Wireless link reliability

[alqasim]

In any specs datasheet for a microwave link , the manufacturer used to mention the reliability value for the link ( e.g. Highly reliable – typically 99.995% availability )?

what does it mean? and depends on what factors ?

 

[tony_lth]

Because the wireless channel model is a propability function on time and amplitude etc.

 

[pstuckey]

Each link will have different factors that affect reliability. A stationary Ka band satellite link will be primarily affected by weather so the reliability estimate will use statistics on local rainfall. A cell phone tower will only be slightly affected by rainfall, but will be strongly affected by the users mobility, time varying mulitpath, loss of LOS, etc.

 

[alqasim]

Does it depends on the technology used itself , sampling rate ...etc. ?

 

[pstuckey]

Does it depends on the technology used itself , sampling rate ...etc. ?

Yes. Everything in the analog and digital Rx/Tx path can affect the link reliability.

Maybe you could be a little more specific in your question.

 

[alqasim]

Actually my question is :

1) Under which condition this percentage ( the reliability percentage ) is given ?
2) How could i calculate the reliability of the link under different environment ?

Best regards

 

[pstuckey]

Each radio link will be different, but in general you will design the link so that it works with everything that you know absolutely, then put some margin in to handle the unknowns.

For instance, a Ka band satellite link will be strongly affected by rainfall. So you will need to get the local statistics for rainfall. This would be a table of the form rainfall exceeds xx mm/hr for yy% of the year. For each rain rate, you will calculate the amount of attenuation (dB). This would tell you how much rain fade margin you will need to meet your availability requirement. If you move this link to an area with more rainfall, then you would need to either increase the margin (larger antenna, larger PA, etc.), or advertise a lower availability number.

 

[tony_lth]

I calculated the satellite telecommunications channel model 17 years ago. First determine channle model, say Rice or Rayleigh, then found some old papers that had equations for that. Some IEEE papers is very helpful.