Hi,
I see this all the time. Firstly I would suggest increasing the 'missed snmp poll threshold' within the settings for that group.
I've also setup the alert to only send when the device is down for 10 mins so avoid alerts for reboots or intermittant network issues.
Add a second condition for 'minutes down threshold' to greater than 10 (or whatever is appropriate for you), as below.
Hope that helps.