Aerospace and Electronic Systems - May 2019 - 73

Guan et al.
The result shows a remarkable power relationship
between HTS affordability and the satellite throughput
(Rtotal). Several observations can be made based on Figure 8:
i. There are clear and substantial economies of scales in
terms of affordability ($/Gbps) to be reaped in designing higher throughput satellites. A significant amount
of the variability in affordability for GEO HTS is
explained by throughout alone (R2 ¼ 0.93). The
following power model captures this scaling effect:
AffordabilityGEO-HTS 

Cost of acquisition
Throughput

(1)

¼ 167:3ðRtotal ÞÀ0:886
ii. The "knee" in the affordability curve occurs around
100 Gbps throughput. As a result, it will become
increasingly more difficult to justify the acquisition
of small- or medium-size GEO HTS below this
throughput threshold.4 In other words, the marginal
cost of throughput of the satellite decreases significantly after 100 Gbps, and a minimal cost advantage
is obtained for acquiring a 200 Gbps instead of, say,
a 250 Gbps satellite.
iii. The three very large GEO HTS shown in Figure 8
(Echostar 19, Viasat 2, and Echostar 24) have
reached or dipped below the threshold of $1 million/
Gbps. This provides them with a significant competitive advantage and brings them close to terrestriallike economics for telecommunication solutions;
iv. Launch cost, which are not reflected in (1) and
Figure 8, will raise the affordability curve upward,
with a more pronounced increased toward the lower
throughputs than the higher ones. Launch cost will
be "amortized" over an increasingly larger throughput, and thus it will have a smaller effect on affordability as throughput increases.
In "Cost Per Bit Per Second Decision Tree: Analyses and
Implications" section, we examine the key technical drivers of HTS affordability.

COST PER BIT PER SECOND DECISION TREE:
ANALYSES AND IMPLICATIONS
Having first discussed high-level strategic considerations
of HTSs and the market disruptions they are likely to
bring, we then examined more tactical matters including
the affordability-throughput map and cost per bit per second of HTS. In this section, we further narrow down our
focus and examine technical considerations that drive the
4

Unless satellite manufacturers can figure out how to significantly compress the cost of low/medium throughput satellites
beyond their current values (by a factor of 2 or 3). This is not
likely to happen in the near future.

MAY 2019

cost per bit per second of these satellites. The integrated
perspective on these technical considerations can help for
example guide the development of a coherent R&D portfolio in support of HTS.
Figure 9 displays our proposed cost per bit per second
decision tree. One important objective for all satellite
operators is to reduce this cost of connectivity. Figure 9
shows different levers for doing so. The cost levers are
self-evident and will not be further discussed here. We
focus instead on the technical lever for increasing satellite
throughput, the upper branches in Figure 9. For an HTS,
the are three broad ways of increasing its throughput, and
they each require a slew of supporting technologies:
1. Widen the usable bandwidth allocated to the satellite;
2. Increase the frequency reuse (with multi-spot beam
coverage);
3. Increase the spectral efficiency.
A brief introduction to each of these levers follows.
The constraints for pulling on these levers are also discussed, and some of the important interconnectedness and
tradeoffs highlighted.

ALLOCATED BANDWIDTH LEVER FOR INCREASING
THROUGHPUT
The various steps in "transforming" the allocated bandwidth ðBw Þ to throughput (Rtotal) are shown in Figure 10,
and they are reflected in 2:


À
Á
Np Nb
Rtotal ¼ b
Bw 1 À hguard
Nc
8
b : spectral efficiency
>
>
>
>
>
>
< Np : number of polarization ð1 or 2Þ
Nb : number of spot beams
>
>
>
> Nc : number of colors ð! 3Þ
>
>
:
hguard : guard band between subbands; typically 5%-10%:

(2)
The different terms in (2) will be examined shortly.
For the time being, we only consider the input and the
output in Figure 10, namely the allocated bandwidth and
the total throughput of the satellite. Notice that the total
throughput increases linearly with the allocated bandwidth
(subject to transmission power constraint discussed later).
As a result, the first lever that can be thought of to increase
throughput is to request an increase in the allocated bandwidth to the satellite. The bandwidth allocated is determined by spectrum availability and by international
regulatory constraints (ITU) for each frequency band and
each geographic region. Allocations are further complicated
by national considerations from country to country, and in
some cases, they require coordination with other terrestrial
users. An examination of spectrum availability can be found
in Evans and Thompson [22] and Bousquet et al. [23].

IEEE A&E SYSTEMS MAGAZINE

73
Aerospace and Electronic Systems - May 2019

Table of Contents for the Digital Edition of Aerospace and Electronic Systems - May 2019
