Hello.
On 09/16/2017 09:08 AM, StarNight wrote:
2017-09-15 20:26 GMT+08:00 Stefan Schmidt <stefan@xxxxxxxxxxxxxxx>:
Hello.
On 09/10/2017 11:30 AM, Jian-Hong Pan wrote:
LoRa is an implementation of LPWPAN.
Chirp Spread Spectrum (CSS) modulation, which is used by LoRa has been
specified by IEEE 802.15.4a.
This driver implements IEEE 802.15.4 mac over LoRa physical layer not
LoRaWAN.
Please see my architectural comments on how to fit in LoRa transceiver in my
other mail. I have a very specific question on this driver and its use with
the mac802154 subsystem though.
I had a look at the first datasheet I could find for it
(http://www.semtech.com/images/datasheet/sx1276_77_78_79.pdf) and stumbled
over this sentence:
"Standard GFSK, FSK, OOK, and GMSK modulation is also provided
to allow compatibility with existing systems or standards such
as wireless MBUS and IEEE 802.15.4g."
Does this mean the transceiver can run in a mode where it would be
compatible with an plain (no LoRA support) IEEE 802.15.4g transceiver on the
same frequency band, subHGZ here?
It is the datasheet that I refer to when I write the driver.
OK
This series of chips (Semtech SX1276/77/78/79) have two operating modes:
LoRa mode and FSK/OOK mode
Its registers may be in different meaning and usage with in different operating
mode.
The overall registers comparing list can be found at "Table 41 Registers
Summary" of the datasheet.
However, the operating frequency of the chip is an issue.
It shuold meet regional regulatory requirements.
The meeted frequency could be found in the series of chips.
Sure, you would need to make sure that the 802.15.4g compliant
transceiver on the other side operates actually in the same frequencies
(based on regional regulations).
What I try to understand is if we can setup this device in a 802.15.4
compliant phy mode. In that case I have no problem with it landing in
the ieee802154 subsystem.
If this is not possible and we could only run mac802514 over them but
they would only work against the same device I do not think it should go
into ieee802154. In that case we should rather work out what parts of
802.15.4 can be shared with a full LoRa subsystem and work on that.
While we do not have support for the 15.4g extension (MTU size, etc) in our
stack right now this would be a very interesting first test for the driver.
If this would be the case.
According to the datasheat, the FIFO buffer size of this series of chips is:
- 256 bytes in LoRa mode (Chapter 4.1.2.3. LoRa Mode FIFO Data Buffer)
- 64 bytes in FSK/OOK mode (Chapter 4.2.10. FIFO)
How did you test your driver so far? Making sure that the device gets
detected and setup correctly I assume. Did you also do full transmission and
receive tests with this transceiver and mac802154? If yes, was it the same
hardware on both sides or different?
I use a Raspberry Pi as the board and Archlinux ARM as the OS.
I make two SX1278 transceivers as SPI slaves connect to the Raspberry Pi,
because the Raspberry Pi has two SPI chip selector addresses.
Then, prepare the Device Tree overlap and install it before modprobe the driver
as a module.
There wil be two wpan interfaces in the system and show them with "ip addr".
I refer to the shell script in the "Setup a 6LoWPAN test network" section of
http://wpan.cakelab.org/ to setup the wpan interfaces.
iwpan dev wpan${i} set pan_id $panid
ip link add link wpan${i} name lowpan${i} type lowpan
ip link set wpan${i} up
ip link set lowpan${i} up
Two lowpan interfaces will be added and brought up.
Use "ip addr" again and will see the IPv6 addresses of the lowpan interfaces.
I also use an echo server and a client written by myself in Python 3 to test the
UDP communication between the two interfaces.
console of server side:
server.py <listening IPv6 address> <listening port>
console of client side:
client.py <src IPv6 address> <dest IPv6 address> <dest port> <data string>
Client will send the data string to server.
Server will capitalize the recieved data string from the client and send back to
the client.
Client will receive the capitalized data string and print it out.
By the way, those could be found in the project on my GitHub:
https://github.com/starnight/LoRa
Thanks for the explanation.
You tested full communication up to UDP over 6LoWPAN adaption, but only
between the same devices. Did these devices run in LoRa or in FSK/OOK mode?
regards
Stefan Schmidt
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
