The pressure to develop industry standards does have one benefit. It means that service providers are continually working to improve their services to meet expectations and to set the bar for expected service higher than all the other companies providing the same service. The result is a fierce competition that drives a set of protocols that govern how your VoIP service is provided in five areas:

• Integration: At least for now, VoIP must integrate with existing telephone service (which is called the Public Switched Telephone Network, or PSTN). VoIP calls are transported via the Internet; however, unless the call is conducted from VoIP subscriber to VoIP subscriber, the call is delivered using the PSTN. The integration of VoIP and PSTN is essential to growth in the VoIP industry. Protocols govern how calls are handed from VoIP networks to PSTN and vice versa.

• Interoperability:You'd think integration and interoperability are the same. They're not. Each company that offers VoIP service also offers different types of equipment. Without some guidelines for how that equipment should interact with the equipment that other vendors provide, you could be stuck using only one vendor without the possibility of calling someone else who uses equipment from another vendor. This interoperability of equipment makes it possible for the VoIP industry to expand and will eventually lead to a single device that all manufacturers and service providers rely on. It's like the telephone that you use today. It works regardless of who made it or who provides your phone service. Protocols and standards will ensure that VoIP phones and equipment will eventually reach that point.

• Scalability: Scalability is the ability of an application or network to grow and shrink according to the amount of traffic using that application or traveling over that network. Some protocols are designed to govern the scalability of VoIP services. Past predictions of how fast the VoIP industry will grow were overly enthusiastic, calling for growth to happen much faster than it has. Even though the industry and public adoption of VoIP hasn't developed as quickly as those early estimates, it's still happening much faster than the adoption of telephone service did. Revised estimates call for 75 percent of the population to be using VoIP by 2010, a much more realistic view of how quickly VoIP will catch on. Still, without some guidelines governing scalability, the VoIP industry would falter under that kind of growth. These protocols help to ensure that VoIP infrastructure will continue to grow and meet the demands that consumers place on it.

• Quality: Quality has been the biggest sticking point for the VoIP industry. Early VoIP services were plagued with poor quality. However, the protocols that govern VoIP quality have come a long way since those first services. It is these protocols that provide guidelines for how voice should be carried over the Internet, which was originally designed to carry data. Protocols that address quality provide guidance for turning voice to digital data and then back to voice, as well as guidance for how that data is transported from one place to another once the change has been made.

• Security:The Internet is plagued with security risks. And voice conversations that travel over the Internet are not exempt from that challenge even though VoIP services have not been hard hit by security threats as of yet. It's only a matter of time before it is and VoIP is as easy a target as any other type of information technology service. Once converted to packets that are transported via the Internet, your conversation is subject to the same security threats that any other data that travels over the Internet is subject to. Hackers that capture other types of information can (and do) capture voice data, and viruses that target cell phones or e-mail can (and do) target VoIP. Even denial of service (DoS) attacks are conducted against VoIP services. Without security protocols to govern how VoIP is protected, your security risks could be much higher. Currently there's no consensus on what the best method of security is for VoIP—secure socket layers (SSL), authentication, encryption, tunneling—all of these types of security play a role, and security protocols ensure that some form of security is protecting your conversation.

In the simplest terms these standards have to work together because they dictate how computers find each other on the Internet and how information is exchanged between the computers to allow VoIP packets to flow between destinations. In addition there has to be an agreed-upon payload format so that the contents of the VoIP packet are recognized and properly decoded on both ends of the conversation. This is no easy process. It requires that the technologies surrounding VoIP be flexible enough to serve the many needs of many companies and many consumers.

As you can see, VoIP protocols and standards are needed for numerous reasons. It's not necessary for you to understand the inner workings of VoIP to be aware of some of the most frequently used protocols. You do, however, need to know what those protocols govern and how they affect you because over time they will mature and change, and as they change, your service will change as well.

Two types of protocols affect the VoIP industry: call signaling protocols and device control protocols. Call signaling protocols set up communications between two endpoints or an endpoint and gateway. Device control protocols control the functions of the device itself. Following are some of the most common call signaling and device control protocols in use in the industry.

H.323

One of the oldest of the sets of VoIP protocols in use, H.323 is a call signaling protocol that was recommended by the International Telecommunications Union (ITU) and originally adopted for use in 1996. H.323 is really a set of protocols that provides guidelines for any type of audio-visual packet communications. For example, if you attend a Web conference or video conference, chances are H.323 guides how the conference is delivered to you. The set includes protocols for VoIP, video conferencing, and other methods of data sharing.

The H.323 standard was originally developed to govern video conferencing, but it has been extended to include VoIP. The current version of H.323 came into effect in early 1998, and it governs both point-to-point communications, such as telephone calls, and multipoint communications, such as video or Web conferences.

This standard ensures interoperability between vendors by governing the endpoints of communications—terminals, gateways, gatekeepers, and multipoint control units:

• Terminals: The terminal is the endpoint that provides two-way, real-time communications—it's the ATA in VoIP. Under this protocol, all H.323-compliant terminals must support the use of channels, call signaling, real-time transport protocol (RTP), and registration administration status (RAS).
  
• Gateways: The gateway is the interface between the PSTN and the Internet. At the gateway, data is converted from analog to digital or vice versa and then delivered to the receiver, and the H.323 protocol governs how that conversion takes place.
  
• Gatekeepers: The gatekeeper is like the administrator for the gateway. It controls authorizations, addresses, signaling, and bandwidth management, among other things. H.323 outlines the parameters under which all of this administration takes place. Each gatekeeper has numerous endpoints registered with it and it performs these administrative tasks for all of those endpoints in what is called a zone.
  
