1. INTRODUCTION
Voice over Internet Protocol (VoIP) technology is capable of passing voice traffic, video and data in the form of packets over an IP network. IP network itself is a network data communications packet-switch, so the phone using an IP network or Internet. With phone calls using VoIP, a lot of benefits that can be taken them are obvious cost is cheaper than traditional phone rates, because the IP networks are global. So for international relations can be reduced by 70%. In addition, costs maintenance to the press because of separate voice and data network, so the IP Phone can be added, moved and changed. This is because VoIP can be installed in any ethernet and IP address, unlike traditional phone must have its own port in Central or PBX.
The development of Internet technology is very rapid push toward convergence with
other communications technologies. Standardization of communication protocols in VoIP technologies such as H.323 has enabled communication integrated with other communication networks such as PSTN. Extensive communications network that has unfolded in Indonesia is managed by the PSTN network
For these networks need to be determined Network Operations Center position
(NOC), Point Of Presence (POP), Router, Gateway and the development of links between cities – cities strategic and efficient. In designing a VoIP network, stressing that in this time is a matter of delay and Bandwidth. Delay is defined as the time required to transmit data from the source (sender) to the destination (receiver), while the bandwidth is the maximum speed that can be used to perform data transmission between computers on an IP network or the Internet.
other communications technologies. Standardization of communication protocols in VoIP technologies such as H.323 has enabled communication integrated with other communication networks such as PSTN. Extensive communications network that has unfolded in Indonesia is managed by the PSTN network
For these networks need to be determined Network Operations Center position
(NOC), Point Of Presence (POP), Router, Gateway and the development of links between cities – cities strategic and efficient. In designing a VoIP network, stressing that in this time is a matter of delay and Bandwidth. Delay is defined as the time required to transmit data from the source (sender) to the destination (receiver), while the bandwidth is the maximum speed that can be used to perform data transmission between computers on an IP network or the Internet.
1.1 delays
In designing network VoIP, the delay is a problem that must be taken into account because the sound quality good or not depends on the time delay. The amount of the maximum delay recommended by the ITU for the application of sound is 150 ms, while the maximum delay with
sound quality is still acceptable user is 250 ms. End to end delay is the sum delay conversion of analog sound - digital delay time can be called also a long delay
package and the network delay at t (time)
Some of the delay that could interfere with voice quality in VoIP network design can
grouped into:
• Propagation delay (delay caused by transmission through the distance between sender and receiver)
• serialization delay (the delay during the process of laying bits into the circuit)
• Processing delay (delay that occurs when the process of coding, compression, decompression and decoding)
• Packetization delay (delay that occurs when the process of digital voice samples paketization)
• Queuing delay (delay due to waiting time for the package until it is served)
• Jitter buffer (delay due to the buffer to cope with jitter)
grouped into:
• Propagation delay (delay caused by transmission through the distance between sender and receiver)
• serialization delay (the delay during the process of laying bits into the circuit)
• Processing delay (delay that occurs when the process of coding, compression, decompression and decoding)
• Packetization delay (delay that occurs when the process of digital voice samples paketization)
• Queuing delay (delay due to waiting time for the package until it is served)
• Jitter buffer (delay due to the buffer to cope with jitter)
In addition, parameters - Other parameters that affect the Quality of Service (QoS), for
same sound is obtained by using traditional telephone (PSTN). Several parameters
that affect the QoS, among others:
• Meeting the needs of bandwidth
• Delay of data (latency)
• Packet loss and desequencing
• Types of data compression
• Interoperability equipment (different vendors)
• Types of multimedia standards used (H.323/SIP/MGCP) To communicate using VoIP technology to real time is the jitter, echo
and packet loss.
same sound is obtained by using traditional telephone (PSTN). Several parameters
that affect the QoS, among others:
• Meeting the needs of bandwidth
• Delay of data (latency)
• Packet loss and desequencing
• Types of data compression
• Interoperability equipment (different vendors)
• Types of multimedia standards used (H.323/SIP/MGCP) To communicate using VoIP technology to real time is the jitter, echo
and packet loss.
Jitter is a variation of delay that occurs due to a lapse of time or interval between
packet arrivals at the receiver. To overcome the jitter of the collected data packets coming
used in the jitter buffer during the time allowed until the package is acceptable to the side
receiver with the correct sequence. Echo caused by impedance differences of tissue
using four-wire with two-wire. Echo effect is an effect that experienced hearing
own voice as a conversation. Hearing his own voice at a time more than
25 ms can cause the cessation of talks. Packet loss (packet loss) when their
peak load and congestion (packet transmission congestion due to traffic congestion which must be served) within certain time limits, then the frame (a combination of payload and header data that is transmitted) votes will be discarded as the treatment of other data frames in IP-based network. One alternative solution to the above problems is to build links between nodes on the network VoIP QoS specifications and dimensions with a well and can anticipate changes traffic increase until at a certain limit.
packet arrivals at the receiver. To overcome the jitter of the collected data packets coming
used in the jitter buffer during the time allowed until the package is acceptable to the side
receiver with the correct sequence. Echo caused by impedance differences of tissue
using four-wire with two-wire. Echo effect is an effect that experienced hearing
own voice as a conversation. Hearing his own voice at a time more than
25 ms can cause the cessation of talks. Packet loss (packet loss) when their
peak load and congestion (packet transmission congestion due to traffic congestion which must be served) within certain time limits, then the frame (a combination of payload and header data that is transmitted) votes will be discarded as the treatment of other data frames in IP-based network. One alternative solution to the above problems is to build links between nodes on the network VoIP QoS specifications and dimensions with a well and can anticipate changes traffic increase until at a certain limit.
