Actor Priority

Unreal Engine does not guarantee that all actors replicate during a network update. This is due to the restricted nature of networking resources. A primary limiting factor is a connection's bandwidth. The bandwidth of a connection is the maximum data transfer capacity of that connection. A connection becomes saturated when the connection exceeds its capacity. When a connection is saturated, Unreal Engine’s replication system uses a load-balancing technique that assigns all actors a numerical priority. This priority gives each actor a fair share of the available network bandwidth resources based on how important the actor is to gameplay. Actors with higher relative priority are those that are more important to replicate, so they receive more bandwidth for replication.

Obtain an Actor’s Priority

Each actor has a floating-point AActor::NetPriority property. Higher NetPriority corresponds to more bandwidth for the current actor relative to others. For example, an actor with NetPriority == 2.0 is given more resources than an actor with NetPriority == 1.0. The only thing that matters with priorities is their ratio; you cannot improve Unreal Engine’s network performance by scaling all actors' net priorities.

As a baseline, here are the initial values used by some common Unreal Engine classes:

NetPriority is a baseline used for low bandwidth or saturated connections. AActor::GetNetPriority determines an actor’s current priority based on a number of factors, including base NetPriority, distance to viewer, and time since last replicated.

Retrieve Actor’s Current Priority

The network driver determines an actor’s current priority for replication to a particular connection with a call to GetNetPriority. This is handled automatically by the network driver.

Override Actor Relevancy

You can customize actor priority by overriding the virtual function GetNetPriority in your AActor-derived class as well as changing the base network priority with NetPriority.

How Priority is Determined

An actor's current network priority is calculated based on the time since the actor was last replicated, as well as a variety of additional factors, to obtain a floating-point priority.

Parameters

Actor network priority is based on the following input parameters:

Priority Logic

Most of the work of AActor::GetNetPriority is done to compute a multiplicative factor for the constant AActor::NetPriority based on distance from and sightlines to the Viewer as well as the time since the current actor last replicated.

Network priority is determined as follows:

• If both of the following conditions hold, then the current actor uses its owner's network priority.

  • The current actor has an owner.

  • The current actor is set to use its owner's network relevancy.

• If at least one of the following conditions holds, then the current actor's network priority increases.

  • The current actor is the current connection's pawn.

  • The current connection's pawn is the instigator of some action.

• If neither of the previous two points are the case, then distance based calculations are performed to determine the current actor's network priority:

  • If the current actor is in front of the viewer, then priority decreases inversely proportional to set distances.

    • If the distance between the current actor and the viewer is greater than CLOSEPROXIMITY but less than NEARSIGHTTHRESHOLD, then the priority is multiplied by 0.2.

    • If the distance between the current actor and the view is greater than NEARSIGHTTHRESHOLD, then the priority is multiplied by 0.4.

  • If the distance between the current actor and the viewer is less than FARSIGHTTHRESHOLD and the viewer is looking at the current actor, then the priority is multiplied by 2.0.

  • If the distance between the current actor and the view is greater than MEDSIGHTTHRESHOLD, then the priority is multiplied by 0.4.

Priority Reference

Functions

Properties