Performance Verification of LMR Systems

Report
Using TIA TSB-88.3c Methodologies
2011APCO Western Regional Conference, Ontario, CA
David Buchanan



This series of documents was prepared in
response to requests from APCO, LMCC and
the NCC
These organizations recognized that new
technologies and the drive for more efficient
use of spectrum required better methods for
frequency coordination.
TSB-88 also covers:
◦ Independent Methods of Performance Modeling
◦ Propagation and Noise






Area Coverage
Covered Area Reliability
Delivered Audio Quality (DAQ)
Mean and Median Signal Levels and other ugly
statistical terms
Outdoor Coverage testing
Developing a test methodology to verify
system performance
Yucaipa, CA
Example of a coverage Boundary








% of the Area?
% of the Area with X% of Confidence
What level of signal?
Faded or not faded?
What % of time
Why do we need all these statistics?
Unfortunately radio propagation is not
constant over time and area
It can only be characterized using statistics

Start with signal level
-94 dBm
-106 dBm
-116 dBm
-123 dBm
Higher value for Interference
limited systems or Portable
coverage
Faded Threshold
Static Threshold
Inferred Noise Floor
Test Tiles
We need to divide our coverage area into tiles to test for the signal level
throughout the coverage area
Estimate of proportions
 Tt = Z2pq/e2
Where:
Tt = number of test tiles
Z = Standard Deviate unit (confidence level
97% = 1.881)
p = Service Area Reliability (95% = .95)
q = 1-p
e = Sampling error allowance (e =2% = .02)
And Tt >= 100

Tt = Z2pq/e2
Tt = 2.2362*.95*(1-.95)/.022 = 594
2.236 = 99% confidence level as recommended
by TSB-88
594>100 so Service area needs to be divided
into 594 equal area tiles
Also for outside tests Tile size ≥ 100λ x 100λ
but less than 2 km x 2 km




We now need to test each tile one time to
determine the signal level
How do you test outdoors?
You need to measure 50 subsamples to produce
a 90% confidence level that the measured value is
within +/- 1 dB of the true value. Also the
subsamples should be measured over 40
wavelengths. Use the median value of these
measurements for each tile.
A tile passes if the measured value is >= to the
CPC value (level from slide 7)


To find out divide the number tiles that pass by
the total number of tiles. Example: 594 total
tiles and 580 passed = .976 or 97.6 %. If your
criteria was 97% area coverage – the system
passes. If not your vendor has some problems.
You should also do subjective voice checks for
analog systems particularly for simulcast systems
as they can show good levels but not be
optimized in the overlap areas. For digital
systems measure the bit error rate to check for
voice quality.





Use an automated system for the testing
Don’t mix mobile and portable testing if the
service areas differ
Not all tiles may be accessible for testing and
should be excluded
Some tiles may be critical to have coverage
and those should be noted to the vendors
before system design
There are methods for indoor coverage
testing and mobile data system testing in the
document

similar documents