wiki:802.11/Benchmarks/Rx_Char

Version 6 (modified by chunter, 9 years ago) (diff)

--

Receiver Characterization

The IEEE 802.11 Standard specifies a minimum receiver sensitivity that all 802.11 devices must achieve. This sensitivity is specified as a receive power (in dBm) in which a device must achieve no worse than a 10% packet-error-rate (PER) for PPDU receptions of 1000 octets. The receiver minimum input sensitivity requirements are specified in Section 18.3.10.2 of 802.11-2012 and are reproduced here:

Rate Modulation / Coding Minimum Sensitivity (dBm)
6 Mbps BPSK 1/2 -82
9 Mbps BPSK 3/4 -81
12 Mbps QPSK 1/2 -79
18 Mbps QPSK 3/4 -77
24 Mbps 16-QAM 1/2 -74
36 Mbps 16-QAM 3/4 -70
48 Mbps 64-QAM 1/2 -66
54 Mbps 64-QAM 2/3 -65

The values specified in this table recognize that higher rate transmissions require higher SNR (and therefore higher delivered power) than lower rate transmissions.

Receiver Sensitivity Test

To test the Mango 802.11 Reference Design Rx PHY against these measures, we use the Keysight N4010A Wireless Connectivity Test Set to generate known-good 1000 octet transmissions and deliver those transmissions to WARP v3 at a specified power. Then, we use the WLAN Experiments Framework to measure PER.

No image "experimental_setup.jpg" attached to 802.11/Benchmarks/Rx_Char
Experimental Setup

Methodology

  1. Construct PPDU waveforms in MATLAB for each PHY rate that have the following characteristics:
    1. 1000 octet PPDU length (i.e., a 972 byte payload + 14 byte MAC header + 4 byte FCS)
    2. 10 different scrambling sequence start points to generate waveforms with different peak-to-average-power (PAPR) features
  2. Load each waveform in to the N4010A's volatile memory as segment files.
  3. Create N4010A sequence files that send each waveform segment 1000x in a row
  4. Use WLAN Experiments Statistics to determine how many of the transmissions were fully received with a good FCS.

We have used the above steps determine the Mango 802.11 Reference Design's receive sensitivity across 7 different WARP v3 kits and RF interfaces.

No image "rx_sensitivity" attached to 802.11/Benchmarks/Rx_Char

The above plots shows the results of this experiment. In particular, notice the receive powers that cross the 10% PER horizontal line -- these can be compared directly to the minimum sensitivity values specified by the standard. Specifically,

Rate Modulation / Coding Minimum Sensitivity (dBm) Mango 802.11 Sensitivity (dBm) Difference
6 Mbps BPSK 1/2 -82 -85 3dB better than minimum
9 Mbps BPSK 3/4 -81 -83 2dB better than minimum
12 Mbps QPSK 1/2 -79 -83 4dB better than minimum
18 Mbps QPSK 3/4 -77 -81 4dB better than minimum
24 Mbps 16-QAM 1/2 -74 -78 4dB better than minimum
36 Mbps 16-QAM 3/4 -70 -75 5dB better than minimum
48 Mbps 64-QAM 1/2 -66 -71 5dB better than minimum
54 Mbps 64-QAM 2/3 -65 -69 4dB better than minimum

Attachments (5)

Download all attachments as: .zip