Chapter 3: Asymmetric Link Budgets

Often, a simple link budget only tells you half of the story. Unless your devices at each end are identical, you need to calculate a link budget for each direction - and use the worst of the two. This section walks you through the process.

Sometimes, you will have different types of radio on each end of a link. A common scenario for this would be a small client-side device (such as a Ubiquiti NanoStation Loco M5), connecting to a sector (such as a Rocket M5 with a ninety-degree sector antenna attached). In these cases, it can be worth generating two link-budgets. This will show you whether the link is likely to be limited by the transmission power of the sector, by the transmission power of the customer-side unit (CPE – Customer Premises Equipment), or both. This can also give you a good idea as to whether using a larger customer-side radio can help for your planned link, or if you need to reconsider your choice for a tower-side radio.

For example, suppose that on the tower-side we have a Rocket M5, attached to a 90 degree sector antenna with a gain of 20 dB. On the client side, we would like to use an inexpensive Ubiquiti NanoStation Loco M5. We hope to provide service to a lot of clients, so using an inexpensive customer-side device can help the budgetary bottom-line significantly. We are hoping to provide service to customers up to 15 kilometers away. Once again, we’ll use the 5.805 ghz center frequency.

We start by calculating the Free Space Path Loss for 15 km:

FSPL (dB) = 20 * log10 (15.0) + 20 * log10 (5805) + 32.44
FSPL (dB) = 20 * 1.176 + 20 * 3.76 + 32.44
FSPL (dB) = 131.16

Since Free Space Path Loss will be the same in both directions, we only have to calculate this once.

A link budget from the Rocket/Sector to the Loco M5 looks like this (using numbers obtained from Ubiquiti’s data-sheet):

Item Value
Transmitter Power 27
Transmitter Antenna Gain 20
Receiver Antenna Gain 13
Free Space Path Loss -131.16
Receive Sensitivity -75 dB
Fade Margin 4 dB (-71 + -75)

While performance won’t be great at the full 15 kilometers, we can reasonably expect the Loco M5 to be able to receive data from the tower at this distance.

A link budget from the Loco M5 to the Rocket/Sector looks like this:

Item Value
Transmitter Power 23
Transmitter Antenna Gain 13
Receiver Antenna Gain 20
Free Space Path Loss -131.16
Receive Sensitivity -75 dB
Fade Margin 0 dB (-75 + -75)

It appears that the Loco M5 will just be able to transmit loudly enough for the Rocket/Sector to receive its signal, but there is absolutely no fade margin whatsoever! Even the slightest disruption at 15 km will prevent the tiny Loco M5 from sending data to the tower. Therefore, if reliable service is required a larger customer-side radio is required (or we can reduce our distance estimate from 15 kilometers). Ideally, we should calculate this link budget for other distances and determine that we can use the inexpensive Loco M5 for customers at up to 10km, and require a larger NanoStation M5 for customers at up to 15 kilometers. We can even consider NanoBridge/NanoBeam/PowerBeam products to extend the range even further.

The good news is that by performing this calculation before we showed up at a customer and promised them great service with a really inexpensive client-side device – we calculated that there would be problems, and avoided an expensive and frustrating visit to the customer’s roof.

There are two major points to glean from this example:

  1. In most cases such as this, the client-side CPE is the limiting factor – so when you are calculating future link-budgets to see if a potential customer can be given service, you should start with the customer-side. You may discover that you need a bigger client-side device!
  2. This example highlights a much-recited fact amongst WISP operators: antenna gain is far more important than transmission power. When calculating a link-budget from both directions, you get to count the gain of each antenna in both directions – but you only get to utilize a high-power transmitter in one direction. You effectively get to count the antenna twice!
« Chapter 3: Link Budgets Up To Contents Chapter 3: Modulation »