Understanding call flow from PSTN to endpoints

[esa_fresa]

I'm slightly confused on how calls flow from the PSTN to, say, an end user's phone. I'm hoping someone can point me in the right direction or explain this.

Cell>PSTN>my ITSP>Voice Gateway>UCM>End phone (This is a common configuration, right?)

My understanding is that if I make a call to a DID number from my cellphone, the call is routed through the PSTN and eventually to my ITSP. The ITSP will have an address (IP address or number? how do they know where to send it? Can there be a failover address?) to route the call to. After reaching the voice gateway, it will match an inbound and outbound dial peer (does it need to match a dial peer? what happens if there isn't a dial peer configured?). The voice gateway's outbound dialpeer will then, in some configurations at least, send the call to UCM. What CSS does the call use when it's received from the voice gateway?

These questions have been bugging me for a while and I'm having trouble putting together all the snippets of answers I find online into a complete picture. Any help will be greatly appreciated. Thank you!

[Jaime Valencia · Accepted solution]

Your ITSP will map the number to your premises depending on what you have, TDM or SIP, if there can be a failover for inbound calls or not, that's up to the telco.

Assuming you use SIP or H.323, yes, if you don't have dial-peers to route the calls, you simply have no configuration to route the calls, and no calls will work. for inbound call leg you might match the default dial-peer, but without dial-peers with destination-pattern, there's no routing to CUCM.

You route the calls by using the dial-peers, you can find plenty of documentation about dial-peers how to configure, matching, etc. on cisco.com

If MGCP, no dial-peers, CUCM takes care of all the routing.

You use the inbound CSS from the GW configuration in CUCM.