pkg: Semtech LoRaMAC #8264

aabadie · 2017-12-14T10:12:03Z

Contribution description

This PR imports the Semtech LoRaMAC reference implementation as a package and proposes a simple API for interacting with a LoRaWAN network.

What has been tested so far:

OTAA and ABP activation
Emission and reception of confirmed and unconfirmed messages
Message reception sent from the LoRaWAN backend
Adaptive datarate

This PR also provides a test application for setting/getting some LoRaMAC parameters, joining to a network and sending/receiving data to/from the network.

This PR has been tested with SX1272 and SX1276 devices, using Nucleo 64 boards and the b-l072z-lrwan1 board.

Of course there's room for improvement:

better management for errors/problems
adaption to netdev is missing and it's still unclear to me how to achieve this
add other LoRaWAN functionnalities: more parameters (RX2 delays, etc), link check and LoRaWAN compliance tests

Issues/PRs references

refs #7331

This PR is based on ~~#7592~~ and ~~#8261~~

PeterKietzmann · 2017-12-14T10:22:22Z

Is this an alternative to #6645?

aabadie · 2017-12-14T11:32:00Z

Is this an alternative to #6645?

Not really. #6645 doesn't use the semtech loramac as a package but reimplement everything in gnrc. It's also totally outdated (still using netdev2, etc).

The PR here doesn't provide the netdev/netif adaption but this can be done in follow-up PRs.

aabadie · 2017-12-14T12:37:58Z

I also forgot to mention that this PR is an adaption of an initial work from @jia200x. So most of the credits go to him.

miri64 · 2017-12-14T13:19:49Z

I also forgot to mention that this PR is an adaption of an initial work from @jia200x. So most of the credits go to him.

Why doesn't the commit authorship then? ;-)

aabadie · 2017-12-14T13:38:34Z

Why doesn't the commit authorship then? ;-)

Lot's of refactoring adaption so it was not easily possible. But I kept the credits in the files headers.

miri64 · 2017-12-14T13:42:08Z

As long as @jia200x is fine with that, I don't mind.

aabadie · 2017-12-14T16:55:20Z

@kaspar030 @smlng any idea why Murdock is failing here ? because locally it builds fine for me. I'm wondering if some git caching could be involved.

dylad · 2017-12-14T20:05:05Z

Is there a way to test this without setting up the whole TTN stuff ? Can we test it with Semtech tools on a gateway ?

aabadie · 2017-12-14T21:23:15Z

Is there a way to test this without setting up the whole TTN stuff ?

There are other LoRaWAN providers (Actility, LoRiot, etc) and I think you can also setup your own LoRaWAN server (see here).

Can we test it with Semtech tools on a gateway ?

Maybe yes but I have to admit that I don't know how. Looking at loraserver.io, they also provide a LoRa gateway bridge application that can run on the gateway itself. That looks interesting.

aabadie · 2017-12-15T09:10:14Z

any idea why Murdock is failing here ?

I found the problem, it was in the package Makefile. Now there are only platform related issues remaining.

aabadie · 2017-12-15T10:18:07Z

Ci is now fixed ! I had to blacklist the MSP430 based boards because of the following issue:

semtech_loramac_crypto.a(aes.o):(.debug_loc+0x596): relocation truncated to fit: R_MSP430_16_BYTE against `no symbol'
collect2: ld a retourné 1 code d'état d'exécution

Not sure how and if it can be fixed.

aabadie · 2018-01-18T08:11:03Z

comments addressed @kaspar030, directly squashed and rebased. @miri64 is it ok for you ?

aabadie · 2018-01-18T08:21:22Z

pkg/semtech-loramac/Makefile.loramac_mac

@@ -0,0 +1,5 @@
+MODULE := semtech_loramac_mac
+
+CFLAGS += -Wno-missing-field-initializers -Wno-unused-parameter -Wno-sign-compare


I just moved this CFLAGS rules here and here so that it's not applied globally to the package module

aabadie · 2018-01-18T08:24:38Z

07aac76 is improving the CFLAGS use in the package modules

kaspar030 · 2018-01-18T13:27:46Z

There's a typo in one of the commit messages (remane). Otherwise the code looks good now.

smlng · 2018-01-18T13:29:50Z

typo in one of the commit messages (remane)

@aabadie please fix on short notice, @kYc0o had pointed that out too, see above.

aabadie · 2018-01-18T14:33:35Z

There's a typo in one of the commit messages

Damn, I fixed it yesterday but on another personal computer and this morning I forced a push from my Inria laptop... overriding the change.
Btw, this is fixed now. @miri64 do you ACK?

no response in time to get this into release 2018.01.

kYc0o · 2018-01-18T16:44:37Z

no response in time to get this into release 2018.01.

Anyways, congrats @aabadie !

jia200x · 2018-01-18T16:46:29Z

Congratz @aabadie!!

dylad · 2018-01-18T16:49:03Z

Wonderful job @aabadie !

aabadie · 2018-01-18T16:50:00Z

Wow !! This really makes my day :D
And many thanks to all people involved, mainly @jia200x and @fjmolinas for the initial work, but also reviewers!

fjmolinas · 2018-01-18T17:28:51Z

Congrats @aabadie !!

miri64 · 2018-01-18T17:36:43Z

Congrats @aabadie. But

no response in time to get this into release 2018.01.

