drivers/at86rf215: implement MR-OFDM by benpicco · Pull Request #14010 · RIOT-OS/RIOT

benpicco · 2020-05-04T01:38:23Z

Contribution description

This adds MR-OFDM modulation to the at86rf215 driver.

This also adds new commands to ifconfig to configure the new modulation parameters.

Testing procedure

Switch to the new PHY mode with ifconfig 7 set phy_mode mr-ofdm. Note that the L2-PDU has changed to 2022 bytes (802.15.4g can send 2047 byte frames).

2020-05-04 03:41:16,173 # Iface  7  HWaddr: 01:57  Channel: 0  NID: 0x23  PHY: MR-OFDM 
2020-05-04 03:41:16,180 #            Option: 1  MCS: 0 (BPSK, rate 1/2, 4x frequency repetition) 
2020-05-04 03:41:16,184 #           Long HWaddr: 1A:F9:8F:5F:30:5D:05:8E 
2020-05-04 03:41:16,191 #            TX-Power: 0dBm  State: IDLE  max. Retrans.: 3  CSMA Retries: 4 
2020-05-04 03:41:16,197 #           AUTOACK  ACK_REQ  CSMA  L2-PDU:2022 MTU:1280  HL:64  RTR  
2020-05-04 03:41:16,199 #           6LO  IPHC  
2020-05-04 03:41:16,202 #           Source address length: 8
2020-05-04 03:41:16,204 #           Link type: wireless
2020-05-04 03:41:16,210 #           inet6 addr: fe80::18f9:8f5f:305d:58e  scope: link  VAL
2020-05-04 03:41:16,213 #           inet6 group: ff02::2
2020-05-04 03:41:16,216 #           inet6 group: ff02::1
2020-05-04 03:41:16,219 #           inet6 group: ff02::1:ff5d:58e
2020-05-04 03:41:16,220 #           
2020-05-04 03:41:16,223 #           Statistics for Layer 2
2020-05-04 03:41:16,226 #             RX packets 0  bytes 0
2020-05-04 03:41:16,230 #             TX packets 4 (Multicast: 4)  bytes 172
2020-05-04 03:41:16,233 #             TX succeeded 4 errors 0
2020-05-04 03:41:16,236 #           Statistics for IPv6
2020-05-04 03:41:16,239 #             RX packets 0  bytes 0
2020-05-04 03:41:16,244 #             TX packets 4 (Multicast: 4)  bytes 256
2020-05-04 03:41:16,247 #             TX succeeded 4 errors 0
2020-05-04 03:41:16,247 # 
2020-05-04 03:41:30,034 #  ping6 fe80::2068:3123:59f5:d23a%7
2020-05-04 03:41:30,053 # 12 bytes from fe80::2068:3123:59f5:d23a%7: icmp_seq=0 ttl=64 rssi=-41 dBm time=12.051 ms
2020-05-04 03:41:31,054 # 12 bytes from fe80::2068:3123:59f5:d23a%7: icmp_seq=1 ttl=64 rssi=-40 dBm time=12.051 ms
2020-05-04 03:41:32,054 # 12 bytes from fe80::2068:3123:59f5:d23a%7: icmp_seq=2 ttl=64 rssi=-41 dBm time=12.051 ms

You should be able to configure the OFDM option (ifconfig 7 set option [1…4]) as well as the Modulation and Coding Scheme (ifconfig 7 set mcs [0…6]):

2020-05-04 03:49:00,938 #  ifconfig 7 set mcs 2
2020-05-04 03:49:04,010 #  ifconfig 7 set option 4
2020-05-04 03:52:49,320 #  ping6 fe80::2068:3123:59f5:d23a%7
2020-05-04 03:52:49,348 # 12 bytes from fe80::2068:3123:59f5:d23a%7: icmp_seq=0 ttl=64 rssi=-42 dBm time=19.828 ms
2020-05-04 03:52:50,348 # 12 bytes from fe80::2068:3123:59f5:d23a%7: icmp_seq=1 ttl=64 rssi=-42 dBm time=19.828 ms
2020-05-04 03:52:51,348 # 12 bytes from fe80::2068:3123:59f5:d23a%7: icmp_seq=2 ttl=64 rssi=-42 dBm time=19.827 ms
2020-05-04 03:52:51,348 # 
2020-05-04 03:52:51,353 # --- fe80::2068:3123:59f5:d23a PING statistics ---
2020-05-04 03:52:51,358 # 3 packets transmitted, 3 packets received, 0% packet loss
2020-05-04 03:52:51,362 # round-trip min/avg/max = 19.827/19.827/19.828 ms

Nodes can talk to each other if they use the same option, but different Modulation & Coding Schemes.

