|Thread Links||Date Links|
|Thread Prev||Thread Next||Thread Index||Date Prev||Date Next||Date Index|
Piers Dawe submitted code that would automate measurement of OMA in TWDP. A subcommittee consisting of Jim McVey, Petar Pepeljugoski, Norm Swenson, Tom Lindsay, Jan Peeters Weem, and Sudeep Bhoja was charged with reviewing the code and deciding whether it should be included in D2.2. The subcommittee was chaired by Tom Lindsay. At the discretion of the chair, Piers Dawe was added to the subcommittee.
The subcommittee met three
The subcommittee evaluated the proposed method against two independently developed methods, using the waveforms on the LRM website and additional waveforms with known OMAs that were simulated and experimentally measured during the study. All three methods work over the variety of waveforms. However, the peak-peak spread in OMA values among the methods was 0.7 dB across the waveforms. The average value of the spread was 0.27 dB with a standard deviation of 0.175 dB. No method consistently was highest or lowest.
Greater than 25% of the subcommittee was concerned that the variation was too large and that further study, testing, and development are required before making changes to the document.
· There was a range of views on how to best implement OMA-automation into the document. Since OMA-automation is independent of the core of TWDP, choices considered include making the code informative or normative, and whether to include it within the body of TWDP or as a separate routine, or possibly in an annex.
As presented in San Francisco, other
improvements in addition to power automation are being considered for
TWDP. These include alternative power normalization, finite EQ length, and
adjustments to the pass/fail limits. These potential improvements are all
inter-related. It may be premature to make changes until all work is wrung out
and coordinated. Per the TP2 summary presentation and discussion in
A summary presentation of this
work is planned. To progress OMA
automation, the subcommittee will make the three OMA extraction methods
available on the web site for folks to test and evaluate. Members of P802.3aq should continue
to investigate improvements to TWDP. This must be done in a