5.9 GHz V2X Modem Performance Challenges with Vehicle Integration

Size: px

Start display at page:

Download "5.9 GHz V2X Modem Performance Challenges with Vehicle Integration"

Audrey Merritt
5 years ago
Views:

1 5.9 GHz V2X Modem Performance Challenges with Vehicle Integration October 15th, 2014

2 Background V2V DSRC Why do the research? Based on p MAC PHY ad-hoc network topology at 5.9 GHz. Effective Isotropic Radiated Power (EIRP) limited to +33 dbm (2 watts) in U.S. for the On-Board Unit (OBU) transmitters. Communication range up to several hundred meters under near-optimal, unobstructed conditions not always possible. Some factors affecting communication range: Environmental conditions (e.g., signal blockage and reflections) Transmitter (TX) power and Receiver (RX) sensitivity Antenna gain and cable losses (both TX and RX vehicles) Antenna configuration (single antenna vs. diversity system) Receiver baseband decoding (channel equalization during mobile reception) The result: Varying amounts of notification time to the driver of potentially hazardous situations. From a vehicle integration perspective, what can be done to optimize the communication range performance?

3 Theoretical Communication Range The free-space path loss limits the maximum theoretical communication range. The received signal power can be calculated from the Friis transmission equation: P r = P t + G t + G r + 20 log 10 ( ) Example: Assume two OBU s with the following characteristics: where: P r = Power at receiving antenna (dbm) P t = Power at transmitter antenna (dbm) G t = Transmitter antenna gain (dbi) G r = Receiver antenna gain (dbi) = Wavelength (meters) R = Distance between antennas (meters) Parameter Value Comments Frequency GHz, Ch. 178 = meters TX Power +23 dbm RX Sensitivity -92 dbm. AWGN Channel Antenna Gain 0 dbi Both TX and RX vehicles Cable Loss 0 db Both TX and RX vehicles Solving for R, the maximum theoretical communication range = meters

4 But..Wireless Channel Impairments The ideal free-space conditions assumed in the Friis equation are rarely achieved. Additional real-world effects include: Signal reflections and scattering from buildings and other objects. Channel characteristics rapidly change due to vehicle motion. Gain and phase offsets between OFDM subcarriers in the 10 MHz channel. Baseband equalization / channel tracking algorithms required. Attenuation due to obstructions in the transmission path. Greater than -20 db attenuation due to building 5 GHz. Antenna polarization mismatches (e.g., ground reflections, etc.) RX sensitivity calculation is often specified in a static Additive White Gaussian Noise (AWGN) channel. (Does not account for time-varying channel conditions).

5 Why test a Vehicle? - Integration Considerations Empirical data has found the reliable communication range is typically only a fraction of the maximum theoretical distance. Questions from a vehicle integration perspective: How does the V2V communication range vary in different environments? What impact does additional system loss have on the V2V communication range (e.g., antenna cable loss, reduced RX sensitivity, etc.)? How much extra communication range is achievable with a dual antenna diversity system? These questions are not easily answered by simplistic models! One solution: Perform pseudo-controlled field-evaluations and vary the system parameters. Use the empirical results to drive the system integration configuration.

6 What did we test? - Field Evaluation Scenarios Summary of field evaluation test conditions: Wide range of conditions - from completely unobstructed to NLOS with various amount of local reflecting structures. Test Use Case RX Vehicle TX Vehicle Environmental Conditions 1a Intersection Awareness Stationary 40 Km/Hr Suburban industrial area. Obstructed NLOS path with minimal reflecting structures. 1b Intersection Awareness 24 Km/Hr 40 Km/Hr Suburban industrial area. Obstructed NLOS path with minimal reflecting structures. 2 Intersection Awareness Stationary 40 Km/Hr Urban business area. Obstructed NLOS path with many nearby reflecting structures. 3a Vehicle Awareness Stationary 56 Km/Hr Open, unobstructed area. No nearby reflecting structures. 3b Vehicle Awareness 56 Km/Hr 56 Km/Hr Open, unobstructed area. No nearby reflecting structures. Not intended to be all-inclusive of every possible environment!

