The Impact of Directional Antenna Models on Simulation Accuracy Eric Anderson, Gary Yee, Caleb Phillips, Douglas Sicker, and Dirk Grunwald eric.anderson@colorado.edu University of Colorado Department of Computer Science 25 June 2009 � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 1 / 21
Outline Introduction 1 Physical Layer Simulation Current Models of Directivity EDAM – The Effective Directivity Antenna Model 2 Error Idea Parameters Case Study 3 Overview Metrics Results Conclusions 4 � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 2 / 21
Phy-Layer Simulation Framework Environmental Factors Antenna Properties Line of Sight Line of Sight In/Outdoors In/Outdoors Terrain Motion Path Loss Fading Directivity Model Two−ray COST 231 ITU1238 ... Rician Rayleigh ... EDAM Antenna Gain Only None Position Temporal Variation Direction / Orientation � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 3 / 21
Directivity – Current Models � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 4 / 21
Directivity – Current Models Fading & path loss Node a gain Node b gain P rx = P tx ∗ X ∗ f a ( θ 1 ) ∗ f b ( θ 2 ) � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 4 / 21
Example �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 5 / 21
Example �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� f a ( θ 1 ) f b ( θ 2 ) � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 5 / 21
Example �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� ) f θ 3 b ( ( θ 4 f a ) � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 5 / 21
Example �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� �������������������������������������������� � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 5 / 21
Outline Introduction 1 Physical Layer Simulation Current Models of Directivity EDAM – The Effective Directivity Antenna Model 2 Error Idea Parameters Case Study 3 Overview Metrics Results Conclusions 4 � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 6 / 21
How Bad Is It? Patch−Panel Antenna 10 0 dB Relative to Peak Mean −10 −20 −30 −40 Patch−Outdoor−B Patch−Outdoor−A Patch−Indoor−A −50 Reference 0 25 50 75 105 140 175 210 245 280 315 350 Angle, Degrees Counterclockwise � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 7 / 21
How Bad Is It? 24dBi Parabolic Dish, Indoors 10 0 dB Relative to Peak Mean −10 −20 −30 −40 Parabolic−Indoor−C Parabolic−Indoor−B Parabolic−Indoor−A −50 Reference 0 21 50 76 106 140 171 205 236 270 301 335 Angle, Degrees Counterclockwise � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 7 / 21
EDAM – The Effective Directivity Antenna Model Key Idea: Model offset between the expected (“pure”) antenna gain and observed effect. Offset is environment-specific and impractical to compute (but easy to measure) Distribution of offsets can be predicted well. Construct distributions, sample, repeat. � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 8 / 21
EDAM Distribution Parameters Mean offset is based on antenna gain and environment type. Offset variance and packet signal variance are based on environment type. See “Modeling environmental effects on directionality in wireless networks,” WiNMee 2009 (tomorrow). � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 9 / 21
Outline Introduction 1 Physical Layer Simulation Current Models of Directivity EDAM – The Effective Directivity Antenna Model 2 Error Idea Parameters Case Study 3 Overview Metrics Results Conclusions 4 � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 10 / 21
Case Study: Data Striping for Security Compare simulated and real results (S. Lakshmanan et. al, 2008) Propagation-sensitive system: APs beam-form to client Packet must be received from all APs to decode. � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 11 / 21
Experimental Design Directivity Path Loss Fading EDAM Two ray Log-normal Pure ITU-1238 Ricean None/Omni Implicit Gaussian None Directivity Null Hypotheses 1 “Pure:” Antenna gain fully describes directional effects. 2 “None/Omni:” There is no predictable directional effect. (indoor) � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 12 / 21
“Ground Truth” Measurement Measurement Points Signal strength from 192.168.99.2 35 30 25 y coordinate 20 15 30 −20 20 10 Y position −40 Signal strength, dBm 10 5 −60 −80 0 0 −100 0 5 10 15 20 25 30 � X position 0 5 10 15 20 25 30 x coordinate Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 13 / 21 foo
Application Performance: Vulnerability Area Application goal: Minimize physical area of reception. Actual performance, outdoor test Metric: Fraction of Complete packet reception probability packets decodable at 1.0 each location. Distribution, not scalar. 30 0.8 CDF of packet reception probability by location, outdoor 1.0 0.6 y coordinate 20 0.8 Cumulative fraction of nodes 0.4 0.6 10 0.4 0.2 0.2 Observed 0.0 0.0 Sim: NA, NA, NA 5 10 15 20 25 30 x coordinate 0.0 0.2 0.4 0.6 0.8 � Fraction of packets received completely Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 14 / 21
Simulation Accuracy: Distribution Similarity CDF of packet reception probability by location, outdoor 1.0 Accuracy = similarity of 0.8 application performance to reality. Cumulative fraction of nodes 0.6 Kolmogorov-Smirnov (KS) test: Maximum divergence 0.4 between distributions. Evaluation: Factorial 0.2 ANOVA on KS test across all configurations. Observed 0.0 Sim: pure, two ray, lognormal 0.0 0.2 0.4 0.6 0.8 1.0 Fraction of packets received completely � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 15 / 21
Results (subset) EDAM Pure Gain Indoor Outdoor � Eric Anderson (CU Boulder) The Impact of Directional . . . WiOpt ’09 16 / 21
Recommend
More recommend