nes-proj/examples/platform-specific/cc26xx/very-sleepy-demo
2017-06-29 16:46:07 +02:00
..
Makefile Re-arrange examples, in particular, move platform-specific examples to examples/platform-specific 2017-05-16 17:47:57 +02:00
Makefile.target Re-arrange examples, in particular, move platform-specific examples to examples/platform-specific 2017-05-16 17:47:57 +02:00
project-conf.h Re-arrange examples, in particular, move platform-specific examples to examples/platform-specific 2017-05-16 17:47:57 +02:00
README.md Re-arrange examples, in particular, move platform-specific examples to examples/platform-specific 2017-05-16 17:47:57 +02:00
very-sleepy-demo.c Added RPL-lite, a lightweight implementation of RPL. This is the new default. 2017-06-29 16:46:07 +02:00

CC13xx/CC26xx Very Sleepy Demo

This example demonstrates a way of deploying a very low-consuming, very sleepy node. The node has two modes of operation:

  • Normal: ContikiMAC duty-cycles the radio as usual. The node is reachable.
  • Very Sleepy: Radio cycling mostly off, except when we need to perform network maintenance tasks. In this mode, the node is unreachable for most of the time.

The node will operate in RPL leaf mode. This means that it will be reachable downwards, but it will not advertise the DODAG and it will not participate in routing.

After booting, the node will enter "normal" mode.

The node exposes an OBSERVEable CoAP resource. It will notify subscribers with a new value for this resource every interval seconds. It will then stay in normal mode for duration seconds. During this time window, it will be reachable over the network in order to e.g. receive a new configuration. When this time window expires, the node will switch back to very sleepy mode. This will only happen if very sleepy mode has been enabled by setting mode=1 as per the instructions below.

When the node is duty-cycling the radio, either because it is in normal mode or because network maintenance is taking place, it will keep its green LED on thus providing an indication that it is reachable (red LED for the CC1350 tag).

A normal mode stint can be manually triggered by pressing the left button.

Requirements

To run this example you will need:

  • A border router operating with the same RDC, same channel, same radio mode (e.g. IEEE or sub-ghz), same PAN ID. Alternatively, you can use 6lbr with a suitable slip-radio.
  • The Copper (Cu) addon for Firefox

Configuration

To configure the node, send a CoAP POST message to the very_sleepy_config resource. The POST message's payload must specify at least one of:

  • mode=0|1: Send mode=1 to enable very sleepy mode, mode=0 to disable it.
  • interval=n where n is the number of seconds between two consecutive normal mode periods. This interval also dictates the OBSERVEr notification period.
  • duration=n where n is the number of seconds that the node will stay in normal mode before dropping to very sleepy mode. This value is only relevant if mode==1.

A POST request must contain at least one of the above, but they are otherwise all optional. So, for example, a POST may simply specify interval=n. To send multiple values, delimit them with &. So you can send something like mode=1&interval=60&duration=20

The current running configuration can be retrieved by sending a GET request to the same CoAP resource.

Running the example

  • Deploy your border router or 6lbr
  • Turn on the very sleepy node.
  • Fire up the Copper addon
  • Select .well-known/core and hit GET
  • Configure very sleepy operation:
    • Select the very_sleepy_config resource
    • In the Outgoing pane, type your POST payload as per the instructions above. For example, you can type: mode=1&interval=30&duration=10
    • Hit POST
  • Select the sen/readings resource and hit OBSERVE

Caveats

If you click on a resource in the Copper resources tree while you are observing a different resource, the OBSERVEr for the latter will be stopped without notifying the CoAP server. This will result in the server sending out OBSERVE notifications that will be responded to with port unreachable ICMPv6 messages. This will continue for quite a while, until the server detects that the OBSERVEr has been lost (a test currently performed once every 20 notifications). In order to prevent this from happening, hit the "Cancel" button for the OBSERVE before switching views to a different resource. This will unregister the observer.

In very sleepy mode, the radio is not truly always off. The contiki core needs to perform other periodic tasks in order to maintain network connectivity. For that reason, this example will allow the radio to turn on periodically even while in very sleepy mode. Thus, you may see that the node becomes briefly reachable every now and then. However, do not count on those periods of reachability to perform any tasks, as they will be brief and will be disrupted without warning.