VoIP essentially involves sending voice information via a packet distribution network rather than the circuit-switched network of the PSTN.
Circuit switching connects two points in both directions whereby the connection is termed a circuit. This circuit must be set up before a conversation can be established and a dedicated path is issued for the entire duration of the call.
Hence any idle time during the conversation is simply wasted bandwidth.
In comparison, VoIP employs packet switching. This technique maintains the connection long enough for packets to be sent across the network. This minimizes the connection time and thus reduces the network load.
Before VoIP can be utilized, IP telephones as well as digital IP_PBX's need to be installed. The IP_PBX converts voice data into IP packets which are then pushed onto the network. Once the packets reach the destination gateway called the IP host, the message is depacketized and converted back into voice signals and sent out via local phone lines. The IP_PBX also determines the destination for mapping the number dialled and verifies correct format of the number. A session is established between the caller's IP_PBX and receiver's IP host. The connection between the caller's phone and IP_PBX closes when the call is terminated[1].
Standards and Protocols
There are two main protocols that define ways for devices to connect to each other using VoIP, as well as provide specifications for audio codecs. A codec is vital for IP telephony as it changes an audio signal into a compressed digital form for transmission and back into an uncompressed audio signal for replay. [2] One of the aforementioned protocols, H.323, was created by the International Telecommunications Union (ITU). H.323 encompasses many protocols and therefore is responsible for many specifications such as interactive videoconferencing, realtime, audio applications and data sharing. Real Time Protocol (RTP) is an important protocol from this suite, and is also the most widely used. RTP defines a standardized packet format for delivering audio and video over the Internet. Due to this, it is used to transmit VoIP traffic in a majority of implementations.
The other protocol, Session Initiation Protocol (SIP), was specifically created by the Internet Engineering Task Force (IETF) for IP telephony. as opposed to H.323. Although later versions of H.323 have been modified to be more in line with VoIP. SIP is a much more efficient protocol in that it uses preexisting protocols on the PSTN to handle certain parts in the process. Media Gateway Control Protocol (MGCP) is one such protocol. MGCP establishes a gateway connecting to the PSTN system.
Megaco, or H.248, has the same function as MGCP, with the main difference being the product of a collaboration of the IETF and the ITU. SIP itself operates in the Application layer of the Open Systems Interconnection (OSI) model, while the categorization of H.323 in the OSI model is more complicated. This is because not all of the protocols in the suite operate in the same layers. RTP itself is located in the Transport layer while most audio and video protocols are located in the Application layer.
Current Issues
All of the current issues regarding VoIP revolve around the fact that "Quality of Service" is paramount. If the technology cannot prove itself to be reliable and cost effective then no-one will implement it.
There are a number of factors that affect voice quality including latency, jitter, and packet loss. Latency (or delay) is the average time taken for a packet to traverse the network from sender to receiver. Latency takes two forms within networks; propagation delay and handling delay. Propagation delay refers to the time lag between the departure of a signal from the source and the arrival of the signal at the destination[1]. Handling delay refers to the time taken for a device to digitize a signal, compress it and packetize it for transmission onto the data network[2]. The reverse is required for the receiver to hear the message. Latency can cause an annoying delay between when a word is spoken and when it is received. Delays of more than 400ms are audibly noticeable and therefore unacceptable.
Jitter occurs when there is a variation between when a voice packet is expected to be received and when it actually is received, causing a discontinuity in the real-time voice stream[3]. A sudden burst of network activity can cause congestion, varying the expected arrival time of voice packets. Jitter causes a distortion of sound and if the jitter is severe enough it can leave the speaker unintelligible.
Packet loss is the percentage of undelivered packets in the network. When voice packets are lost or arrive at the destination late, they are discarded, possibly causing incomprehensible gaps in the conversation. The maximum packet loss percentage allowable is 3%.
Bandwidth is abundant and cheap but passing large amounts of traffic through the Internet may cause problems as discussed above. Compression is used to relieve the network of large cumbersome communications by compressing a signal normally requiring 64Kbps to 5Kbps whilst retaining good voice quality. Much of the compression is achieved by removing the gaps in speech prior to sending the signal, which can account for as much as 50% of the conversation^]. The use of compression increases the handling delay time.
Networks are not reliable and not always available to the degree that Plain of Telephone Systems (POTS) are (see table 1). Even the best networks have not been able to achieve the standards set by POTS.
Hardware associated with VoIP is expected to have a life span equivalent to current networking equipment (3-5 years) but the technology is so new that this has not been studied to date.
Updating the technology regularly could disintegrate the possible savings that many businesses seek. VoIP will require additional Uninterruptible Power Supplies (UPS) and networking equipment to support the infrastructure. Additionally, telephones under the POTS system are powered and paid for by the PSTN through the Cat 3 telephone wires whereas VoIP handsets will need to be powered and paid for by the business. The handsets can be powered using a separate power supply or via one of the unused pairs currently available on the Cat 5 media.
Conclusion
VoIP yields various benefits such as efficiency and thus stands as being a promising telecommunications tool for the future. Utilizing packet switching to deliver voice data, it abandons the traditional PSTN and replaces it with an IP_PBX to connect calls. Nonetheless, there are many constraints involved including voice quality, overall cost and reliability which need to be acknowledged before pursuing the growing technology. Overtime, many of these issues concerned with VoIP will most likely be overcome; however until then, this form of telecommunication may not be viable for all businesses and such issues need to be carefully considered before giving VoIP the green light.
References
[1] http://computer.howstuffworks.com/ip-telephony2.htm (Accessed 12-
08-04)
[2] http://computer.howstuffworks.com/ip-telephony3.htm (Accessed 15-
08-04)
[3]
http://searchnetworking.techtarget.com/sDefinition/0,,sid7 gci939061 ,00.ht
ml (Accessed 14-08-04)
[4] Chong, H.M and Matthews, H.S, "Comparative Analysis of Traditional
Telephone and VoIP Systems," 0-7803-8250-1/04/$20.00 © 2004 IEEE
(Accessed 14-08-04)
[5]
http://www.cisco.com/en/US/products/hw/routers/ps221/products configur
ation guide chapter09186a008007c9a1.html (Accessed 14-08-04)
[6] Australian Technology, "Voice of IP: Security, Stability, Success" Pg.
94, August 2004 (Accessed 14-08-04)