I also had no time to check this... It leaves a bitter taste that my concerns about the architecture are just dismissed like that, just so that a (because of that potentially unstable) feature makes it into the release. As far as I can see my concerns are addressed, but I'm still not sure they actually are...

aabadie · 2018-01-19T08:40:04Z

It leaves a bitter taste that my concerns about the architecture are just dismissed like that

Don't worry, your concerns are WIP in another branch. I'll continue to work on that.

a [...] potentially unstable feature makes it into the release

I understand your concerns but I would rather see it as a first step. Maybe we can just mention in the release notes that this is an experimental integration of LoRaWAN into RIOT ?
Also notice that the actual version of the package points to the latest release of Loramc-node and it'll have to be adapted to the coming one (that should be released soon).

miri64 · 2018-01-19T08:52:11Z

I understand your concerns but I would rather see it as a first step. Maybe we can just mention in the release notes that this is an experimental integration of LoRaWAN into RIOT ?

Agreed.

aabadie added Area: LoRa Area: LoRa radio support Area: network Area: Networking Type: new feature The issue requests / The PR implemements a new feature for RIOT Area: pkg Area: External package ports labels Dec 14, 2017

aabadie force-pushed the pkg_semtech-loramac branch 2 times, most recently from 55a595c to 83d5d4f Compare December 14, 2017 12:54

aabadie force-pushed the pkg_semtech-loramac branch 3 times, most recently from fbd88e2 to 771914f Compare December 14, 2017 15:19

aabadie added the CI: ready for build If set, CI server will compile all applications for all available boards for the labeled PR label Dec 14, 2017

aabadie force-pushed the pkg_semtech-loramac branch 2 times, most recently from f2c7bee to e89cf1b Compare December 14, 2017 15:35

aabadie requested a review from dylad December 14, 2017 16:40

aabadie force-pushed the pkg_semtech-loramac branch from e89cf1b to 1b65657 Compare December 14, 2017 16:45

aabadie force-pushed the pkg_semtech-loramac branch 4 times, most recently from df8a2bb to f52d27a Compare December 15, 2017 09:09

aabadie force-pushed the pkg_semtech-loramac branch from f52d27a to 03f8a58 Compare December 15, 2017 09:23

aabadie force-pushed the pkg_semtech-loramac branch from 30fcaa7 to 85df39b Compare January 18, 2018 08:19

aabadie commented Jan 18, 2018

View reviewed changes

aabadie force-pushed the pkg_semtech-loramac branch from 07aac76 to f5957f7 Compare January 18, 2018 08:24

aabadie force-pushed the pkg_semtech-loramac branch 4 times, most recently from d2ee3ee to ae2c023 Compare January 18, 2018 09:40

kaspar030 approved these changes Jan 18, 2018

View reviewed changes

aabadie added 3 commits January 18, 2018 15:30

sys/net/loramac: rename default device class define

f9eabb5

pkg: add Semtech LoRaMAC package

96e059d

tests/semtech-loramac: add test application

e3ff143

aabadie force-pushed the pkg_semtech-loramac branch from ae2c023 to e3ff143 Compare January 18, 2018 14:31

smlng merged commit 0352681 into RIOT-OS:master Jan 18, 2018

This was referenced Jan 19, 2018

RFC: LoRaWAN in RIOT #7331

Closed

pkg: add LoRaMAC-in-C package (WIP) #7640

Closed

aabadie deleted the pkg_semtech-loramac branch February 26, 2018 10:31

		@@ -0,0 +1,5 @@
		MODULE := semtech_loramac_mac

		CFLAGS += -Wno-missing-field-initializers -Wno-unused-parameter -Wno-sign-compare

pkg: Semtech LoRaMAC #8264

pkg: Semtech LoRaMAC #8264

Uh oh!

Conversation

aabadie commented Dec 14, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Contribution description

Issues/PRs references

Uh oh!

PeterKietzmann commented Dec 14, 2017

Uh oh!

aabadie commented Dec 14, 2017

Uh oh!

aabadie commented Dec 14, 2017

Uh oh!

miri64 commented Dec 14, 2017

Uh oh!

aabadie commented Dec 14, 2017

Uh oh!

miri64 commented Dec 14, 2017

Uh oh!

aabadie commented Dec 14, 2017

Uh oh!

dylad commented Dec 14, 2017

Uh oh!

aabadie commented Dec 14, 2017

Uh oh!

aabadie commented Dec 15, 2017

Uh oh!

aabadie commented Dec 15, 2017

Uh oh!

aabadie commented Jan 18, 2018

Uh oh!

aabadie Jan 18, 2018

Choose a reason for hiding this comment

Uh oh!

aabadie commented Jan 18, 2018

Uh oh!

kaspar030 commented Jan 18, 2018

Uh oh!

smlng commented Jan 18, 2018

Uh oh!

aabadie commented Jan 18, 2018

Uh oh!

kYc0o commented Jan 18, 2018

Uh oh!

jia200x commented Jan 18, 2018

Uh oh!

dylad commented Jan 18, 2018

Uh oh!

aabadie commented Jan 18, 2018

Uh oh!

fjmolinas commented Jan 18, 2018

Uh oh!

miri64 commented Jan 18, 2018 • edited by aabadie Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

aabadie commented Jan 19, 2018

Uh oh!

miri64 commented Jan 19, 2018

Uh oh!

Uh oh!

aabadie commented Dec 14, 2017 •

edited

Loading

miri64 commented Jan 18, 2018 •

edited by aabadie

Loading