Interference of Simulated IEEE 802.11 Links with Directional Antennas Michael Rademacher, Karl Jonas first.last@h-brs.de Hochschule Bonn-Rhein-Sieg March 30, 2017, IEEE Wireless Days, Porto, Portugal 1
Table of Content Introduction Network Architecture Related Work Methodology Results Conclusion and Future Work 2
A Different Technology for Broadband in Rural Areas Commercial off-the-shelf (COTS) WiFi/802.11 transmitter and directional antennas ... ◮ Inexpensive (low CAPEX) ◮ Free-to-use band (low OPEX) ◮ Low energy consumption (low OPEX) ◮ Well developed and documented ... used in a controlled Multi-Radio Multi-Channel Wireless Mesh Network (WMN) 1 ◮ Our main research fields: - Channel Allocation [1] - MAC-layer optimization [2–4] - Propagation modeling [5, 6] - Wireless-SDN [7] 1WiFi-based Long Distance networks (WiLD) [8] or Coordinated Wireless Backhaul Network (WBN) [9]. 3
WiLD Architecture and Motivation Example of a WiLD Rhein-Sieg testbed ◮ The nodes imply a static placement (vs. MANET) ◮ The nodes are equipped with multiple mesh-radios (vs. SR/DR-WMN) ◮ ≈ 3000 of these networks in the US [10]. Vendors: Ubiquity, Cambium or Mimosa [11–13]. Motivation: Multiple self-interference effects despite high gain antennas in our long-distance testbed. 4
Related Work: Three Types of Interference a) b) c) 2 3 E 0 1 2 0 1 0 1 Different types of interference a) Intra-flow b) Inter-flow c) External Interference [14]. 0 1 2 3 Cone (3D) / pie-slice (2D) model of interference. A transmission from 0 to 1 does not interfere with a transmission from 2 to 3. Widely used [15–18]. ◮ Based on testbed observations, this model seems idealistic [19]. ◮ Interference measurement in the testbed are challenging -> Simulation 5
Methodology - ns-3 Simulation ◮ Based on a recent ns-3 module called SpectrumWiFiPhy [20] ◮ New ’File-Antenna’ model: Selected NS-3 parameters Parameter Choice Standard 802.11n Freq./Width 5180/20 MHz MAC-Layer AdhocWifiMac Station Manager ConstantRate Data Rate MCS7 Control Rate MCS3 A-MPDU-Size 8192 Bytes RTS/CTS Disabled (a) Antenna diagram data-sheet based on .ant file TxPower 5 dBm Simulation Time 30 s Traffic UDP, saturated Distance 1 km Payload Size 1450 Bytes Routing IPv4 Static Delay Model ConstantSpeed Error-Rate Model Nist (b) Verification of the antenna diagram with ns-3 Propagation Model Friis MAC-Layer timings Depended Antenna: Ubiquity PowerBridge M5 6
Methodology - Setup ◮ All antennas are perfectly aligned. Same channel everywhere. ◮ Trace signals directly before the radio (SpectrumWiFiPhy module) ◮ Energy Detection Threshold (EDT): ◮ A signal < EDT = noise. Carrier sensing reports idle medium. ◮ Notation for interference: source-node → interference-node. Inter-Flow Intra-Flow ◮ α = 0 ... 180 ◦ ◮ ∆ y = 5 ... 1250 m ◮ n 0 → n 2. n0 to n1 interferes at n2. ◮ n 1 → n 3. n1 to n0 interferes at n3. ◮ n 0 → n 3. n0 to n1 interferes at n3. ◮ n 1 → n 2. n1 to n0 interferes at n2. 7
Results - Intra-Flow - One Stream ! n 0 → n 2, n 0 → n 3: wave form according to the side-lobes of the antenna. ! Fluctuating throughput. 1 At α ≈ 11 ◦ : ◮ n 0 → n 3 > EDT: n3 is vulnerable to the transmission from n0. 50 50 −40 n0 ⇒ n3 n0 → n2 n0 → n3 ◮ n 0 → n 2 < EDT: n2 is not Throughput [Mbps] Throughput [Mbps] 40 40 −54 able to detect a transmission Signal [dBm] from n0 to n1. 30 30 −68 ◮ n0 and n2 are const. 20 20 −82 transm. Interference at n3. 10 10 −96 2 At α ≈ 21 ◦ : ◮ n 0 → n 2, n 0 → n 3 < EDT. 0 0 −110 0 0 20 20 40 40 60 60 80 80 100 100 120 120 140 140 160 160 180 180 Angle, α [ ° ] Angle, α [ ° ] ◮ No interference UDP: n 0 ⇒ n 3 via n1 and n2 8
Results - Intra-Flow - Two Streams ◮ Similar results compared to the one stream case. 50 50 −40 n0 ⇒ n3 n3 ⇒ n0 n0 → n2 n0 → n3 Including the interesting Throughput [Mbps] Throughput [Mbps] angles α ≈ 11 ◦ and α ≈ 21 ◦ . 40 40 −54 Signal [dBm] 30 30 −68 20 20 −82 10 10 −96 0 0 −110 0 0 20 20 40 40 60 60 80 80 100 100 120 120 140 140 160 160 180 180 Angle, α [ ° ] Angle, α [ ° ] Two UDP streams: n 0 ⇒ n 3 and n 3 ⇒ n 0 9
