SIP has been defined in the era of H323 protocol suite, where people were longing for one protocol that will encapsulate and unify the communication protocols.
Being close to PSTN / POTS telephony in the years of the VOIP, makes it ideal for telephony vendors who captivate the gap between telecom companies at prices and the cheap Internet prices. The price gap allows sellers to step in to the telecom arena, offer some discount, and earn big commission on the margins.
So, as big technology break through exhausted and the need to make profit out of lower the telecom services, SIP was accepted as the Internet Telephony Signaling. Equipment vendors were using it as common standards for telephony components. Software vendors (some of the open-source) were using it to develop software based PBXs. As the de-facto standard protocol it gained acceptance, and today telephony does use it as a base protocol for its operation.
Today, telecom companies have lowered their prices, and other telecom companies appear and offer cheap telephony around the world (about $ 20 per telephone flat-fee per month) and all of them use SIP as their standard common protocol. Anybody attempting to become player in the VOIP telephony should be an expert in this protocol. The protocol itself has signaling of:
a. "Invite" – initiating a call.
b. "Trying" – respond to the initiating call.
c. "Bad request" – initiating credential (username / password) to be transferred.
d. "Ringing".
e. "OK" – Indicate that the other party accepted the call and the two ends can talk.
As you can see the protocol closely follows the regular POTS / PSTN telephony. Extensions to the SIP protocol enable support in Instant Messaging and video conferences. Using the SIP protocol is a safe bet when it comes to creating telephony equipment. SIP has some inherit flow in it:
The first flow is not using and relying on the socket mechanism of the TCP / IP communication. Not using the socket requested that socket information has to be embedded in the SIP signaling regardless of the socket in use. This created redundancy and was prune to errors.
The second flow was not relaying on the TCP protocol of the TCP / IP suite. The TCP protocol is an abstraction mechanism that delivers reliable communication over the unreliable packets of the internet. The TCP contains specific signaling and "handshaking" to make sure that all the packets are delivered reliably and in the same order – to the upper communication layer. Almost every program running on the Internet relies on this protocol implementation to get its information reliably. Not using this protocol forces the SIP stack implementers to create similar mechanisms to the TCP in their stack. So, the stack is spread on two communication layers instead of one. This inserts big complexity to the stack manufacturer. Until these days I can see messages in the support group that are related to TCP mechanism implementation.
The above two inherit problems did not stop the SIP protocol to dominate the VOIP world. Over the Internet, there is another communication protocol related to 'presence'. Presence is now the "killer application" which vendors are looking for. The presence system is about 'buddy list' and allows the user to know in what state his list is on. Are people in the list online? Talking? Busy? The preferred method of communication in the presence world is chat which seems to be native to Internet communication. Using a presence system allows people to make much more social interactions with friends and collections and speed further the pace of living.
The presence world is dominated by another protocol called "Jabber" or XMPP. The VOIP and presence world are colliding. SIP got presence extension called "SIMPLE" while XMPP got VOIP capabilities extension called "Jingle". Jingle is supported by Google which gives it a big boost. The winning protocol will be determined by the hand-help devices. How much will they resemble phone? How much will they clash lap-tops? Apple as we know is marketing both types of devices: iPhone and iPad.
We need to see where the hand-held devices are heading. Which device will eventually dominate the Internet? The SIP protocol is and will be dominate in the Internet and VOIP arena. As the de-facto protocol it will be used as long as VOIP is being used.