Even though a sync loss between a sensor device and its PAN coordinator is usually something temporal, it can become permanent if the PAN coordinator is no longer reachable.

If this is the case, the sensor device transitions into what is known as an orphan state. In this state, the sensor device enters a low power mode, and only periodically awakes to attempt to reconnect to its PAN coordinator. If the PAN coordinator is in fact no longer reachable because it has failed or it has been removed, the sensor will indefinitely (or until its power source fails) attempt to reconnect to it. Therefore, if the sensor is to remain active in the network, the implementation of a PAN coordinator switching procedure is necessary.

This first switching scenario is rather simple, as the most basic implementation only requires a modification to the application layer of the sensor example. Taking this into account, the aspects that must be considered when implementing the coordinator switching procedure at the application layer of the sensor device are:

• The condition under which the procedure is triggered: How much time should the sensor spend trying to rejoin its original PAN coordinator.
• The implementation of a network disassociation process which does not rely on an acknowledge from the PAN coordinator. In the existing implementation of the disassociation process, either the sensor or the coordinator initiates the process, and the process must be acknowledged by the other side.
• The periodical use of active or passive scans (depending on the TI 15.4 mode of operation) to search for a replacement PAN coordinator.