This read-write Unsigned32, ranging 1 to 1000, sets the number of consecutive delta_ts (denoted 'n' in the description) required to transition the availability state, used in computing availability and unavailability. An administrator adjusts this to control how persistent a loss event must be before it's counted as a real unavailability period rather than a brief blip. For example, setting esaProbeCOSConfigSoamPmExtConDeltaTsForAvail to 10 means ten consecutive delta_ts of excessive loss must occur before the link is marked unavailable, filtering out single transient spikes.