Installed HW - Theoretical Communication Range Configuration of radio parameters: Parameter Value Comments Frequency 5.890 GHz, Ch. 178 = 0.0509 meters TX Power +21 dbm. ---- RX Sensitivity -92 dbm.

7 Installed HW - Theoretical Communication Range Configuration of radio parameters: Parameter Value Comments Frequency GHz, Ch. 178 = meters TX Power +21 dbm RX Sensitivity -92 dbm. Est. value AWGN Channel Antenna Gain -1.0 dbi Both TX and RX vehicles Cable Loss -2.0 db Both TX and RX vehicles Calculation of communication range from Friis equation: R = ( meters / 4 p) * 1 / 10 [(-92 dbm (+21 dbm - 3 dbi - 3 dbi)) / 20] R = meters = Maximum theoretical free-space communication range

Example 1a Calculate straight line communication distance @ 10% PER: TX vehicle transmits BSM @ 10 Hz rate (including GPS coordinates).

8 Example 1a Calculate straight line communication 10% PER: TX vehicle transmits 10 Hz rate (including GPS coordinates). RX vehicle continuously logs GPS coordinates. Post-processing compares TX and RX log files to calculate 10% PER distance. Test Case 1a

9 10% PER (Meters) Test Results Unobstructed Single Antenna Configuration The unobstructed routes provided the greatest range: Avg. of 461 meters (~51% of free-space value) with one vehicle moving. Avg. of 300 meters (~33% of free-space value) with both vehicles moving. Equivalent excessive loss of -6 db to -10 db over the free-space value: Presumed to result from ground reflections, transient signal fading, and non-gaussian channel conditions (not included in Friis equation) Theoretical Free-space Communication Range Unobstructed Use Cases Obstructed Use Cases Single Antenna a 1b 2 3a 3b Test Case

10 10% PER (Meters) Test Results Obstructed Single Antenna Configuration The obstructed NLOS routes had the smallest range: Avg. between 73 to % PER. Equivalent excessive loss of -19 db to -22 db over the free-space value: Presumed to be dominated by building attenuation and scattering loss. (Other studies have cited building attenuation greater than GHz) Theoretical Free-space Communication Range Unobstructed Use Cases Obstructed Use Cases Single Antenna a 1b 2 3a 3b Test Case

11 10% PER (Meters) Dual-Antenna Diversity Improvement Diversity improvement achieved in both LOS and NLOS environments: Between 4% to 24% range increase in obstructed NLOS cases. Between 0% and 26% range increase in unobstructed cases. Physical mechanisms behind the environmental dependency are not fully understood, but results are directionally consistent with other field trials Unobstructed Use Cases Obstructed Use Cases Single Antenna Dual Antenna a 1b 2 3a 3b Test Case

12 Conclusions Maximum reliable V2V communication range is a function of many variables and is only a fraction of the theoretical, free-space value. Communication range is dominated by the physical environment outside of the control of the vehicle Distances can range between 73 meters and 461 PER have been measured with a single antenna and +21 dbm TX power. Dual antenna diversity increased the communication range up to about 26% in some test cases performance improvements will carry a cost To understand the solution you have to understand the problem

13 Page 13 Slide 16

The Radio Channel. COS 463: Wireless Networks Lecture 14 Kyle Jamieson. [Parts adapted from I. Darwazeh, A. Goldsmith, T. Rappaport, P.

The Radio Channel. COS 463: Wireless Networks Lecture 14 Kyle Jamieson. [Parts adapted from I. Darwazeh, A. Goldsmith, T. Rappaport, P. The Radio Channel COS 463: Wireless Networks Lecture 14 Kyle Jamieson [Parts adapted from I. Darwazeh, A. Goldsmith, T. Rappaport, P. Steenkiste] Motivation The radio channel is what limits most radio