Issues/PRs references

split off #12128
requires #13979 to send full 2k frames

benpicco · 2020-05-25T18:47:50Z

@tinstructor I heard you like MR-OFDM, want to give this a try? 😉

tinstructor · 2020-05-25T18:53:58Z

@benpicco I'll run through your testing procedure quickly tomorrow morning. Mind you, due to Corona lockdown measures I currently don't have access to a signal analyzer should things need troubleshooting.

tinstructor · 2020-05-26T09:48:36Z

@benpicco Remind me again, which example application do you typically use to test drivers like this?

benpicco · 2020-05-26T09:49:45Z

examples/gnrc_networking should work

tinstructor · 2020-05-26T10:41:35Z

@benpicco I've quickly done some testing with two OpenMote-B nodes and it seems to work just fine. Of course you'd probably want to run a few hundred pings for each interface / each combination of option + MCS to be sure

fjmolinas

Some nitpicks, otherwise code looks. Will test against different configurations.

sys/shell/commands/sc_gnrc_netif.c

drivers/include/at86rf215.h

drivers/at86rf215/include/at86rf215_internal.h

drivers/at86rf215/at86rf215_ofdm.c

fjmolinas · 2020-06-02T13:41:18Z

@benpicco I've quickly done some testing with two OpenMote-B nodes and it seems to work just fine. Of course you'd probably want to run a few hundred pings for each interface / each combination of option + MCS to be sure

Your application that looped over the different configurations would be very usefull here :)

benpicco · 2020-06-02T13:43:51Z

Your application that looped over the different configurations would be very usefull here :)

That one?
Probably needs an update to match the upstream API now.

fjmolinas · 2020-06-02T15:09:06Z

Your application that looped over the different configurations would be very usefull here :)

That one?
Probably needs an update to match the upstream API now.

Did some testing mixing up configurations and it all seems to work. trying to see if I can cover them all automaticaly.

fjmolinas · 2020-06-02T19:42:10Z

Your application that looped over the different configurations would be very usefull here :)

That one?
Probably needs an update to match the upstream API now.

Did some testing mixing up configurations and it all seems to work. trying to see if I can cover them all automatically.

I used #13612 to script testing over all possible combinations. Worked for all pining both ways with -c 100 -s 100 -i 100. For the options/mcs allowing higher datarates I was able to ping at much higher rates.

fjmolinas · 2020-06-02T19:59:02Z

@benpicco would you mind uncrustifying the files?

benpicco · 2020-06-02T20:40:12Z

I removed some of the crust, although IMHO that added to the taste.
Or do you mean also uncrustify the existing code?

fjmolinas · 2020-06-03T06:23:42Z

I removed some of the crust, although IMHO that added to the taste.
Or do you mean also uncrustify the existing code?

The new code, I noticed that uncrustify -c uncrustify-riot.cfg drivers/at86rf215/at86rf215_ofdm.c suggested many changes :)

fjmolinas · 2020-06-03T13:42:09Z

Please squash @benpicco (BTW I don't get nitified when you push a fix, so please ping me when you do so I notice :) )

fjmolinas · 2020-06-03T13:44:45Z

drivers/at86rf215/at86rf215_ofdm.c

+                        _TXDFE_SR(option) | _TXDFE_RCUT(option));
+
+    /* set channel spacing */
+    at86rf215_reg_write(dev, dev->RF->RG_CS, _channel_spacing_kHz(option) / 25);


Just noticed this 25, why is it?

Oh yeah 25Khz resolution

Can you add a comment like for o-qpsk?

/* set channel spacing with 25 kHz resolution */

Done! And I also undid some of the damage done by uncrustify in the next line.

