|
|
How to use this web site:
What's here:
At this web site, you can listen to audio samples of simulated voice radio traffic received over real terrain along the Pacific northwest rail corridor. Some of the samples have acoustic locomotive cab noises in the background (for the locomotive-to-base, or talk-in, transmissions). These audio recordings were transmitted over land-mobile radio between a locomotive and base station in analog FM and digital Project 25 (TIA-102) modulation formats, and the received audio was digitally recorded at the distant end. The recordings are sorted by GPS-derived latitude/longitude coordinates, and can be heard by clicking on specific locations on the map which follows. This enables you to compare the relative intelligibility of speech as processed by the two different modulation formats, and as would be heard from the same, or nearly the same, geographic location. This was just one part of our investigation; the other topics were: (1) implementing a railroad trunked radio system in Portland, Oregon, and (2) comparing the data performance of Project 25 (TIA-102) radios to the railroad standard Spec-200 data radios. All of these topics will be fully discussed in a forthcoming NTIA Report entitled "Data, trunking, and voice performance of TIA-102 land-mobile radios in a railroad environment."
What we did:
We created digital audio files (in ".wav" format) using several groups of Harvard phonetically-balanced sentences. These sentences are phonetically balanced in the sense that the occurrence of English-language phonemes is more or less uniformly distributed.
We made an additional set of these sentences, except that the second set had acoustic locomotive cab noises in the background. The background noise had been previously recorded with a high-quality microphone and digital audio tape recorder in actual freight locomotives under various conditions of engine throttle settings, airhorn blasts, and operator alerts, and for two styles of freight locomotives. Absolute sound pressure levels were measured for each recording so the proper pressure level could be replicated in the laboratory when making the source audio speech recordings. The purpose of having acoustic noise in the background of the source speech files was two-fold: (1) to simulate a locomotive cab environment for talk-in transmissions, and (2) to facilitate the natural tendency of the talkers to speak louder into the microphone when contending with loud extraneous background noise. Both facets of this simulated environment provided for more realistic inbound voice reception.
The microphone that was used in both cases (background noise and no background noise) was that which was furnished with the locomotive radio. Using this microphone ensured that any nuances peculiar to the microphone element itself would be accounted for when applying the ".wav" files to the radios' microphone inputs. The files with no background noise simulated the outbound dispatcher (base station) element and the files with the added acoustic noise simulated the inbound locomotive element.
Two computer programs (one at the Capitol Peak base station site and one on the locomotive) running in tandem and synchronized by continuously updated GPS-derived time information, alternately transmitted and received (and digitally recorded on computer) the audio files sent over the air, first in the talk-in direction and then in the talk-out direction. Date and time of transmission, name of ".wav" audio file sent, date and time of reception, name of ".wav" audio file received (recorded), and in the case of the locomotive segment, latitude, longitude, and velocity information, were logged to data files on each computer. This enabled us to associate the recorded audio files, at both locomotive and base station, to the train's location and to the appropriate source audio files on the sending computer.
The actual radio equipment was outfitted onto a passenger locomotive which had a fixed and repetitive route/schedule between Seattle, Washington, and Portland, Oregon. Over the course of several trips, the audio transmissions were collected in both analog FM and digital Project 25 modulation formats, and included simulating the different styles of freight locomotives (by virtue of the different background noise sets on the talk-in ".wav" files). The method of RF modulation and noise set used was the same for an entire trip so that each trip corresponded to a particular condition of modulation (analog FM or Project 25 digital) and style of freight locomotive.
What you have to do:
As you glide your computer mouse over the map image that follows, and illuminate those segments of railroad track (magenta-colored lines on the map) where radio transmission recordings were made (along a route between Seattle and Portland), you will get a visual indication of a "hyperlink" on your display monitor. By clicking on that hyperlink, you will view information unique to that location, such as latitude, longitude, terrain profile between base station and locomotive, whether the associated audio recordings were in the locomotive-to-base station (talk-in) direction or in the base station-to-locomotive (talk-out) direction, the modulation format used (analog FM or Project 25 digital), and so forth. There are also hyperlinks to listen to the original source audio that was transmitted and the audio that was heard by the receiver, for each type of modulation and each type of simulated freight locomotive. There are also hyperlinks to information about the locomotive RF electronics and the Capitol Peak base station site.
IMPORTANT! -- open the hyperlink on the next page that says "For a description of how to use this site, click here" and READ IT!
Note that the analog audio source file will not necessarily be the same source file as was used for the digital modulation case, because the analog and digital samples were gathered on different trips, and variations in speed and lengths of time spent boarding and deboarding passengers along the way meant that the train would cross a given latitude/longitude at slightly different times (relative to the train's initial departure time) on any given trip.
On the map, you can also view information about the base station site by clicking on the "Capitol Peak" hyperlink (at the magenta-colored point to the lower left of the "Capitol Peak" label). There, you will be able to view information about various parameters of that site, such as transmitter power, transmit and receive antenna heights, etc.
If your mouse is positioned anywhere else on the map (not on railroad tracks, or on tracks where we weren't), nothing will happen when you click your mouse.
Click here to continue on to the map.
For more information, or to obtain a clone of this web site on CD-ROM, contact:
John M. Vanderau, P.E.
Radio Systems Engineer
TEL: (303) 497-3506, FAX: (303) 497-5969
Email:
Institute for Telecommunication Sciences
NTIA/ITS.P, 325 Broadway, Boulder, CO, USA 80305-3328