# Acceptable link margin for voice communication in uhf band

I want to understand the theory behind how link margin or fade margin is decided? If I have to have a point to point UHF voice communication link from ground to aircraft, for example, with a reliability of 95%, how is the link margin decided? What is acceptable link margin for these links? Are there any standards for these?

• I don't have time to look it up right now, but for example in Theodore S. Rappaport's Wireless Communications: Principles and Practice second edition there were derivations for cell coverage with different reliability levels. So your point-to-point system would act as a one-cell cluster. I think that you might be able to directly use derivations, but I don't remember the specifics right now and I don't have time to look it up these days. Don't forget that aircraft's altitude will also figure in the calculations! Sep 21, 2015 at 17:10
• – user
Oct 25, 2015 at 14:05

It all starts by determining the link budget. According to your model and your system characteristics (i.e., bandwidth, frequency band, bit rate, antennas, etc.), the link may have different kinds of behaviors. For instance, the motion of the aircraft will produce Doppler effect, which might be important for high frequencies (e.g. S-band). Anyway, a Doppler effect combined with specular reflections of the radio wave on the communication environment results in a time varying communication channel, commonly described by its coherence time. Time varying channels are also referred to as fading channels. On the other hand, you may also encounter atmospheric effects, which come from the interaction of the radio wave with the particles on the atmosphere (so it depends on the frequency and the altitude of the spacecraft). So again, everything depends on your model. Sometimes these effects can be neglected.

Once you can describe the behavior of the channel, then you can define a worse case scenario and determine the required link margin in order to cope with the fading events. The fading statistics of typical channels are known, so you can compute the probability of having a fading event larger than X dB. Having that, you can design your system to achieve a certain availability objective.

The ITU provides recommendations for the design of radio links. You may take a look at the ITU-R P.618-7 to get an idea (it is for an Earth-Space communication system), but possibly you may find something closer to your application.

If you have to estimate the loss that your signal will face whilst propagating to your receiver you perform a link budget for your design. Ignoring antenna gains/losses we generally have,

$$P_{TX} = P_{RX} + L_{TOTAL} + L_{MARGIN}$$

where $$P_{TX}$$ is the transmit power, $$P_{RX}$$ is the minimum receive power you need, $$L_{TOTAL}$$ is an estimate of the total link loss and $$L_{MARGIN}$$ (link margin) is the difference you want between the the receive sensitivity and the actual receive power. $$L_{MARGIN}$$ is the additional losses your scheme can withstand over and above standard losses.

and $$L_{TOTAL} = L_{PATH} + L_{SLOW} + L_{FAST}$$

where $$L_{PATH}$$ is the path loss, $$L_{SLOW}$$ is the slow fading loss and $$L_{FAST}$$ is the fast fading loss.

The term "fading margin" is not consistently used, it can be used to either mean i) the link margin ($$L_{MARGIN}$$) or ii) the total estimated loss due to fading ($$L_{TOTAL}$$).$$L_{MARGIN}$$ is chosen at the designers discretion, if $$L_{MARGIN}$$ is 0 then the average BER will be $$10^{-9}$$, a higher value of $$L_{MARGIN}$$ will lead to a lower (and hence better) BER.

$$L_{TOTAL}$$ is computed by individually computing and summing up $$L_{PATH}$$,$$L_{SLOW}$$ and $$L_{FAST}$$.If you do not know how these individual components are computed then you might want to ask separate questions about how to calculate each type of loss.