Define options for IEEE 802.15.4g MR-OFDM as well as shell commands to set them via ifconfig.

With the added features, the driver doesn't fit onto those ATmega boards anymore.

fjmolinas

ACK

benpicco added Type: new feature The issue requests / The PR implemements a new feature for RIOT Area: drivers Area: Device drivers labels May 4, 2020

benpicco requested review from PeterKietzmann, cgundogan, haukepetersen and miri64 as code owners May 4, 2020 01:38

benpicco requested a review from fjmolinas May 4, 2020 01:38

benpicco force-pushed the at86rf215-mr-ofdm branch from 1f7841b to eba4410 Compare May 4, 2020 01:44

benpicco added the CI: ready for build If set, CI server will compile all applications for all available boards for the labeled PR label May 4, 2020

benpicco requested review from MichelRottleuthner, aabadie, leandrolanzieri and smlng as code owners May 4, 2020 02:35

benpicco requested a review from roberthartung May 15, 2020 11:23

fjmolinas reviewed Jun 2, 2020

View reviewed changes

drivers/at86rf215/at86rf215_ofdm.c Outdated Show resolved Hide resolved

fjmolinas added the Reviewed: 3-testing The PR was tested according to the maintainer guidelines label Jun 2, 2020

fjmolinas added Reviewed: 1-fundamentals The fundamentals of the PR were reviewed according to the maintainer guidelines Reviewed: 2-code-design The code design of the PR was reviewed according to the maintainer guidelines labels Jun 2, 2020

fjmolinas self-assigned this Jun 3, 2020

fjmolinas added the Reviewed: 4-code-style The adherence to coding conventions by the PR were reviewed according to the maintainer guidelines label Jun 3, 2020

fjmolinas reviewed Jun 3, 2020

View reviewed changes

benpicco added 4 commits June 3, 2020 15:58

ieee802154: add definitions & config options for MR-OFDM

25a34a1

Define options for IEEE 802.15.4g MR-OFDM as well as shell commands to set them via ifconfig.

drivers/at86rf215: implement MR-OFDM

cc5fbdf

tests/driver_at86rf215: update Makefile.ci

8a63ca5

With the added features, the driver doesn't fit onto those ATmega boards anymore.

at86rf215: remove obsolete at86rf215_switch_mode()

164e239

benpicco force-pushed the at86rf215-mr-ofdm branch from 3930a60 to 164e239 Compare June 3, 2020 13:59

fjmolinas approved these changes Jun 3, 2020

View reviewed changes

benpicco merged commit 3a1ee49 into RIOT-OS:master Jun 3, 2020

benpicco deleted the at86rf215-mr-ofdm branch June 3, 2020 14:39

miri64 added this to the Release 2020.07 milestone Jun 24, 2020

Conversation

benpicco commented May 4, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Contribution description

Testing procedure

Issues/PRs references

Uh oh!

benpicco commented May 25, 2020

Uh oh!

tinstructor commented May 25, 2020

Uh oh!

tinstructor commented May 26, 2020

Uh oh!

benpicco commented May 26, 2020

Uh oh!

tinstructor commented May 26, 2020

Uh oh!

fjmolinas left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

fjmolinas commented Jun 2, 2020

Uh oh!

benpicco commented Jun 2, 2020

Uh oh!

fjmolinas commented Jun 2, 2020

Uh oh!

fjmolinas commented Jun 2, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

fjmolinas commented Jun 2, 2020

Uh oh!

benpicco commented Jun 2, 2020

Uh oh!

fjmolinas commented Jun 3, 2020

Uh oh!

fjmolinas commented Jun 3, 2020

Uh oh!

fjmolinas Jun 3, 2020

Choose a reason for hiding this comment

Uh oh!

fjmolinas Jun 3, 2020

Choose a reason for hiding this comment

Uh oh!

fjmolinas Jun 3, 2020

Choose a reason for hiding this comment

Uh oh!

benpicco Jun 3, 2020

Choose a reason for hiding this comment

Uh oh!

fjmolinas left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

benpicco commented May 4, 2020 •

edited

Loading

fjmolinas commented Jun 2, 2020 •

edited

Loading