►
From YouTube: 20211002 Technical Steering Committee
Description
1) Ensure sufficient architectural clarity for daily development work
2) Ensure the technical roadmap is clearly defined into sub-projects and communicated with the wider community. In particular, we will aim for concise definitions of development steps that interested members (or prospective members) can get involved in.
Links mentioned today:
https://github.com/phase4ground/documents/tree/master/Engineering/Requirements/Architecture
https://github.com/phase4ground/documents/tree/master/Remote_Labs
A
Okay,
then,
I
guess
we'll
just
start
here,
so
I
made
some
slides,
there's
no
kind
of
divine
truth
in
them.
I
just
thought
it
would
be
good
to
have
something
to
serve
as
a
kind
of
backbone
to
any
conversations
so
yeah.
It's
kind
of
my
my
opinion,
my
views,
but
hopefully
it
can
inspire
some
discussion.
A
A
As
I
see
it
is
that
we
want
to
make
sure
that
there's
basically
clarity
of
the
architecture
so
that
when
people
are
working
on
things
day
to
day,
they
don't
have
to
get
confused
and
get
lost
and
what
it
is
they're
trying
to
achieve
and
then
also,
I
think
it's
important
and
there's
been
a
lot
of
discussion
with
juno
and
other
people
that
we
need
to
make
sure
that
we
basically
communicate
what
we're
trying
to
do
to
the
wider
community
and
provide
options.
So
people
can
join
the
project
easily.
A
Yes,
it's
clear
what
we're
trying
to
do
and
where
we
need
help
and
then
a
kind
of
a
overarching
objective
is
this
of
this
is
to
yeah,
try
and
maintain
focus
of
the
goal
of
getting
our
system
in
orbit.
So
yeah,
that's
basically
kind
of
objective.
As
I
see
it
so
yeah.
I
think
they
kind
of
it's
important
to
look
at
the
system
partitioning.
A
So
there's
the
p4x
dmt
payload,
which
is
the
actual
transceiver
that
will
receive
data
from
from
the
ground
and
then
yeah
marks
all
the
different
channels
together
and
transmit
that
back
down
into
dvdx
yeah
transmission,
so
yeah,
we
can
see
here
so
I've
simplified
some
of
the
at
least
for
this
discussion.
Here.
I've
simplified
some
of
the
things
highlighted
in
the
original
dmt
definition
document,
just
because
I
kind
of
feel
that
simplicity
might
be
quite
advantageous
here.
A
So
of
course
everything
is
a
certain
amount
of
configurability,
but
I
think
it's
a
kind
of
a
high
level
system
dimensioning.
I
think
that
that
kind
of
reflects
my
understanding
of
what
we're
trying
to
do
so
then,
to
do
that.
We
also
need
to
put
this
thing
inside
a
satellite
that
can
basically
maintain
attitude,
provide
power,
yeah,
basically
a
host
for
the
actual
payload
itself.
A
A
So
yeah
large
gateway
can
put
gateway
component,
which
is
a
term
that
seems
to
be
increasing
popularity
for
vhdl
and
verilog,
but
yeah
there's
also
non-travel
firmware
and
hardware
that
really
allows
this
thing
to
operate
as
a
whole,
so
yeah.
I
think
we
have
pretty
good
progress
on
these
two
elements:
the
gse,
the
dbb
s2x.
A
So
the
work
of
of
andre
and
onshore
have
really
pushed
this
quite
far.
Actually
at
the
moment,
and
then
we've
got
the
other
aspects
of
the
m17
receiver.
We
need
to
look
at
and
then
obviously
all
the
kind
of
support
and
circuitry
boards
and
all
that
kind
of
stuff,
so
yeah.
This
is
obviously
a
very
high
level
diagram.
I
think
it
would
be
good
to
discuss
what's
missing
from
a
diagram
where
we
need
to
have
some
clarity
on
and
stuff.
So
I
guess
this
had
different
layers
to
it.
A
B
I
can
speak
a
little
bit
about
the
the
number
of
uplink
channels
it
doesn't
have
to
be
so
many
there's
been
talk
about
even
more
having
more
than
a
thousand
and
having
a
relatively
narrow
band
signal
uplink
for
each
operator,
but
it
might
be
worth
thinking
about.
You
know
down
to
like
like
90
90,
something
or
or
you
know,
half
of
5
12..
So
so
that's
also
something
that
we've
we've
talked
about
in
case.
B
B
Now
that
ignores
some
of
the
filtering
requirements
and
stuff
like
that,
and
it
also
flirts
with
the
well.
If
you
have
a
thousand
channels,
can
your
onboard
processing
deal
with
with
all
of
that?
So
so
that's
an
active
area
of
discussion.
I
don't
see
any
problem
with
512
or
this
or
a
smaller
number
being
used
as
a
baseline.
A
B
The
second
thing
would
be
that
there
is
a
couple
of
block
diagrams,
including
a
data
flow
that
we
did
for
airax
project
and
that
we
kept
revising
and-
and
that
may
be
informative
here,
because
it
includes
some
failure.
You
know,
failure,
analysis
or
failure.
Defense-
and
you
know-
has
has
some
of
the
circuitry
that
would
need
be
needed
for
that
some
of
the
rf
switching
in
in
there.
So
the
you
know.
B
B
A
Okay,
yep
into
the
gsc
encoder
yeah
yeah.
I
mean
that's
one
of
the
things
that
I
reviewed
the
the
previous
paper
and
there's
a
lot
of
options
on
the
uplink.
I
think
it
would
be
beneficial
to
basically
choose
one
thing
and
develop
for
that.
One
thing
with
the
view
that
we
can
extend
the
options
in
the
future,
but.
C
C
So
we
did
have
a
discussion
with
m17
group,
but
I
followed
the
path
of
ip
and
I
need
to
take
up
this
channel
back
again
with
m17.
What
all
do
I
need,
but
to
give
them
the
concrete
answer?
I
want
my
simple
solution
with
ip
two
gsc
work
and
then
I
will
be
in
a
better
position
to
go
back
to
m17
and
discuss
what
we
need
there.
A
A
A
D
A
B
This
one,
though,
is,
is
really
good.
It's
just
this
is
better
than
than
developing
yet
another
digital
protocol.
So
yes,
something
like
m17
up
m17
would
be
the
native
one.
But
honestly
any
traffic
could
be
digitized
packaged
up
and
then
put
into
that
particular
operator's
time
slot
in
in
the
tdm
downlink,
so
yeah,
it's.
It
definitely
is
asymmetric
where
dvb,
s2
and
s2x
are
transmitted
on
the
downlink
and
you
you
have
a
preferred
or
a
native
digital
protocol
that
went
presumably
would
allow
you
to
to
more
easily
access.
B
A
B
Correct
anybody
that
can
receive
any
station
that
is
listening
should
be
able
to
receive
all
the
traffic
if
they
are
capable
enough,
and
so
since
we
we
have
adaptive
coding
and
modulation
here,
you
you
could
see
where
there
be
situations
where
a
station
could
listen
to
everything,
no
problem,
you
know,
but
even
a
small,
portable
or
marginal
station.
They
once
they're
part
of
the
mix
once
they're
they're
in
the
system.
You
know
they
at
first.
They
may
not
be
able
to
to
receive
everything.
B
Everybody
can
receive
the
physical
layer,
header
and
if
you
look
at
that-
and
you
go
oh,
you
know
this
is
below
my
my
limit
of
being
able
to
receive.
Then
you
can
see
where
the
link
would
adapt
to
that
particular
radio.
That's
kind
of
a
neat
thing
about
acm,
so
we
have
the
opportunity
for
doing
something:
pretty
pretty
cool
and
powerful
here
and
and
that's
kind
of
the
that's
that's
one
of
the
big
advantages
of
of
dbbs
too.
A
Okay,
so
yeah,
obviously,
we've
had
some
discussions
with
aaron
previously
a
week
ago,
last
week
anyway,
we
discussed
open
cpi,
which
I
think
sounds
really
positive,
and
it
will
be
good
to
try
and
integrate
that
into
the
project.
Hopefully
leave
some
of
our
pain
points
and
make
things
more
capable
in
the
future.
So
I'm
I'm
excited
to
see
how
that
progresses.
A
So
yeah
this
is
where
I'm
just
really
fully
my
opinion,
but
I'm
just
putting
things
out,
so
I
tried
to
think
of
what
kind
of
minor
goals
we
could
have
in
three
months
and
six
months.
I
mean
obviously
the
time
scale
so
well,
yeah
people
work
when
they
can
so
it's
not
exactly
something
enforced.
But
I
was
trying
to
think
of
kind
of
steps,
so
kind
of
immediate
steps
is.
It
would
be
great
to
start
trying
to
get
the
the
dvds
to
transmitter
work.
A
The
second
immediate
thing
is
getting
a
more
detailed
architecture
to
find
so
like
the
thing
that
anchor
mentioned,
how
we're
going
to
connect
the
different
pieces
together
so
that
it's,
if
not
crystal
clear
but
at
least
quite
clear,
at
a
glance
of
how
different
what
different
parts
are
and
how
they
talk
to
each
other
and
then
kind
of
secondary
steps.
I
think
it
would
be
good
to
start
looking
at
getting
rf
and
baseband
board.
A
Schematics
started
start
getting
some
designs
down
on
paper,
so
to
speak
because
yeah
that
will
be
a
big
piece
of
work.
So
the
sooner
we
start
that
process
the
better
and
then
yeah,
basically
trying
to
get
a
let's
say,
a
prototype,
rx
tx
data
path,
so
a
single
channel
of
uplink
to
the
jse
and
back
down
to
the
dvb
s2x,
so
yeah.
A
A
Okay,
so
kind
of
key
enablers-
I
I
see
I
was
getting
open,
cpi
flow
up
and
I
think
will
be
really
powerful
and
it
will
hopefully
help
with
some
of
that
over
the
testing,
because
we
can
abstract
away
the
horrible
interface
to
the
adi
chips
yeah.
Maybe
some
hardware
contributors
might
want
to
get
involved
with
the
schematic
and
pcb
aspect,
so
the
hardware
side.
A
So
I
want
to
have
a
think
about
defining
what
we
need
and
maybe
putting
some
work
factors
out.
People
might
be
interested
or
might
not,
but
yeah
it'd
be
good
to
see.
If
people
want
to
work
on
that
and
then
you
can
eventually
touch
this,
but
yeah
possible
can
well.
I
know
you
guys
are
collaborating
with
m17,
I'm
not
really
up
to
date
and
what
discussion
has
been
had
there,
but
obviously,
if
we're
using
m17
and
the
uplink
it's
good
to
yeah
maintain
contact.
A
B
There
is
an
effort
by
watches
callsign
but
rob
to
to.
He
started
working
with
a
pink
and
verilog
and
is
trying
to
just
to
start
the
process
of
hdl.
For
for
m17.
I
invited
him
to
to
feel
free
to
use
the
lab
and
to
to
share
share
his
work.
So
there
is
at
least
some
movement
towards
towards
hdl
from
m17.
B
I
would
call
it
kind
of
slow
and
and
and
whenever
he
has
time-
and
I
think
that
that's
something
that
we
should
start
looking
at,
supporting
and
collaborating
with
him
on
to
to
kind
of
get
it
into
into
our
into
fpga
or
asic
hardware,
and
there
is
also
another
effort.
We
just
ordered
pcbs
with
a
rfic
in
order
to
have
the
first
sort
of
development
stations
for
hardware
for
m17,
because
you're
exactly
right.
Up
until
now,
this
has
been
almost
completely
dominated
by
software.
B
C
Just
yeah
this
is
another
place
where
opencpi
can
help,
because
if,
if
we
have
software
versions
of
m17,
those
can
be
implemented
and
be
running
on
hardware,
I
mean
on
the
on
the
targeted
devices
and
then
once
hdl
we
have
some
of
those
components
in
hdl
we
can
pair
them
up
and
then
eventually
convert
convert
them
all
into
hdl,
but
I'm
just
throwing
that
out
there.
Thank.
A
C
B
B
A
F
Yep
cool,
if
I
may
ask
is,
is
this
the
radio
portion
or
the
radio
part
of
the
whole
mission?
If
that's
progressed
quite
a
bit,
what
hasn't
progressed?
When
do
we
actually
get
to
discuss
like
what
are
the
other
challenges?
Over
the
last
few
months,
I've
been
definitely
seeing
a
lot
of
progress.
You
know
in
various
things,
radio
related,
but
what
else
hasn't
been
done
or,
and
is
it
time
now
to
put
on
the
table
that?
F
A
Yeah,
absolutely,
in
my
opinion,
the
hardware
for
the
the
payload
is
still
a
considerable
amount
of
work.
It's
going
to
be
a
big
fpga
with
all
the
problems
that
come
with
that
and
rf.
So
it's
you
know
it's
not
trivial,
to
do
the
hardware
aspects
so,
in
my
opinion,
yeah.
We
should
start
getting
some
prototypes
in
that.
F
They
told
me
that
the
only
reason
I
I
was
able
to
do
18
months
of
like
developing,
to
get
to
trl
level
six
and
fly
in
seven
and
it
worked
perfectly-
was
that
I
had
been
running
a
test
loop
hardware
in
the
test-
loop
forever
accidentally
by
myself
for
myself
for
my
own
purposes,
nothing
to
do
with
formal
education
programming,
and
that
gave
us
the
co
that
gave
them
the
confidence
to
approve
it.
That
was
like
really
funny.
F
I
just
didn't
know
that
that
was
so
much
important
if
I
hadn't
done
it.
If
I'd
just
done
a
science
experiment
and
then
showed
up
at
the
flight
path
launch
pad,
it
would
have
probably
failed
just
like
everywhere
every
thing
other
fields,
that's
what
I
also
heard
from
the
university
of
minnesota
team.
You
know
those
students
were
really
like
they're
pushing
the
hundred
thousand
feet
boundary
model
right
now,
but
they've
flown
up
to
thirty
thousand
feet
in
two
different.
You
know
programs.
A
F
Yeah,
I'm
just
hoping
to
add
to
that
part
I
mean
I
would
like
to
get
testing.
You
know
if
you
guys
are
ready
to
release
code
batch
code
and
okay,
then
give
me
a
login
and
I'll
just
be
done
at
least
getting
some
experience
with.
You
know,
pushing
bits
and
packets
and
frames
yeah
absolutely
so
the
developing
team
needs
to
give
out
their
baby.
Is
it?
What
is
it
called
like?
They
need
to,
like
you
know,
say
goodbye
to
the
development
code
version
one
and
then
move
on
to
version
two
or
something
right.
F
B
Yeah,
you
don't
need
to
wait
for
any
code
to
be
released.
We
have
to
publish
as
we
create
so
at
any
point
in
time.
The
code
is
available
for
you
to
to
use
test.
Look
at
what
probably
would
help
would
be.
You
know
a
little
bit
of
versioning
to
say:
okay.
Well,
here's
a
here's,
a
a
gentle,
stopping
point.
It
sounds
like
that
might
help,
but
you
don't
have
to
wait.
We're
not
holding
back
any
code.
It's.
F
All
released
as
it's
created
well
michelle
the
reason
I'm
saying
this.
I
suppose
that
the
code
is
designed
for
some
particular
hardware.
That's
you
know
currently,
particularly
in
one
of
the
remote
labs,
and
I
don't
have
it.
I
have
my
own
test
bed,
but
you
know
my
test
bed
was
designed
for
another
purpose.
F
B
Yeah
the
target
is
the
fpgas
that
we
have
in
the
remote
lab.
So
in
the
case
of
somebody
wanting
to
help
test
and
and
do
things,
you
would
get
an
account
at
the
remote
lab
you
would
log
in
to
to
that.
You
would
target
the
hardware
that
we
have
identified
as
what
we're
working
with
and
and
then
you
would
test
there
if
you
wanted
to
test
it
on
different
hardware,
which
is
of
value,
as
we've
talked
before
about
wanting
this
to
be
why,
as
widely
used
as
possible,
which,
with.
F
B
F
Most
easiest
one,
if
I
logged
in,
if
I
logged
in
and
there
were
these
questions
or
privileges,
I
would
want
to
be
in
a
group
of
account
where
the
privileges
would
be
for
testing
and
not
for
code
development.
I
would
actually
test
the
system
test.
Do
the
very.
I
would
try
to
reduce
the
workload
of
the
development.
B
Yeah,
that's
very
good,
and-
and
I
strongly
concur
with
with
your
your
sentiments
and
and
everything
in
terms
of
like
privileges,
it
would
I'd
if
you,
if
you
wanted
some
sort
of
change
to
the
developer
privileges,
because
that's
what
we
do.
We
set
you
up
as
a
developer.
If
you
wanted
some
sort
of
change,
then
the
remote
lab
staff
can
handle
that.
B
F
If
you
formalize
as
part
of
a
plan,
then
during
the
review
process,
the
design
review
process
gate,
you
would
say
that
well,
yeah
development
team
did
this
and
then
they
pushed
it
out
and
then
the
testing
team
found
this
and
you
could
describe
it
in
a
proper
software
maturity
model,
something
or
other
some
kind
of
practice.
Yeah.
That's
right!
F
B
A
So
another
thing
I
say
is
that
you'll
be
the
interface
tom
and
I
was
going
to
say
that
at
the
moment
it's
it's
not
very
exciting
to
test
things,
because
I
mean
it's
proper
development
practice,
but
everything's
kind
of
unit
test.
So
basically
you
hit
run
tests
and
then
it
says
pass
and
that's
done,
but
once
we
get
into
the
next
phase,
which
is
really
kind
of
integrating
these
pieces
together,
then
it
will
be
more
of
an
interesting
task
to
actually
start
learning
stuff
in
the
hardware
and
actually.
F
F
A
A
A
C
A
Yes,
the
payload
is
the
basically
the
the
transceiver
that
well
yeah
that
we
we
work
on
and
then
typically
it
you'd
have
basically
the
bus
which
is
the
satellite,
so
that
has
the
basically
all
the
housekeeping
know.
It
has
solar
panels
and
power
generation,
batteries,
attitude,
control,
determination
and
control.
Basically,.
A
F
B
F
B
Yeah,
thank
you.
I
think
I
can
answer
your
question
in
general.
Our
mission
is
to
provide
the
communications
package,
that's
the
price
priority
and
that's
what
we
got
funding
for.
We
want
to
put
it
into
space.
We
have
several
partners
that
are
very
interested
in
helping
us
out
here.
It
may
be
that
we
have
to
purchase
a
lot
of
the
non-communications
payload
stuff
and
that's
totally
okay.
B
So
we
do
have
options,
opportunities,
potential
partners,
but
what
we
need
to
do
is
focus
entirely
on
the
communications
package
nail
it
show
it
off
set
it
up,
terrestrially
keep
doing
what
we're
doing,
and
then
we
have
a
compelling
argument,
for
we
really
do
deserve
to
to
get
a
donation
or
some
help
or
some
collaboration
with
people
where
the
the
power
supply
or
the
thruster
is
their
primary
motivation
and
their
primary
area
of
expertise.
True.
F
You
know
three,
but
three
by
one
used
times
two
in
parallel
or
sequence,
you
can
stack
it
and
make
it
cubic,
but,
however,
where
I'm
afraid
of
what
I've
gone
through
in
my
small
experience
is
that
if
I
hadn't
paid
attention
to
the
interior
of
the
structure,
the
protrusions,
the
switches,
the
rb,
are
removed
before
flights,
which
is
the
power
buses,
the
connectors,
the
other
devices
that
we're
going
to
be
interfering
with,
etc,
etc,
etc,
and
step
back
up
and
saying.
Well,
that's
my
my
available
volume
or
potential
available
volume.
F
Then
the
board
that
I
designed
that
had
to
fit
in
that
potential
volume
dictated
what
I
could
actually
put
on
where
you
know,
because
in
the
newer
cubesat
quote
standards
unquote,
they
have
even
limited
the
amount
of
the
type
of
clocks
that
you
can
run
on
your
payload
board.
I
mean
because
there
are
other
flight
computers
that
are
affected
by
those
say,
the
harmonics
of
the
clock
frequencies.
F
Now
I'm
just
asking
because
I
use
switch
mode
regulators,
I
did
low
level
power
control
whatever
and
if
I
was
lucky
that
it
worked
the
unlucky,
but
the
next
version
I
wouldn't
have
been
lucky
because
it
would
have
interfered
with
something
else
so
before
the
card
before
the
horse
or
the
horse
before
the
car.
If
you
give
me
an
outer
volume,
that's
great,
but
what
about
the
interior
space?
If
we
just
pick
a
bus,
I.
B
A
F
So
whatever
it
is,
I
mean
it
would
be
nice
to
know
what
tokens
I
can
design
a
board,
and
you
know
give
the
outline
I've
done
that
so
many
times
I
have
libraries
built
up
like
okay,
it's
a
it's
a
pumpkin
board
or
it's
xyz
board
or
whatever
or
I
can
just
make
it
like.
You
know
I
can
shove
that
to
you
to
tom
for
his
keycard
work
or
my
keycard
work.
Whatever
it
can
be
done,
that's
right.
We
can
start
to
lay
out
things.
B
I
think
it
would
probably
be
a
little
too
early
to
lay
out
until
after
the
sorts
of
things
that
we're
talking
about
today
or
demonstrated
as
quickly
as
possible.
You
know
to
get
an
end-to-end
system
working
in
the
lab,
so
to
do
the
lab
call
that
would
be
first
and
then
the
the
there
are
some
things
that
we
already
know
about
the
layout
just
based
on
the
fpga
that
we're
targeting
and
the
fact
that
we
want
to
be
very
highly
resilient.
B
So
those
two
things
you
know
the
particular
fpga
and
the
needs
of
the
fpga
for
layout
and
defending
it
and
then
the
particular
resiliency
higher
level
resiliency
functions
that
are
in
the
architecture
document
linked.
Those
two
things
determine
a
large
amount
of
the
bill,
materials
and
a
large
amount
of
the
layout,
but
not
all
of
it.
The
things
that
you're
talking
about
are
still
to
be
determined,
because
we
don't
know
where
exactly
in
the
interior
of
a
spacecraft,
we
will
be
so
those
things
will
have
to
be
driven
by
those
opportunities
when
they
arise.
B
Those
opportunities
can
arise
as
soon
as
I
have
a
a
good
demo
that
I
can
take
out
into
the
world,
and
I
promise
you
that
I
will
do
absolutely
everything
I
can
to
show
it
off
to
whoever
will
look
at
it
and
to
see
the
value
that
we
bring
and
all
of
that.
So
I
think
it
might
be
a
little
early
to
actually
start
the
work.
But
it
is
definitely
the
right
time
to
bring
it
up
and
I
think
you've
stated
it
very
well.
F
And
the
reason
I'm
I'm
saying
is
that
this
is
my.
You
know.
You
know
cool
powerpoint,
basically
right
here,
this
fpga
device,
I'm
like
on
your
your
test
boards
and
dev
boards
and
everything
is
like
you
know,
probably
wired
to
some
kind
of
you
know.
You
know
components
of
of
fpga
of
die
components,
so
the
timing
closures
I
mean
you
know
eventually
on
that
board-
are
already
defined.
F
Where
would
you
place
the
system
design
constraint
so
that
when
we
go
to
board
making,
we
don't
end
up
like
well?
Okay,
excuse
me,
I
can't
put
the
chip
here.
I
can
put
the
chip
there,
but
you've
got
to
change
your
code
see
the
flight
code
has
to
mature
well
in
advance
of
the
hardware
or
vice
versa,.
A
Yeah,
I
mean
that's
because
that's
more
in
the
payload
domain,
but
I
mean
largely
split
between
it
and
yeah:
a
baseband
board,
baseband
rf
board
a
digital
board
and
then
an
unspecified
rf,
the
actual
rf
board,
but
I
think
it's
kind
of
well
defined.
That
is
each
of
those
boards
will
be
one
new
structure
or
one
new
form
factor.
Okay
and
then,
as
michelle
says,
a
lot
of
the
things
kind
of
fall.
Naturally,
from
that.
F
So
you're
going
to
be
able
to
say,
defer
the
decision
to
position
the
chips
and
thereby
constraining
the
pathways
between
memory
and
and
fpga
until
a
later
stage
right
I
mean
that's
a
decision
that
you're
going
to
show
to
a
later
stage
or
you
can
just
do
it
right
and
say:
okay,
then
the
code
can
mature.
Your
code
has
to
mature,
sooner
or
later.
A
F
For
the
payload,
so
if
I
understand
you
tom,
I'm
thinking
that
you're
saying
that
we
should
at
least
go
with
a
version
one
like
the
decision
tree
of
what
goes
where
and
leave
it
at
then
and
then
design
the
software
to
catch
up.
And
then
we
at
least
know
the
outer
boundaries
that
we
can't
have.
You
know
fpga
here
and
at
two
feet
away
a
memory
chip
or
something
right.
A
B
Yeah,
there's
things
that
we
can.
We
already
know
about
the
layout
based
on
the
target
fpga,
based
on
how
much
memory
we
think
we
need
and
based
on
how
much,
how
resilient
we
have
to
be
to
radiation.
Those
things
drive
large
chunks
of
this,
but
not
all
of
it.
So
there
are
some
things
that
are
going
to
have
to
be
left
until
we
figure
out
exactly
what
package
we
have.
Those
are
integration
questions,
but
the
layout
for
the
things
that
we
know
that's
under
our
control.
F
B
The
next
step
is
to
use
the
trends
gear,
which
gives
you
more
visibility
to
the
to
the
fpga
and
then
step
into
our
own
prototype
board.
That's
been
in
the
roadmap
as
presented
by
by
thomas
and
others
at
like
at
ham
expo
when
he
gave
a
walk
through
of
the
roadmap,
so
so
yeah
I
mean
it
seems
like
we
would
have
to
have
prototype
hardware
and
then,
after
that
would
be
whatever
is
needed
to
get
to
to
flight,
which
is
a
whole
whole
other
set
of
work,
but
yeah.
A
So
on
the
satellite
side
of
the
rugby
earth,
michelle
that
it's
too
much
to
bite
off
to
try
and
build
our
own
satellite
from
scratch,
but
I
do
think
so.
There's
yeah,
basically
some
high
level
work
that
can
be
done
or
needs
to
be
done.
Actually
we
basically
have
to
have
some
idea
of
budgets
for
power
link,
budget
thermal
budget
proportion
and
well
adcs,
maybe
less
so,
but
we
have
to
have
basically
an
idea
of
the
dimensions
that
we're
dealing
with
well
order
of
magnitude
of
power
links
and
thermal.
B
Correct
we
have
a
really
good,
a
very
good
thermal
analysis
package
with
thermal
desktop
that
we
can
use.
So
as
soon
as
we
have
anything
to
analyze,
then
we
have
access
to
that.
That's
a
community
resource
that
we
we
got.
We
have
a
large
amount
of
work.
The
by
far
the
best
area
in
your
list
is
the
link
link
budget.
B
We
have
some
propulsion
work
with
ais,
but
those
those
are
our
small
ion
engines.
So
you
know
that's
not
a
good
match
for
a
larger
spacecraft,
but
that's
an
area
where
we're
active
and
keeping
keeping
up
power,
I
would
say,
has
not
been
addressed
very
much
and
we
don't
really
know
our
power
budget.
We
assumed
a
very
large
number
and
we
assumed
the
largest
number
that
we
could
get
and
not
run
into
thermal
problems
and
that's
as
far
as
we
got
on
that
on
that
side
and
adcs
we're
actually
not
in
terrible
shape.
A
Yeah,
I
think
that's
good.
I
think
my
opinion
is
that
we
should-
and
I
can
do
this
volunteer
myself
do
it,
but
that
we
should
so,
I
like
to
think
in
numbers
actual
numbers,
even
if
you
know
that
the
numbers
are
going
to
only
be
approximate
but
I'd
like.
I
think
it's
good
to
have
a
page
where
we
say
this
is
the
number
we're
targeting
for
power
link
thermal
proportion
just
so
that
there's
some
dimensional
people
have
a
good
idea
of
the
dimensions.
A
F
Can
I
ask,
if
can
I
ask
if,
in
terms
of
what's
just
at
a
very
high
level,
could
could
I,
if,
if
I
built
a
subsystem
of
a
cluster
of
becky,
mark
thrusters
and
I've
designed
this
before
for
other
missions
and
I've
designed
this
for
my
own
hardware
and
it's
flown
in
space
and
everything?
So
I'm
I'm
starting
from
that
point?
I
don't
have
to
reinvent
the
bill.
A
F
1.4
kilowatts
of
a
theoretical
per
square
meter.
So
if
we
have
a
couple
of
we
fold
out
panels,
we're
gonna
have
lots
of
watts.
Now
you
can
store
them
or
you
can
let
some
of
the
float.
Basically,
that's
extra
of
the
charge
just
be
utilized
for
propulsion,
and
I
I've
designed
those
systems.
You
know
that
have
actually
gone
out
there.
So
what
I'm
asking
is
suppose
that
I
just
came
up
with
kind
of
like
a
first
order.
A
rough
order.
F
F
Okay,
at
any
time,
50
hertz
times,
let's
say
four
or
five
or
six
seven
thrusters,
so
that's
35,
watts,
seven
thrust:
seven
total
thrusters
at
50
hertz
will
give
you
35
times
one
micro
newton's
at
35
newton
seconds
of
thrust.
At
any
time
I
mean
it's,
really
it
really
can
scale
up.
The
question
I
have
is:
what
would
you
when?
Would
you
like
to
know
it?
F
What
would
you
like
to
know
it
I
mean:
do
you
want
a
model
that
works
in
sdk
or
or
gmat
first
to
see,
if
it
does
it
or
you're
going
to
give
me
the
model
and
then
say,
go
come
up
with
a
solution?
What
would
I
like
I'd
like
to
I'd
like
to
do
an
academic
I'd
like
to
do
an
analysis
and
present
it
through
or
I
publish
it.
A
F
C
F
Yeah
you
you,
please
invite
me
to
a
meeting.
You
know
that
you're
gonna
host
okay,
so
that
you
can
begin
to
collect
your
notes,
I'm
just
recording
that,
maybe
today
or
maybe
next
day
or
next
month,
but
you've
got
zoom
whatever
recording
and
ori
obligations,
just
go
ahead
and
just
put
me
on
a
list
of
people
that
you
want
to
talk
to
sure.
Okay,
thank
you.
A
Yeah,
so,
basically
kind
of
knowing
that
slightly
is
yeah.
Basically,
I
think
the
only
real
actionable
work
we
can
do
on
this
at
the
moment
is
just
to
have
yeah
an
idea
of
order
of
magnitude
on
these
budget
items,
and
then
we
have
to
wait
a
bit
longer
before
we
can
have
much
more
maturity
in
the
payload
before
we
can
really
take
the
satellite
site
any
further.
A
F
If
possible,
I
mean
we
could
do
this
on
a
rolling
basis
and
I'll
ask
you
to
just
you
know:
do
it
on
a
linear
time,
like
you
know,
semesters
quarters
whatever,
so
that
you
know
we
can
you
know
we
can
all
commit
time.
There
is
a
level
of
commitment
to
do
this
and
then
and
other
people
will
be
using
that
from
that
point.
So
six
months
is
like
way
down
the
line.
Nine
months,
that's
like
next
year.
A
A
C
I
just
want
to
summarize
and
highlight
one
important
point:
one
of
the
main
thing
going
forward
and
short-term
goal
for
me
personally
and
I
think
so
is
also
working
on
that
and
just
wanted
to
bring
it
to
your
tom
is
to
get
end-to-end
system
working.
That's
what
I
have
been
trying
with
ip
and
get
the
transmitter
end-to-end
working
and
get
all
the
components
you
talk
of
gsc.
You
talk
of
the
dbvs
to
protocol
implementation,
so
bringing
that
all
together
and
in
innocent
keeping
in
a
simple
term.
I
want.
C
B
Strong
strongly
agree:
we
have
an
opportunity
this
coming
thursday
to
demonstrate
a
beacon,
an
example
of
our
system
as
a
beacon
in
los
angeles,
to
one
of
the
technical
groups.
So
I
think
I'm
going
to
go
ahead
and
take
take
that
as
a
a
really
good
short
deadline
to
get
something
transmitting
over
the
air.
B
So
you
may
have
seen
some
of
the
videos
that
are
posted
about
the
beacon
work
but
like
this
could
be
part
of
this
end-to-end
system,
and
I
could
not
strongly
more
strongly
agree
working
over
the
air
is
to
me,
that's
the
that's
all
that
matters
is
and
that
it
should
work
quickly
so
that
we
fail
quickly
and
figure
out
what's
wrong.
So,
yes,
I
have
a
little
bit
of
work
that
and
I've
pulled
in
a
couple
of
people
to
help
me
with
the
python.
B
So
when
we
talk
about
an
end-to-end
system
in
a
lab
call
to
me
a
lab,
a
lab
call
like
a
lab
phone
call.
Laboratory
phone
call
means
that
things
are
working,
but
it
might
look
ugly.
You
know
it's
a
bunch
of
dev
boards
strung
together
code,
that's
holding
hands,
you
know,
and,
and
all
of
that
and
the
the
other
thing
that
we
that
we
have
talked
about
in
the
past
that
we
haven't
really
addressed
today
is
like
a
some
sort
of
end
in
model
or
something
like
a
simulation
in
the
past.
B
We
assumed
that
good
new
radio
would
be
a
good
place
for
this.
Gnu
radio
has
some
significant
limitations
that
we've
run
into
I've
talked
with
both
ben
hilburn
and
derek
kozell,
and
a
couple
of
other
people
in
the
architecture
team,
and
we
really
can't
do
our
full
simulated
system
end
to
end
with
the
bandwidths
and
speeds
and
numbers
of
channels
that
we're
talking
about.
So
I
think
that
we
need
to
kind
of
maybe
back
up
and
say.
Okay,
do
we
need
to
do
everything
in
matlab
or
octave?
Do
we
need
to?
B
A
All
right,
I
think,
see
open
cpi
is
probably
the
best
candidate
for
helping
us
there.
I
mean
if
we
have
c
plus
fast
implementations,
which
basically
is
going
to
be
radio
implementations,
then
we
can,
as
my
understanding,
we
can
basically
create
yeah
just
a
simulation
on
a
pc
which
matches
the
hdl,
maybe
mention
more.
C
Yeah
I
mean
yeah
that
is
possible,
but
you're
again
we're
gonna
run
into
that
same
implementation.
Like
a
new
radio
has
this?
Is
you
can't
feed
the
dac
with
you
know
the
10
megahertz
signal
and
everything's
running
in
software?
Then
there
are.
There
are
some
limitations
there,
but
if
you
want
to
do
a
simulation,
that's
offline
and
and
then
that's
that's
a
possibility
as
well,
so
it
just
depends.
A
F
F
F
A
A
A
F
A
D
B
A
B
Yeah,
I'm
sorry,
I
I
didn't
have
anything
to
add
other
than
the
idea
that
we
have
have
something
at
the
top
level
that
people
can
can
click
on
and
then,
if
they
want,
then
it
gives
them
the
opportunity
to
get
the
high
level
overview
very
quickly
and
to
see
this
this
laid
out
and
then
it
naturally
sort
of
organically
flows
down
into
more
and
more
detail
so
that
people
can
find
it.
So
I
I
think
that's
good
and-
and
you
know,
being
a
simple
person-
that's
that's
about
it.
F
F
A
F
A
Yeah,
I
think,
to
just
repeat
what
many
people
have
said.
I
think
the
most
important
thing
now
is
that
we
can
get
an
end-to-end
system,
because
I
really
think
once
we
can
start
playing
with
in-depth
system,
it
will
change
a
lot
of
things,
it'll
be
really
cool
to
see
and
it
will
yeah
inspire
more
people.
D
B
F
B
Yeah
yeah
originally
was
the
a
a
r
e
x
rx
diagram,
correct,
it's
the
it's
under
in.
If
you
go
to
facebook
ground
documents,
let's
say
I'm
backing
up
to
find
it
engineering
requirements.
Architecture
is
where
all
of
that
stuff
is
is
found
the
pdf
and
then
all
the
supporting
drawings
as
separate.
So
the
the
idea
of
a
default
digital
downlink
or
digital
content
coming
at
you
has
been
that
that's
in
the
plan.
B
Ron
economos
about
whether
or
not
we
should
start
out
with
like
having
a
beacon
or
start
out
with
using
mpeg
as
a
transport
stream.
I
don't
really
think
that's
a
good
idea,
long
term,
however
short
term
there's
an
awful
lot
of
people
that
have
amateur
television
stations
that
will
receive
it.
So
it's
a
balance
between
getting
traction
with
the
hardware.
That's
out
there
versus
bringing
something,
that's
a
big
step
forward
to
people
if
we
choose
and
we
have
chosen
gse
over
over
the
other
transport
stream.
B
That
means
it's
on
to
us
to
make
it
easy
for
people
to
to
receive
and
having
a
default
digital
downlink.
You
know
something:
that's
if
there
isn't
any
traffic,
then
it
has
essentially
a
test
signal
that
you
can
then
use
just
like
we
use
microwave
beacons
on
the
air
all
the
time
to
test
our
microwave
stations.
B
Then
you
have
a
test
signal
going
on
all
the
time
that
that
is
of
great
value,
so
it
can
run
through
all
the
different
mod
codes.
It
can
provide
different
types
of
digital
content.
You
can
do
you
know
here's
a
video
signal
at
the
top
of
the
hour.
We're
going
to
go
we're
going
to
run
through
all
the
mod
codes.
Things
like
that.
F
So
if
I
understand
you
michelle
in
this
diagram,
that's
on
the
screen
below
the
dbvs
2x
green
block.
Could
there
be
a
potential
you
know
say
just
for
conveniently
a
test
signal,
but
that
would
all
feed
into
the
rf.
I
mean
it
would
either
be
going
through
a
selector
switch
or
in
parallel
to
the
rf
d2a
converter,
so
that
we
generate
in
the
same
band
separated
by
a
few
megahertz
or
so
two
carriers,
one
your
generate
dvd.
You
know
what
multiplex
stuff
and
another
is
just
occasional
test
signals
for.
B
F
A
E
E
F
Oh
beautiful
I've
I've
been
part
of
the
amsat
community
for
a
while,
and
I
read
some
of
their
old
archives.
They
were
given
to
me
by
perry,
klein
and
he's
been
trying
to.
You
know
guide
me
for
a
number
of
years
now
so
apparently,
the
early
days
of
amsat
hf
was
it.
That
was
the
way
they
would
communicate
back
and
forth.
There's
even
an
older
satellite
spec
and
circuit
diagram
still
around
in
my
archives
of
they
check,
beacon
down
to
earth.
Weird.
E
E
A
B
E
And,
of
course,
the
thing
I'm
looking
it
comes
out
up
to
six
gigahertz,
so
the
missing
element
right
now
is
an
up
converter
to
go
from
we'll
be
running
an
if
of
2.4
gigahertz
and
with
a
and
we
need
to
up
convert
that
to
six
to
10.4
gigahertz
within
eight
gigahertz.
Lo.
So
if
anybody
has
one
of
those
sitting
around
give
me
a
call.
E
E
E
B
F
Mike,
I
I'm
I'm
glad
to
hear
the
word
inflatable,
because
my
master's
thesis
was
on
interplanetary
networks,
where
I
used
sts-77
inflatable
antenna
experiment
as
my
model,
married
with
the
mars
reconnaissance
orbiter,
and
about
150
relay
stations
for
a
universal
network
spanning
the
solar
system
able
to
be
utilized
all
of
the
time,
regardless
of
the
geometry
of
the
planets,
and
that
was
published
in
2011.
But
the
crux
of
the
problem.
The
solution
was
with
jpl
that
I
worked
out
was
with
the
inflatable
antennas
14.6
meter
class.
E
F
A
Okay,
I
don't
really
have
anything
else.
I
think
my
main
action
here
is
I'm
going
to
make
a
more
detailed
diagram
of
this
multi-tiered
version
of
this
and
then
basically
start
sharing
it
with
you
guys.
You
can
tell
me
where
it's
wrong
and
then
we'll
kind
of
iterate
a
few
times
and
once
we're
happy
with
it,
we
can
stick
on
the
website,
so
yeah,
that's!
That's
kind
of.
I
think
the
next
steps
on
my
side
anyway,.
F
A
A
B
A
B
Be
a
good,
a
good
sort
of
gathering
point
for
all
of
the
work
and
yeah.
I
like
the
three
and
six
months
thing
and
the
the
near
term
I'll
be
working
with
the
usual
suspects
to
get
a
sort
of
a
multi
multimedia
beacon
demonstrated
by
by
thursday
and
then
we'll
anything
that
we
do.
Weekly
stand-up
meeting
is
where
that
will
be,
and
we're
going
to
strive
to
keep
it.
What
we've
done,
what
we
have
planned
any
roadblocks
and
any
necessary
resources.
F
Excellent
yeah,
can
I
just
ask
the
group:
do
we
now
shift
to
partitioning
the
well
just
think
of
the
of
the
channels,
but
really
slack
channels
are
great.
You
know,
ad
hoc
mechanisms,
but
like
say
propulsion
could
be
one
track
major
track.
Major
tracks,
like
you
know,
radio
is
a
major
track.
Propulsion
major
track,
orbital
analysis,
major
track,
but
basically
have
stand-ups.
F
F
B
Meetings
are
scheduled
whenever
there's
a
need
so
as
soon
as
there's
enough
work
to
have
a
stand
up
on
things
like
propulsion,
which
spacecraft
propulsion
and
bus
and
a
number
of
other
things
all
currently
have
channels,
so
that,
once
the
activity
level
rises
up
to
the
point
where
it
requires
a
stand-up
to
organize
the
work
further,
then
they
happen.
But
the
meeting
schedules
are
driven
by
the
level
of
activity.