I am testing a mesh network, using the NRF52832 modules, SDK 15.2 and MESH_SDK 3.0, provisioning the nodes with nRF Mesh android app and using a modified firmware for the Light Switch (Client), connected to the PC via serial usb, which sends an on-off command to two groups of nodes (Server): first, sequentially through the implementation of a timer, the client [0] automatically sends the on-off command to the first group, then again the client ([1] to the second group, and so on.
This is the test network:
Client[0] –> Nodes 8-9-10 (group1)
Client[1] -> Nodes 1-2-3-4-5-6-7 (group2)
Group1
Server 1 0x0004
Server 2 0x0008
Server 3 0x0009
Server 4 0x000A
Server 5 0x000B
Server 6 0x000C
Server 7 0x000D
Group2
Server 8 0x000E
Server 9 0x000F
Server 10 0x0010
Log file
I added the log to the network.c file, as in the following code
static bool should_relay(const network_packet_metadata_t * p_metadata)
{
/* Relay feature must be enabled */
#if EXPERIMENTAL_INSTABURST_ENABLED
if (!core_tx_instaburst_is_enabled(CORE_TX_ROLE_RELAY))
#else
if (!core_tx_adv_is_enabled(CORE_TX_ROLE_RELAY))
#endif
{
return false;
}
/* TTL must be 2 or greater */
if (p_metadata->ttl < 2)
{
return false;
}
/* Should not be directed to a unicast address on this device */
if (p_metadata->dst.type == NRF_MESH_ADDRESS_TYPE_UNICAST)
{
// mod And see all the Health messages from the other server nodes in the network
__LOG(LOG_SRC_APP, LOG_LEVEL_INFO, "Packed relayed from Node: 0x%04x with TTL:%u sequence_numb: 0x%04x dest: 0x%04x.\n", p_metadata->src,p_metadata->ttl,p_metadata->internal.sequence_number, p_metadata->dst.value);
nrf_mesh_address_t dummy_addr;
if (nrf_mesh_rx_address_get(p_metadata->dst.value, &dummy_addr))
{
return false;
}
}
/* Relay check callback function approves of the relay */
if (m_relay_check_cb != NULL)
{
if (!m_relay_check_cb(p_metadata->src, p_metadata->dst.value, p_metadata->ttl))
{
return false;
}
}
return true;
}
#endif /* MESH_FEATURE_RELAY_ENABLED */
Testing is done using the following configuration:
- TTL (client and server) modified to 4
- TX_power (client and server) 4dB
- Network and relay configuration, for the application to test, as default.
The messages sent via the generic_onoff_client model (with ACK for both groups of nodes) are all received by the nodes of the network, both from those in range and those out of the client range, but, for each on-off command sent, I riceive back the STATUS of all the interested nodes only from group1, never from group2 (see log file above).
Could it be a network congestion problem due to packets exchanged between the nodes of group2, being these 7 to receive the command instead of 3 as in group 1? Or could it be trivially that I need to increase the TTL?
If so, I would like to know how to configure network level repetitions and network level repetition intervals, in order to solve the problem, as well as relay repetitions and relay repetition intervals. I can also easily do the same with nRF Mesh?
