Animating Channel Condition Number Along a UE Trajectory
This notebook demonstrates how to use NeoRadium to evaluate and animate the condition number of a trajectory-based channel as a UE moves along a trajectory in a DeepMIMO scenario.
The notebook first opens a DeepMIMO scenario, creates a random UE trajectory, and then creates a TrjChannel object from that trajectory. As the UE moves along the trajectory, the notebook computes the channel condition number from the channel matrix and updates an animation.
The animation shows the UE moving along the trajectory on the DeepMIMO map. Below the map, two graphs are updated as the UE moves:
Condition number: shows the average condition number over each animation update interval, with a shaded region indicating the minimum-to-maximum range.
Number of paths: shows the average number of propagation paths over each animation update interval.
Notes
The condition number is computed from the channel matrix generated by the trajectory-based channel model.
The condition-number curve summarizes the channel over each animation update interval rather than showing every individual trajectory point.
The
trajLenparameter refers to the number of DeepMIMO grid points used for the trajectory, not the final number of interpolated trajectory points.The user may need to update the
dataFoldervariable based on the location of the DeepMIMO scenario files on their system.The generated animation is saved as a GIF file and then displayed in the notebook.
[1]:
import numpy as np
import matplotlib
from IPython.display import HTML, Markdown, display
from neoradium import DeepMimoData, TrjChannel, BandwidthPart, AntennaPanel, random
from neoradium.utils import toDb
[2]:
# Replace this with the folder on your system where the DeepMIMO scenarios are stored
dataFolder = "/data/RayTracing/DeepMIMO/Scenarios/V4/"
DeepMimoData.setScenariosPath(dataFolder)
# Create a DeepMimoData object
deepMimoData = DeepMimoData("asu_campus_3p5")
deepMimoData.print()
DeepMimoData Properties:
Scenario: asu_campus_3p5
Version: 4.0.0a3
UE Grid: rx_grid
Grid Size: 411 x 321
Base Station: BS (at [166. 104. 22.])
Total Grid Points: 131,931
UE Spacing: [1. 1.]
UE bounds (xyMin, xyMax) [-225.55 -160.17], [184.45 159.83]
UE Height: 1.50
Carrier Frequency: 3.5 GHz
Num. paths (Min, Avg, Max): 0, 6.21, 10
Num. total blockage: 46,774
LOS percentage: 19.71%
[3]:
random.setSeed(123) # Make results reproducible
bwp = BandwidthPart(numRbs=24, spacing=15) # Create a BandwidthPart object
# Create a random trajectory
trajectory = deepMimoData.getRandomTrajectory(xyBounds=np.array([[-210, 40], [-120, 100]]), # Trajectory bounds
segLen=5, # Number of DeepMIMO grid points on the shortest segment
bwp=bwp, # The bandwidth part
trajLen=100, # Number of DeepMIMO grid points on the trajectory
speedMps=15) # Speed in m/s
trajectory.print() # Print the trajectory information
ax = deepMimoData.drawMap("LOS-NLOS", trajectory) # Draw the map with the trajectory
deepMimoData.drawBsPanel(ax, 180) # TX antenna with a 180-degree bearing angle
Trajectory Properties:
start (x,y,z): (-164.55, 69.83, 1.50)
No. of points: 7,984
curIdx: 0 (0.00%)
curSpeed: [10.64 10.64 0. ]
Total distance: 119.71 meters
Total time: 7.983 seconds
Average Speed: 14.996 m/s
Carrier Frequency: 3.5 GHz
Paths (Min, Avg, Max): 6, 8.90, 10
Totally blocked: 0
LOS percentage: 48.58%
[4]:
channel = TrjChannel(bwp, trajectory,
txOrientation = [180,0,0], # BS antenna facing the left side of the map
txAntenna = AntennaPanel([2,4]), # 8 TX antennas
rxAntenna = AntennaPanel([1,2]), # 2 RX antennas
seed = 123)
print(channel)
def getConditionNumber(channelMatrix):
