Percent Time and Location vs. Reliability

In the EDX software within the propagation dialogue the two statistical mechanisms known as “Time/Location Variability” and “Reliability” are completely independent of one another.

“Time/Location Variability” are propagation parameters that adjust the path loss calculation and are used to gain confidence that you have an accurate mean path loss prediction.

Whereas, “reliability” is determine by a signal level analysis (a study) and is used to understand receiver performance in the presence of a received signal that is variable. Also, even though these two approaches to signal fading are statistical in nature they don’t necessarily use the same methods. The Time/Location calculations are from Longley-Rice, the Reliability study is a regular Log Normal distribution.

Because each approach attempts to model signal fading at different points in the overall signal analysis within the EDX program it is important to note that there is a risk of doubling up on signal margin adjustments and ending up with overly conservative results unless each approach is independently considered.

The Time/Location/Margin entries in the propagation model are simply means of gaining more confidence in the propagation model. One starts with the assumption that the model along with the use of properly tuned clutter data and good terrain data will give you an accurate mean signal level. However, there are also other signal level variants involved such as location or time (due to climate) and the Time/Location % controls are a way of accounting for these in order to give you a more accurate mean path loss. These are simply additional ways to tune the model and increase the confidence in its results.

For example, if you have a mobile receiver and you want a received signal prediction that represents mean signal levels that are achievable for 90% of the locations in the study area you could use 90% Location variability. Doing so adds about 10dB additional path loss which makes the predicted signal levels more conservative. This dB margin is calculated using the “AVAR” routine found the Longley-Rice propagation model computer code and is characterized by a random variable with a distribution similar to a Rayleigh distribution. Originally this routine was created from the information found in the NTIS “Tech Note 101” (January, 1967) document. Put another way, location variability is localized signal level variations that account for changes in the propagation environment which are not explicitly considered in the propagation model.

Alternately, 90% Time variability means a margin is added to the path loss so we are confident that the actual signal level will not fall below the predicted value for at least 90% of the time. Because this variability is due to climate the distance from the transmitter is a factor. Locations at distances less than 30km effectively have little or no Time margin adjustment.

50% Time or Location variability is 0dB margin. In other words, for Location variability, 50% of the locations would have actual signal levels equal to or greater than the predicted mean signal level and 50% of them would be lower than that predicted by the propagation model.

The “Margin” entry in the propagation model parameters is simply a straight dB adjustment. Positive values show up as an increase in path loss. Further details regarding Variability can be found in Appendix A of the Reference Manual.

Now, when doing a calculation for reliability we are asking what is the likelihood that there will be a usable signal level at the receiving equipment. The greater the signal level is above a receiver’s operating threshold, the more “reliable” is the receiver because as the signal fades due to time or location it is less likely to fall below the receiver threshold. Reliability studies allow you to choose the type of statistical distribution that you feel best describe the fading characteristics of the received signal.

You can see that these two statistical mechanisms are independent of one another. “Time/Location Variability” is used to gain confidence that you have an accurate mean path loss prediction and “reliability” is used to understand receiver performance in the presence of received signal that is variable.

A default EDX recommends the Time and Location be set to 50% when doing reliability studies. If you need additional path loss margin to account for propagation loss mechanisms for which the propagation model doesn’t accommodate (unknown clutter effects, wall penetration loss, multipath, etc.), add those in the “dB Margin” entry. Ultimately the best way to determine this value (as well as the dB loss values of your clutter database clutter categories) is to perform drive test measurements.