Data for Investigating the V2V millimeter-wave channel near a vehicular headlight in an engine bay

Both an omnidirectional antenna and a directional horn antenna (RX) were alternately positioned in an engine bay, near the right vehicular headlight. The transmitting directional horn antenna (TX) was placed at equal height, in 10 steps, on a circular path with a radius of 120 cm around the RX, and manually boresight aligned with RX. Two TX/RX antenna pairs were used: Pair 1 consisted of a double ridged guide horn antenna with a typical gain of 20 dBi and a half-power beam width (HPBW) of 20 for frequencies between 24 and 40 GHz, and an ultra-wideband omnidirectional antenna with a typical gain of 6 dBi and an elevation HPBW of 20 for the same frequency range.

Pair 2 consisted of two waveguide horn antennas for the frequency range from 60 to 90 GHz, one with a typical gain of 25 dBi and a HPBW of 10, and one with a typical gain of 10 dBi and a HPBW of 55. Our channel sounding equipment uses the pulse compression method and consists of an R&S SMW200A Vector Signal Generator, an R&S SMZ90 Frequency Multiplier, an R&S FSW85 Signal and Spectrum Analyzer and an R&S RTO2044 Digital Oscilloscope.

The resulting I/Q data are forwarded to the R&S Channel Sounding Software, which autocorrelates the received I/Q data with the originally transmitted signal. The equipment is synchronized by means of direct, calibrated coaxial connections of their time reference and trigger channels. A 2 GHz bandwidth signal was generated and analyzed, providing a time resolution of 0.5 ns, equivalent to a spatial resolution of 15 cm.

For this bandwidth, the numeric bit resolution of the setups analog-to-digital converter is 10 bit, resulting in a 60 dBc dynamic range of the spectrum analyzer. For CFs up to 40 GHz, a filtered Frank-Zadoff-Chu sequence [1] with a length of 120,000 samples was used as a channel sounding waveform, resulting in a channel impulse response (CIR) update rate of 50s. Per antenna constellation 1024 CIR measurements were automatically triggered, and their results were stored, averaged and further analyzed.

For CFs between 60 and 90 GHz, a custom Chirp sequence with a length of 480,000 samples was used; resulting in a CIR update rate of 200s. 256 CIRs were measured for each antenna constellation. The excess loss is defined as: EL(f)[dB] = P_RX - cal(f) - 20log_10(d), where P_RX is the received power, cal(f) are the frequency and setup dependent calibration factors and d is the distance between TX and RX. When during the measurements a clear line of sight signal is observed in a CIR, its relative delay is converted into a distance and used as d.

The root mean square delay spread of the determined power delay profiles (PDPs) is calculated following standard procedures [2]. We only take into account signals larger than our defined noise threshold, 6 dB above the measurement noise floor; based on the average power plus standard deviation of the final 20 % of the PDP-data, in which no multipath components are expected. Each text file named "EL_**p*GHz.txt" contains the determined Excess Loss at a carrier frequency of **.* GHz.

The first column contains the azimuth angle in degrees, and the second and third column contain the Excess Loss and its standard deviation in dBm. Each text file named "RMSDS_**p*GHz.txt" contains the determined Root Mean Square Delay Spread at a carrier frequency of **.* GHz. The first column contains the azimuth angle in degrees, and the second and third column contain the Root Mean Square Delay Spread and its standard deviation in nanoseconds.