# Condition number:
# CN(H) = 20 log10(sigmaMax/sigmaMin)
# NOTE: CN is given on a logarithmic scale in dB. Values between 0 and
# 10 dB are usually considered good for beamforming, while values above 20 dB
# are considered unusable for beamforming.
u, s, v = np.linalg.svd(channelMatrix)
cn = 2 * toDb(s.max(axis=2) / np.maximum(s.min(axis=2), np.finfo(float).eps))
return cn.min(), cn.mean(), cn.max()
# A callback function used to draw condition-number and path-count graphs below the animated trajectory
def handleGraph(request, ax, trajectory, points=None):
if request=="Config":
# Configure both graphs
ax[0].set_xlim(0,trajectory.numPoints)
ax[0].set_title("Condition Number (dB)")
ax[0].grid()
if len(ax) > 1: # Set `numGraphs` to 2 to draw the number of paths in the second graph
ax[1].set_xlim(0,trajectory.numPoints)
ax[1].set_title("Number of Paths")
ax[1].grid()
elif request=="ConfigMap":
# Customize the drawn map
deepMimoData.drawBsPanel(ax, 180) # TX antenna with a 180-degree bearing angle
elif request=="Draw":
# Per-frame updates to the graphs below the map
# ax is an array with `numGraphs` elements
p0, p1 = points
cnMin, cnMean, cnMax, numPaths = [], [], [], []
while channel.trajectory.curIdx < p1:
cn = getConditionNumber( channel.getChannelMatrix() )
cnMin += [cn[0]]
cnMean += [cn[1]]
cnMax += [cn[2]]
numPaths += [channel.numPaths]
channel.goNext()
cnMin, cnMean, cnMax, numPaths = min(cnMin), np.mean(cnMean), max(cnMax), np.mean(numPaths)
global prevInfo
prevMin, prevMean, prevMax, prevNumPaths = prevInfo
ax[0].plot([p0,p1], [prevMean, cnMean], 'red', markersize=1)
ax[0].fill_between([p0,p1], [prevMin, cnMin], [prevMax, cnMax], color='pink', alpha=0.5)
if len(ax) > 1: # Set `numGraphs` to 2 to draw the number of paths in the second graph
ax[1].plot([p0,p1], [prevNumPaths, numPaths], 'blue', markersize=1)
prevInfo = [cnMin, cnMean, cnMax, numPaths]
print(f"\rProcessed trajectory point: {p1}", end="")
TrjChannel Properties:
carrierFreq: 3.5 GHz
normalizeGains: True
normalizeOutput: True
normalizeDelays: True
xPolPower: 10.00 (dB)
filterLen: 16 samples
delayQuantSize: 64
stopBandAtten: 80 dB
dopplerShift: 175.6 Hz
coherenceTime: 2.409 milliseconds
TX Antenna:
Total Elements: 8
spacing: 0.5𝜆, 0.5𝜆
shape: 2 rows x 4 columns
polarization: |
Orientation (𝛼,𝛃,𝛄): 180° 0° 0°
RX Antenna:
Total Elements: 2
spacing: 0.5𝜆, 0.5𝜆
shape: 1 rows x 2 columns
polarization: |
Trajectory:
start (x,y,z): (-164.55, 69.83, 1.50)
No. of points: 7,984
curIdx: 0 (0.00%)
curSpeed: [10.64 10.64 0. ]
Total distance: 119.71 meters
Total time: 7.983 seconds
Average Speed: 14.996 m/s
Carrier Frequency: 3.5 GHz
Paths (Min, Avg, Max): 6, 8.90, 10
Totally blocked: 0
LOS percentage: 48.58%
[5]:
# Create the animation, save it as a GIF, and display it below. This step may take some time.
channel.restart()
prevInfo = [0, 0, 0, 0]
anim = deepMimoData.animateTrajectory(trajectory, numGraphs=1, pointsPerFrame=20,
graphCallback=handleGraph, fileName='AnimateCN.gif')
display(Markdown(""))
# Another option is to use the following command which gives you more controls for running
# the animation.
# # Increase the animation memory limit for HTML-based animation display
# matplotlib.rcParams['animation.embed_limit'] = 100000000
# HTML(anim.to_jshtml())
Processed trajectory point: 7960
[ ]: