It's the whole sector. You definitely don't want anybody at QPSK, and
IMO you don't actually want customers who can't get 64QAM. Anybody
running QPSK would be an unhappy customer and he'd weaken the whole
sector.
The point of the chart was this: I said earlier, "even 900 doesn't work
with a mile of forest in the way". People responded, "but SCADA at
115kbps!". My rebuttal is that crappy speeds are already an option, but
they don't count as "working" if you're selling internet access.
I'm not sure about the SNR question.
------ Original Message ------
From: "Ken Hohhof" <[email protected]>
To: [email protected]
Sent: 9/25/2016 3:30:43 PM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
Something that always seems fuzzy with WIMAX and now LTE is whether you
can have a bunch of subscribers all getting that throughput
simultaneously, or if that’s the entire sector capacity. If the entire
sector capacity is 0.91 Mbps download shared over however many
customers you need to make that basestation profitable, then it’s silly
to talk about an MCS0 link. In fact, even the MCS10 numbers from that
chart wouldn’t really be useful for fixed broadband service. For best
effort connectivity from a mobile client, maybe it’s acceptable.
The other thing is I remember one vendor saying their SNR numbers were
per subcarrier or something, and you had to add a fudge factor of
something like 10 dB to do an apples-to-apples comparison with the non
LTE world. Not sure if that applies here.
From: Af [mailto:[email protected]] On Behalf Of Adam Moffett
Sent: Sunday, September 25, 2016 2:02 PM
To:[email protected]
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
I dunno, but the chart says it theoretically works at an SINR of -6.7db
Seems insane.
At the other end of the chart at MCS28 you're getting 42.46mbps x
6.61mbps.....and that's supposed to work with 24 SINR.
------ Original Message ------
From: "Chuck McCown" <[email protected]>
To: [email protected]
Sent: 9/25/2016 12:17:14 PM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
What is 0-QPSK? CW?
From:Adam Moffett
Sent: Sunday, September 25, 2016 5:09 AM
To:[email protected] ; [email protected]
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
Undoubtedly true guys, but same thing. Assuming this table comes
through, it's showing you the bottom mod levels on LTE at 10mhz
channel size. So yeah, in theory we could hook up somebody at -100
and it would "work", but you'd be spending a lot of money to not get
much capacity. SCADA might "work" for George's internet customers in
the same sense that scraping the bottom on LTE would "work".
Modulation and Coding Scheme
Max troughtput [Mbps]
SINR (dB)
Receiver Sensitivity (dBm)
Minimum by DL/UL split
Required SINR at Cell Edge (dB)
DL MCS
UL MCS
DL
UL
DL
UL
DL
UL
DL
UL
0-QPSK
0-QPSK
0.91 Mbps
0.25 Mbps
-1.2 dB
-1.0 dB
-6.7 dB
-3.3 dB
-106.1 dBm
-102.3 dBm
1-QPSK
1-QPSK
1.18 Mbps
0.32 Mbps
0.0 dB
0.1 dB
-5.6 dB
-2.4 dB
-105.1 dBm
-101.4 dBm
2-QPSK
2-QPSK
1.45 Mbps
0.40 Mbps
0.7 dB
0.7 dB
-4.8 dB
-1.6 dB
-104.3 dBm
-100.6 dBm
3-QPSK
3-QPSK
1.87 Mbps
0.51 Mbps
1.7 dB
1.7 dB
-3.8 dB
-0.5 dB
-103.3 dBm
-99.5 dBm
4-QPSK
4-QPSK
2.38 Mbps
0.65 Mbps
2.7 dB
2.8 dB
-2.8 dB
0.4 dB
-102.2 dBm
-98.5 dBm
5-QPSK
5-QPSK
2.88 Mbps
0.79 Mbps
3.6 dB
3.5 dB
-1.7 dB
1.3 dB
-101.1 dBm
-97.7 dBm
6-QPSK
6-QPSK
3.38 Mbps
0.93 Mbps
4.6 dB
4.2 dB
-0.7 dB
2.2 dB
-100.2 dBm
-96.7 dBm
7-QPSK
7-QPSK
4.07 Mbps
1.12 Mbps
5.6 dB
5.3 dB
0.6 dB
3.4 dB
-98.9 dBm
-95.6 dBm
8-QPSK
8-QPSK
4.57 Mbps
1.25 Mbps
6.5 dB
6.0 dB
1.5 dB
4.2 dB
-97.9 dBm
-94.7 dBm
9-QPSK
9-QPSK
5.24 Mbps
1.44 Mbps
7.6 dB
6.9 dB
2.5 dB
5.6 dB
-97.0 dBm
-93.3 dBm
10-16QAM
10-QPSK
5.24 Mbps
1.58 Mbps
7.7 dB
7.6 dB
2.5 dB
6.2 dB
-97.0 dBm
-92.7 dBm
------ Original Message ------
From: "Colin Stanners" <[email protected]>
To: "[email protected]" <[email protected]>
Sent: 9/25/2016 12:11:30 AM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
Some SCADA systems run at like 115kbaud with -103db receive
sensitivity in a 1mhz channel... that will survive a lot.
On Sat, Sep 24, 2016 at 10:17 PM, Jaime Solorza
<[email protected]> wrote:
Depends on the system.. SCADA type radios have no problem with
trees...
On Sep 24, 2016 7:22 PM, "Adam Moffett" <[email protected]> wrote:
Might be worth a shot I guess. Depends how strong your 900 is
coming in.
Attenuation is attenuation though...whether it's LTE or something
else. If the path literally passes through a mile of trees then
I'm surprised even 900 works.
------ Original Message ------
From: "George Skorup" <[email protected]>
To: [email protected]
Sent: 9/24/2016 3:45:44 PM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
And what happens when the noise floor increases and the eNB can't
hear those shitty CPEs anymore? Nevermind, that question answered
itself.
The boss keeps wanting to try an LTE sector at sites where we have
a mile or more deep trees (where we know 900 FSK barely works now,
not only due to power levels but noise floor too). My fear is that
it actually does "work" (meaning horrible mod levels) and he'll
want to run with it. So we'll get what, a couple Mbps out of a
sector. Sounds a lot like 900 FSK. Sounds a lot like 900 450i with
a horrible noise floor. So we gain nothing and spent a ton of
money. Great idea. And we won't end up getting all of the
customers off of the 900 anyway, that I'm sure.
On 9/24/2016 11:47 AM, Adam Moffett wrote:
In Wimax it's 4x4....I'm pretty sure we'll have 4x4 in LTE as
well, but I think feature was released only a month or so ago.
We have a few places with split sectors, so we'll be able to
compare to 2x2.
From what I understand, LTE's frame structure is such that it can
hang on to a crummy signal longer than Wimax. It was explained
to me that Wimax puts the synchronization data in the pre-amble
which has to be received on every subcarrier, whereas LTE has
that data interspersed among the subcarriers, so where your weak
wimax CPE sometimes cuts in an out, an LTE CPE in the same
conditions can stay connected. It also has lower mod levels
that let it operate right down to the noise floor. And at least
in theory you'll get more throughput than you get in the same
conditions on Wimax.
I have the same reservations as you about the low mod levels
thing. Just because they work doesn't mean you want them. We're
not intentionally installing anything weaker than a -80 RSSI
right now, so we really ought to be ok on that front.
-Adam
------ Original Message ------
From: "George Skorup" <[email protected]>
To: [email protected]
Sent: 9/23/2016 11:23:57 PM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
Well, let me ask this. Are you doing 2x2 or 4x4 on the Telrad?
Obviously 4x4 would give a slight advantage.
My whole thing is, OK, it might work through a shit ton of
trees. Linked up and able to move some traffic is one thing. But
a whole bunch of low modulation customers on a sector is not
worth the investment. LTE, Wimax, 450i 900.. whatever it may be.
I know of a Telrad installation where they couldn't make it
work. Turned out to be interference. They had some guys from
Israel come "fix" it. I won't say any names, but I now see what
they did to get it working. It's in the 3.5 band.. because I can
see them on my 450's spectrum analyzer. From multiple sectors on
multiple towers, so I know what direction it's coming from. And
I have no doubt they're running it over powered.
Welp, we have a BaiCells demo kit, so we'll see what happens.
On 9/23/2016 9:52 PM, Adam Moffett wrote:
We've had Telrad Compact 1000's for around 2.5 years, but
they're running Wimax firmware because we were replacing older
16e installations. We have a number of sites now that have
entirely dual mode CPE so we're about to pull the trigger on
LTE. We're installing four LTE base stations next week on
brand new sites, and assuming those go well we'll upgrade some
existing Wimax sites.
So yeah, within the next few weeks I'll know more. I'll
definitely report back.
It's interesting that you phrase it as "if it works at all."
The issue with Wimax has never been it "working", it's just
that it comes with a lot of quirks and it sucks at
administration and troubleshooting. I'm speaking of Wimax in
general here, not Telrad specifically....and I've used Wimax
from three different vendors now. I have no fear about LTE
working. I am afraid it will turn out to be cut from the same
cloth as Wimax.
------ Original Message ------
From: "George Skorup" <[email protected]>
To: [email protected]
Sent: 9/23/2016 8:04:34 PM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
Aren't you doing Telrad? Please let us know if it works at
all.
On 9/23/2016 4:05 PM, Adam Moffett wrote:
I'll let you know in a few weeks.
------ Original Message ------
From: [email protected]
To: [email protected]
Sent: 9/23/2016 5:01:02 PM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
I wonder what LTE would do with the same RSSI.
From:Adam Moffett
Sent: Friday, September 23, 2016 2:46 PM
To:[email protected]
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
Oh I also have somebody with a -88 who gets about half that.
900 was the last ditch effort for both of these.
With wimax from the same tower we got a big fat nothing at
both locations.
------ Original Message ------
From: "Adam Moffett" <[email protected]>
To: [email protected]
Sent: 9/23/2016 4:44:21 PM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
On the other end of the quality spectrum:
Link Test with Bridging
VC
Downlink
Uplink
Aggregate
Packet Transmit
Packet Receive
Actual
Actual
19
6.07 Mbps
1.32 Mbps
7.39 Mbps, 474 pps
821 (410 pps)
128(64 pps)
That's a -85 on a 5mhz channel. On any wider channel I
lose this guy.
------ Original Message ------
From: "Dave" <[email protected]>
To: [email protected]
Sent: 9/23/2016 4:17:36 PM
Subject: Re: [AFMUG] 900 MHz PMP450i :: Any real numbers?
This is from one of ours
Current Results Status
Stats for LUID: 3 Test Duration: 5 Pkt Length: 1714
Test Direction Bi-Directional
RF Link Test
VC
Downlink
Uplink
Aggregate
Packet Transmit
Packet Receive
Actual
Actual
19
26.13 Mbps
6.78 Mbps
32.92 Mbps, 2367 pps
2389 (477 pps)
9450(1890 pps)
Efficiency
Downlink
Uplink
Efficiency
Fragments
count
Efficiency
Fragments
count
Actual
Expected
Actual
Expected
100%
255254
255254
78%
84124
66231
Link Test ran on 03:59:48 01/09/2011 UTC
Currently transmitting at:
VC 19 Rate 8X/8X MIMO-B
On 09/23/2016 03:12 PM, Bill Prince wrote:
We're looking at doing a forklift on a couple of 900 MHz
FSK APs. What are real world throughput numbers that any
of you pioneers are getting? We would probably want to do
10 MHz channels at first, and I would hope that we could
get > 15 Mbps in download, but maybe I'm being too
conservative?
The places we are looking at do not have Smart Meter
issues.
--
bp
part15sbs{at}gmail{dot}com
--