Results - Inter-Flow - Same Direction ◮ n 1 → n 3 decreases, n 1 → n 2 wave form 1 At ∆ y ≈ 180 m ◮ n 1 → n 3 > EDT ◮ n 1 → n 2 < EDT ◮ If n3 & n1 choose the same backoff slot there is no collision at n0 & n2. 2 At ∆ y ≈ 385 − 405 m or 50 50 −40 60 n1 ⇒ n0 n1 → n3 avg. Backoff ∆ y ≈ 570 − 690 m : Average Backoff [Slots] n3 ⇒ n2 n1 → n2 Throughput [Mbps] Throughput [Mbps] 40 40 −54 48 ◮ n 1 → n 3, n 1 → n 2 < EDT Signal [dBm] 30 30 −68 36 3 At ∆ y ≈ 405 − 575 m or ∆ y ≈ 690 − 940 m : 20 20 −82 24 ◮ n 1 → n 3 < EDT 10 10 −96 12 ◮ n 1 → n 2 > EDT 0 0 −110 0 0 0 200 200 400 400 600 600 800 800 1000 1000 1200 1200 ◮ avg. backoff slots at n1/n3 ∆ y [m] ∆ y [m] increase Two UDP streams, n 1 ⇒ n 0 and n 3 ⇒ n 2 (same direction). 10
Results - Inter-Flow - Different direction ◮ The DCF works for all distances: Average Backoff. 1 At ∆ y ≈ 180 m ◮ n 1 → n 3 > EDT 50 50 −40 60 ◮ n 1 → n 2 < EDT n1 ⇒ n0 n1 → n3 avg. Backoff Average Backoff [Slots] n2 ⇒ n3 n1 → n2 Throughput [Mbps] Throughput [Mbps] 40 40 −54 48 ◮ Remarkable difference: One Signal [dBm] flow dominates the other. 30 30 −68 36 Further analysis is needed. 20 20 −82 24 10 10 −96 12 0 0 −110 0 0 0 200 200 400 400 600 600 800 800 1000 1000 1200 1200 ∆ y [m] ∆ y [m] Two UDP streams, n 1 ⇒ n 0 and n 2 ⇒ n 3 (different direction). 11
Conclusion and Future Work Done: � Results for interference of (simple) WiFi links with directional antennas � Load real-world antenna diagrams in ns-3 ! Side-lobes of antennas have a significant impact on interference ◮ The widely used cone model would lead to significantly different results not reflecting real-world build-ups. ◮ Source code, logs, and high-quality images free available online 2 Current work and next steps: ◮ Load larger topologies based on Google-Earth .kmz format ◮ Calculate propagation based on satellite data (incl. hills) ◮ Evaluate different Channel Allocation algorithms for WiLD 2 http://mc-lab.inf.h-brs.de/wild.xml 12
Thank you very much! Are there any questions? M.Sc. Michael Rademacher Fachbereich Informatik Grantham-Allee 20 53757 Sankt Augustin Tel. +49 2241 865 151 Fax +49 2241 865 8151 michael.rademacher@h-brs.de www.h-brs.de Acknowledgment: This work has been funded by the Federal Ministry of Education and Research of the Federal Republic of Germany (Foerderkennzeichen 16KIS0332). 13
References and Acronyms I [1] M. Rademacher et al. “Towards Centralized Spectrum Allocation Optimization for Multi-Channel Wireless Backhauls”. In: e-Infrastructure e-Services Dev. Ctries. Ed. by J. Nungu, Amos and Pehrson, Bjorn and Sansa-Otim. Springer International Publishing, 2015, pp. 74–83. [2] M. Rademacher, M. Kretschmer, and K. Jonas. “Exploiting IEEE802.11n MIMO Technology for Cost-Effective Broadband Back-Hauling”. In: 5th Int. IEEE EAI Conf. on e-Infrastructure and eServices for Developing Countries. 2013. [3] M. Rademacher. Performance estimation and optimization of the IEEE802.11 MAC layer for long distance point-to-point links. Tech. rep. Hochschule Bonn-Rhein-Sieg, 2015, p. 109. [4] M. Rademacher, M. Chauchet, and K. Jonas. “A Token-Based MAC For Long-Distance IEEE802.11 Point-To-Point Links”. In: VDE ITG-Fachbericht Mobilkommunikation (2016). [5] M. Rademacher and M. Kessel. “An Empirical Evaluation of the Received Signal Strength Indicator for fixed outdoor 802.11 links”. In: VDE ITG-Fachbericht Mobilkommunikation 20 (2015), pp. 62–66. [6] M. Rademacher et al. “Experimental Results For the Propagation of Outdoor IEEE802.11 Links”. In: VDE ITG-Fachbericht Mobilkommunikation 22 (2016). [7] M. Rademacher et al. “Experiments with OpenFlow and IEEE802.11 Point-to-Point Links in a WMN”. In: ICWMC - Twelfth Int. Conf. Wirel. Mob. Commun. IARIA, 2016, pp. 99–105. isbn : 978-1-61208-514-2. [8] J. Simo et al. “Distance Limits in IEEE 802.11 for Rural Networks in Developing Countries”. In: Proc. Conf. on Wireless Rural and Emergency Commun. 2007, p. 5. [9] C. Mannweiler. “Context-Enabled Optimization of Energy-Autarkic Networks for Carrier-Grade Wireless Backhauling”. PhD thesis. TU Kaiserslautern, 2015, p. 259. 14
Recommend
More recommend
