When a node (SM = 8) joins a sleeping network, it receives sync information from the sleep coordinator (SM=7) and immediately goes into sleep mode. If messages addressed to the sleep coordinator are sent whenever the node wakes (CTS polled), we find that up to FOUR sleep cycles might by needed before the messages are received at the coordinator. All the messages appear there at the same time and all the transmit status frames are produced by the source node, having given no responses in the earlier “wake” periods when the messages were applied to its input. Our interpretation of the status frames is that they indicate that the delay in the arrival of the messages at the sleep coordinator was due to the route discovery process. All subsequent messages pass quickly through the network and the transmit status messages at the source node show “success”. Has anyone else found this? Is there anyway to reduce the delay for route discovery?