loam 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 -20 -20 -40 -40 RSS (dBm) RSS (dBm) -60 -60 -80 -80 -100 -100 -120 -120 Horizontal distance (m) Horizontal distance (m) A2U U2A A2U Experimental U2A Experimental A2U U2A A2U Experimental U2A Experimental • Difference 2 ray: 0.06 ms @ 20 cm | 0.04 ms @ 30 cm • Difference 3 ray: 0.12 ms @ 20 cm | 0.11 ms @ 30 cm • Higher precision for high depths • 2 ray model with results only until 1.1 m Novel ns-3 Model Enabling Simulation of Electromagnetic Wireless Underground Networks 22
Recommend
More recommend
