Wanesy™ Management Center (End Of Life)

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Network Delays configuration ======

Network Delay configuration helps target the RX1 window when sending TX messages to an end device. 
It also allows adaptation to networks with high latency.

===== Configuration =====
==== Menu ====
From the backhaul configuration menu of each gateway, the network delay uplink and downlink can be configured. The default configuration is 350ms. 

Path to the menu: Managements -> gateways-> <gateway_EUI> -> configurations -> backhaul

{{:wiki3:network_delays_wmc3.0.png|}}

This value is an indication of the network delays that the gateway has towards and from the LNS server.
==== Samples of configuration ====
<note important>The latency of the gateway can be evaluated from the ping (delay up + delay down) KPI. 

On a fast network, to target the RX1 window with maximum anticipation, the following configuration must be set:
|Network max. delay up* | |Network max. delay down* |
|0| |400|

On a slow network, to target the RX2 window with maximum anticipation, the following configuration must be set:
|network max. delay up* | |network max. delay down* |
|0| |1400| 


</note>

===== Behavior =====

When an end-device sends a LoRa packet, each gateway that receives it will forward it to the LNS server. The LNS server will wait 350ms after having received the first packet, in order to aggregate all the copies of that packet sent by each individual gateway.

If a push is configured for the cluster, the LNS server will wait 150 ms after the push, in order to offer the client the opportunity to reply to the message in the following RX window.

It will then try to see which window is best to send the TX message to the end device. To do so, it will add the //network delay down// to the current time, and see if this fits, with an additional 50ms margin. If this allows to reach the RX1 window delay (usually one second), then it will send it soon. If this doesn't allow to reach the RX1 window, it will target the RX2 window (usually two seconds).

++++ RX window decision method |

<code>
T0_lns = now - N_du
T_est = now + T_agg + T_opp + N_dd
Decision = T_est - T0_lns = N_du + T_agg + T_opp + N_dd < 1000 ?
T_tx = now + T_win - N_dd
</code>

with:

   * ''T0'': packet transmission date by gateway.
   * ''T0_lns'': packet transmission date by gateway estimated by lns.
   * ''now'': packet reception date by LNS.
   * ''N_du'': //Network max. delay up//: the setting on the dashboard (350ms by default).
   * ''T_est'': estimated earliest timestamp for transmission
   * ''T_agg'': Aggregation time: the LNS waits 350ms after the first packet.
   * ''T_opp'': Opportunity time: the LNS waits 150ms after the first packet.
   * ''N_dd'': //Network max. delay down//: the setting on the dashboard (350ms by default).
   * ''Decision'': whether the RX1 window is reachable, 1000ms being the default RX1 value (it can be changed).
   * ''T_tx'': the time at which the LNS will send the packet to the gateway.
   * ''T_win'': the time for the target RX window (1000ms by default for RX1).
++++ 
++++ Example |

Inputs:

   * Delay up: ''350ms'' (configured).
   * Delay up: ''196ms'' (actual).
   * Delay down: ''350ms'' (configured).
   * Delay down: ''212ms'' (actual).
   * Packet sent at ''10:00:00.000'' by gateway (//actual// ''T0'', but not the value computed by LNS).
   * Packet received at ''10:00:00.196'' by LNS (''now'').

Computation:

   * ''T0_lns'' assumed is ''09:59:59.846'' by LNS (''now'' minus 350ms ''Ndu'').
   * ''T_est'' without push is ''now'' + 350 + 350 = ''10:00:00.896''
   * ''T_est'' with push is ''now'' + 350 + 150 + 350 = ''10:00:01.046''
   * Decision: ''T_est'' is greater than 1000ms after ''T0'', so RX1 isn't reachable.
   * Decision: ''T_est'' is less than 2000ms after ''T0'', so RX2 is reachable.
   * ''T_tx'' is ''T0_lns'' + 2000 - 350 = ''10:00:01.496'' (time at which the LNS sends the TX message)

Behavior:

   * The LNS sends the packet at ''10:00:01.496'' (''T_tx'').
   * The packet forwarder receives it at ''10:00:01.708'' (''T_tx+212ms'').
   * The LoRa frame is sent at ''10:00:02.000'' when the RX2 window opens. Actual timing is done with ''tmst'' (SX1301 clock counter).
++++
++++ Notes |

