Low Earth orbit (LEO) satellite networks are designed to provide broadband access in the most remote, hard to reach locations. While they work well for this intended application, they are less successful at delivering fixed wireless access (FWA) in more mainstream markets; in this case, anywhere with a household density greater than 1 household per square mile (1.6 square kilometers). This is contrasted with the 340 households per square mile supported by Tarana’s next-generation fixed wireless access (ngFWA) platform, Gigabit 1 (G1).
We’ll cover some of the reasons for the wide discrepancy and the differences between satellite and ngFWA here. For a more detailed discussion of how we derived these values, check out the full white paper here.
Critical Service Metrics
We get our density number of served households with the following metrics:
- Cell coverage area
- Link capacity
- System (cell) capacity
The following table compares service metrics between a leading LEO satellite broadband provider and G1. We use Starlink as our example given it has the largest and most mature constellation available at this time.
|
Metric |
Satellite |
Tarana G1 |
|
Cell coverage area |
143 miles² (370 km²) |
3 miles² (8 km²) |
|
Link capacity (theoretical) |
1,322 Mbps downlink |
800 Mbps (aggregate) with 1.6 Gbps in an upcoming release |
|
System capacity (theoretical) |
21.2 Gbps per cell |
12.8 Gbps per cell |
The important thing to understand about these numbers is that the link and system capacity for both networks is shown above as theoretical. In reality, based on reported performance, a given satellite will have a realistic maximum system capacity of 10 Gbps and a peak beam (link) capacity of 625 Mbps (16 beams per satellite). In the case of G1, we assume 70% of maximum capacity based on reported real world performance from our customers. For a more detailed discussion of how we derived these values, check out the full white paper here.
From these numbers we can derive the subscriber density using peak hour traffic demand. A typical developed country household consumes 4 Mbps of downlink traffic at peak hour. Peak hour traffic has been rising every year with a projected peak rate of 10 Mbps in the next 5 years. This yields the following:
|
Peak Hour Rate |
Subscribers per Cell (Satellite, 1 beam) |
Subscribers per Cell (G1, 4 sectors) |
|
4 Mbps |
625 Mbps/4 Mbps = 156 subscribers per cell |
9 Gbps/4 Mbps = 1,024 subscribers per cell |
|
10 Mbps |
625 Mbps/10 Mbps = 62 subscribers per cell |
9 Gbps/10 Mbps = 900 subscribers per cell |
Now that we have the number of subscribers per cell, we can determine the supported subscriber density.
|
Peak Hour Rate |
Subscribers Density (Satellite) |
Subscribers Density (G1) |
|
4 Mbps |
1.1 HH per mile² (0.4 HH/km²) |
340 HH per mile² (130 HH/km²) |
|
10 Mbps |
0.4 HH/mile² (0.16 HH/km²) |
300 HH/mile² (115 HH/km²) |
This calculation attempts to use realistic performance numbers based on factors such as interference, weather, distribution of user terminals/CPEs) etc. If we want to run the same exercise using theoretical capacity, the subscriber density for a satellite rises to slightly over 2 subscribers per square mile. This still falls far short of a typical rural population density of 30-50 households per square mile.
As we mentioned at the beginning, LEO satellite is an excellent solution for broadband in the most inaccessible places on Earth. However, when any type of density is applied, satellite broadband fails to satisfy the requirements necessary for a successful fixed wireless deployment. Only a next-generation FWA platform such as G1, which was designed from the ground up specifically for FWA applications, delivers speed and capacity at any subscriber density.
For a more detailed discussion of G1 vs. LEO satellites, check out the full white paper here.
If you just can’t wait to learn more, check out our other blogs or some of our favorite customer links. Or reach out to us at info@taranawireless.com. We’d love to hear from you.