• Multipoint Control Units: A multipoint control unit (MPU) is an endpoint that allows three or more terminals and gateways to participate in a multipoint conference. Using other elements of H.323, the MPU determines the common capabilities of the endpoints.

H.323 also works in conjunction with numerous other protocols, such as H.225, which describes call signaling and packetization; H.235, which provides guidelines for security; and H.245, which governs opening and closing logical channels for data control and exchange. This type of multi-protocol governance is called stacking. Protocols are stacked together to address numerous issues.

SIP

Sessions Initiation Protocol (SIP) is also a call signaling protocol that operates at the application layer for creating, modifying, and terminating VoIP connections. It was first introduced about the same time that H.323 was introduced, but wasn't recognized as a standard until 1999. Since that time it's gone through several revisions and was republished in its current form in 2002.

SIP establishes a common ground on which two separate connections can meet. So, for example, if you make a call using VoIP to someone who does not have VoIP, SIP is the protocol that finds the middle ground, or compatibilities between your equipment and connection and the receiver's equipment and connection.

SIP helps create a connection by directing the following activities:

• Determines the connection point that will be used for communication
  
• Determines how calls will be transferred and terminated
• Determines call parameters at both ends of the call and rings the receiver
• Determines the availability of the user receiving the call
• Determines the capabilities of both the caller and the receiver

The protocol achieves these standards by using two components: user agents and network systems. The user agents are end systems along the network that receive requests and return responses for the user. So, for example, if you're making a call to someone, the user agent gets a small piece of information that tells it you would like to place a call. The agent then processes that information and sends it out across the Internet. It then gets a response that tells you if the phone you're calling is available. If it is, the agent puts the call through. If the phone is not available, the agent returns a response that the call cannot be completed—that response could be a busy signal or a recording of some type that tells you the number is not available.

The second part of SIP is the network systems. SIP determines the communication between each of the three types of servers that make up a network system. Those servers are a registration server that logs a user's current location, a proxy server that forwards requests to the next server down the line, and the redirect server that sends information about the next server in line to the proxy server. Together, these three types of servers make up the path through which data travels from one location to another. Although there are only three types of servers, dozens of those servers might be involved in the transport of your call from your location to the receiver's location. SIP ensures that the call is routed properly and arrives at the destination.

In a lot of ways, H.323 and SIP are similar in function. In fact, fundamentally the two protocols do the same thing. Where they differ is in nature. H.323 was designed for use on a single local area network (LAN) for video conferencing, whereas SIP was designed for use on the Internet for VoIP. H.323 borrows from legacy communication systems and SIP does not. Also, H.323 is a binary protocol whereas SIP is ASCII-based. Because of these differences, H.323 works, but can be cumbersome. SIP, which is built much like the HTTP protocol that is the foundation of the Internet, is much simpler and designed to work with the attributes of the Internet.

Like H.323, SIP works in conjunction with a number of other protocols:

• Real-time Transport Protocol (RTP)
  
• Real-time Streaming Protocol (RTSP)
• Session Announcement Protocol (SAP)
• Resource Reservation Protocol (RSVP)

Others exist as well. And it's not necessary for you to know that RTP is for transporting real-time data and virtually every device uses RTP for transmitting audio and video packets, or that SAP is for advertising multimedia sessions. What's important for you to understand is that SIP is a protocol that was designed with VoIP and the Internet in mind. That makes it more effective and efficient for guiding VoIP services.

H.323 is superior in some ways—it provides for better interoperability with the public telephone network, it provides better support for video, and allows better interoperability with legacy video systems. SIP, on the other hand, isn't designed to address the problems that can be encountered in legacy communications systems. However, SIP is easier to develop and troubleshoot. SIP also seems to be more efficient for VoIP, and many equipment manufacturers seem to be turning to SIP over H.323.

Of course, the fact that SIP is more popular doesn't mean that all service providers use SIP. Nor does it mean that all equipment will work with SIP. To be sure that your equipment works with the protocol that your service provider supports, ask your service provider about compatibility. Even better, get your equipment from your chosen service provider instead of a thirdparty vendor.

Dozens of other protocols affect VoIP, but they're parts of the larger picture. For example, the Media Gateway Controller (MEGACO), which is also called H.248, is a device control protocol; the Media Gateway Control Protocol (MGCP) is also a device control protocol; and the Skinny Client Control Protocol (SCCP) is a proprietary VoIP protocol that was developed by Selsius Corporation.

Each of these different protocols is used in concert with other protocols, each of them directing one aspect or facet of VoIP. Currently, no single protocol directs all VoIP activity. H.323 and SIP are the two most broad-reaching protocols, and in truth it would be more accurate to call them protocol stacks.

As mentioned earlier, H.323 stacks H.225, H.235, and H.245 (among others) together to address VoIP as a whole. If you were to envision the protocol stack, it would look like what's shown in diagram bellow.

The H.323 protocol stack

The same is true for SIP. Because it combines numerous other, more specific protocols, it's a stack and would look very much like next diagram if you envisioned it.

The SIP stack of protocols

Even companies that develop and use proprietary VoIP protocols use more than one. What's more, those companies may use one proprietary protocol and several other common protocols to achieve the best-quality VoIP service. There's still no single "best practice," but the intent to find one is there. The race to be the discoverer is what makes these different protocols interesting. It's also what's pushing the VoIP industry to continually improve the quality of VoIP services and equipment.