   * Due to the packet forwarder processing time, a delay of 30ms to 50ms occurs between the IP packet reception and the TX LoRa frame going out of the radio.
   * The computation algorithm above is only for the RX window **decision**. The actual **scheduling** of the packet is done using the ''tmst'' field of the packet forwarder, and adding the corresponding SX1301 ticks to reach the RX1 or RX2 windows.
   * The LNS will keep the TX messages for as long as possible, and send them at the last moment. This is by design, to allow less load on the packet forwarders. Indeed, if the TX messages were sent as soon as possible, the gateways might get a lot of TX messages queued, which would be problematic on embedded devices.
++++

===== Actual behaviors =====

Several situations may arise:

   * The backhaul is correct (~200ms RTT), with default configuration everything works (nominal case with push).
   * The backhaul is very good (<100ms RTT), with specific configuration everything works (RX1 target with push).
   * The backhaul is not good (~800ms RTT), with default configuration the TX messages aren't transmitted.
   * The backhaul is not good (~800ms RTT), with specific configuration everything works (RX2 target with push).
   * The backhaul is very bad (>1000ms RTT), and the TX messages aren't transmitted properly.
   * The backhaul is correct (~200ms RTT), everything works (nominal case without push).

==== Correct backhaul (~200ms RTT) ====

This is the nominal case. The LNS will always target the RX2. The configured delays on the dashboard are the default values (350ms).

{{:images:326lns_delay_200ms.png|}}

==== Very good backhaul (<100ms RTT) ====

The LNS will target the RX1 depending on the configured network delay. 
The configured network delays on the dashboard are set to 0ms and 400ms for //Network max. delay up// and //Network max. delay down//.

{{:images:326lns_delay_100ms.png|}}

In this example (0ms/400ms), the RX1 is reachable. 
If the //Network max. delay down// had been configured to 100ms, the RX1 window targeted may have been reachable without margin.
With default configuration (350ms/350ms), RX2 window is targeted.

==== Not good backhaul (~800ms RTT) ====

In this case the RX1 window is not reachable. The RX2 window is also missed, unless the network delay is adjusted.

//Network max. delay down// is set with the default configuration:

{{:images:326lns_delay_800ms_ko.png|}}

In this case the RX2 is missed and the TX packet is dropped. The solution is to change the //Network max. delay up// and //Network max. delay down //to a value of 0ms and 850ms:

{{:images:326lns_delay_800ms.png|}}

The RX1 is not reachable, but the RX2 is, thanks to the updated configuration.

==== Very bad backhaul (>1000ms) ====

In this case the RX1 and RX2 windows are never reachable. Indeed, with a RTT of 1200ms, the delay is:

<code>1200 + 350 + 50 + 1200 = 2800ms</code>

The RX2 window has to be changed in order for this to work. On each end device, and on the dashboard configuration of the end device, the RX1 window delay has to be adjusted to a greater value. For example, 2000 (for RX2 target) or 3000ms (for RX1 target) is fine.

Indeed, the delay calculated above is 2800ms, so if RX1 delay is 3000ms, it is reachable.

This is possible at the cost of provisioning every end device on the network so that it can work with high latencies. This can be useful with satellite backhaul, for example.

==== Correct backhaul (~200ms RTT) without push ====

This is the nominal case. The LNS will always target the RX2. The configured delays on the dashboard are the default values (350ms).

{{:images:326lns_delay_200ms_adr.png|}}


===== Edge case: idle cellular radios =====

Mobile network operators, to save bandwidth and power, usually instruct the mobile equipments to reduce the radio state (RRC layer).

For example, in UMTS, there are mainly three states:

   * idle: radio is OFF, but the context is saved both in mobile and network.
   * FACH: radio is in a low latency but low payload active state. Only very limited payloads can be sent.
   * DCH: radio is active and transmitting data.

The behavior is that when there is network activity, the mobile is in DCH. When there is no activity, after a delay, the network will place the mobile in FACH. This allows the mobile to quickly recover an active connection (DCH) and to send very limited payloads. When there is still no activity while in FACH, the network will place the mobile in idle state.

When the mobile is idle and wants to send data, it will have to establish a new radio connection to the mobile network, which will take some time. Once in this state (DCH), the mobile will have a low latency.

So basically, if there is no traffic, the first connection may take up to 1000ms. Subsequent connections are faster, like 100 or 200ms.

==== Solutions ====

One way to circumvent this is to run a periodic ping on the gateway, this will keep the network connectivity in DCH state with a low latency. The downside is that this generates a lot of data traffic, that may be charged a premium by the mobile ISP.

Another way to work around this is to configure a 1000ms //Network max. delay up// and a 200ms //Network max. delay down//. This will allow the LNS to finely compute the delays, and be able to reach the RX2 window.
Wanesy™ Management Center (End Of Life)

User Tools

Site Tools

Page Tools