1.2 Bandwith
Already described above that the bandwidth is the maximum speed that can be used for
perform data transmission between computers on an IP network or the Internet. In VoIP design, bandwidth is one that must be taken into account so as to meet customer needs
be a parameter that can be used to calculate the amount of equipment is needed in a network. This calculation also is indispensable in network efficiency and cost as well as a reference for the development needs in the future. Packet loss (Loss of data packets on the transmission) and desequencing is a problem relation with bandwidth requirements, but more influenced by the stability of routes passed the data on the network, an efficient method of queue, settings on the router, and use control of congestion (data overload) on the network. Packet loss occurs when there is accumulation of data on the track impassable and causing a buffer overflows in the router.
perform data transmission between computers on an IP network or the Internet. In VoIP design, bandwidth is one that must be taken into account so as to meet customer needs
be a parameter that can be used to calculate the amount of equipment is needed in a network. This calculation also is indispensable in network efficiency and cost as well as a reference for the development needs in the future. Packet loss (Loss of data packets on the transmission) and desequencing is a problem relation with bandwidth requirements, but more influenced by the stability of routes passed the data on the network, an efficient method of queue, settings on the router, and use control of congestion (data overload) on the network. Packet loss occurs when there is accumulation of data on the track impassable and causing a buffer overflows in the router.
2. Protocols to Support VoIP Network
2.1 Protocol TCP / IP
TCP / IP (Transfer Control Protocol / Internet Protocol) is a protocol used
on the Internet. This protocol consists of two major parts, namely TCP and IP. Illustration
processing data to be sent using TCP / IP protocol is given in the figure
below.
TCP / IP (Transfer Control Protocol / Internet Protocol) is a protocol used
on the Internet. This protocol consists of two major parts, namely TCP and IP. Illustration
processing data to be sent using TCP / IP protocol is given in the figure
below.
2.1.1 Application layer
the main function of this layer is to transfer files. Transfer files from a system to system
others require a different control system for the lack of different file systems - different. This protocol deals with the application. One examples of applications that have been known as HTTP (Hypertext Transfer Protocol) for Web, FTP (File Transfer Protocol) to transfer files, and TELNET to a remote virtual terminal.
2.1.2 TCP (Transmission Control Protocol)
In the transmit data in two Transport layer protocols are TCP and role UDP. TCP is connection-oriented protocol which means to maintain reliability communication relationship end-to-end. The basic concept of the workings of TCP is sends and receive segment - segment information with data length varies on an internet datagram. TCP ensure reliability of communication links for doing repairs to the damaged data, missing or send errors. This is done by giving the serial number on each octet signals transmitted and requires a positive response from the receiver of the ACK signal (Acknowledgment). If the ACK signal is not received at a certain time interval, then the data will sends back. On the receiver side, the serial number had been useful to prevent errors sequence data and duplication of data. TCP also has a mechanism of control in a way flow include information in the ACK signal on the limited number of data octets that is still allowed
transmitted in each segment received successfully.
In a VoIP connection, TCP is used at the time of signaling, TCP is used to guarantee setup a call to the session signaling. TCP is not used for sending voice data in VoIP because VoIP on a data communications data handling which has been delayed more important than the lost packet handling.
2.1.3 User Datagram Protocol (UDP)
UDP which is one of the main protocols over IP is the transport protocol that is more
modest compared with TCP. UDP is used for situations that are not concerned reliability mechanisms. UDP header contains only four fields are the source port, destination port, length and UDP checksum in which function is similar to TCP, but the checksum facility
the UDP is optional.
VoIP UDP is used to transmit audio streams send continuously. UDP is used in VoIP due to the delivery of streaming audio continues consistently more concerned about the speed of data transmission to arrive at the destination regardless the package is missing, although 50% of the number of packets sent. (VoIP fundamentals, Davidson Peters, Cisco Systems, 163) Because UDP can transmit streaming data quickly, then the UDP VoIP technology is one important protocol that is used as a header on the sending of data other than RTP and IP. To reduce the number of packets lost during transmission of data (since no there retransmission mechanism) then on VoIP technologies of data transmission much done on the private network.
2.1.4 Internet Protocol (IP)
Internet Protocol communications system designed to interconnect computers on the network paket switched. In the TCP / IP network, a computer is identified by IP address. Each computer has a unique IP address, each different from each other. This done to prevent errors in data transfer. Finally, the protocol data access related directly with physical media. In general, this protocol is to handle detection errors during data transfer. For data communications, Internet Protocol implements two basic functions are addressing and fragmentation.
One of the important things in the delivery of IP addressing is a method of sending information and receiver. We have standards that have been used namely addressing IPv4 with address consists of 32 bits. The number of IPv4 addresses that are created with it is not expected to sufficient for the IP addressing so that in the next few years will implemented a new addressing system that uses IPv6 systems 128-bit addressing.
3. H.323VoIP to communicate with other systems that operate on packet-switch network. For can communicate a need for a standard communication system compatible with each other. One of the VoIP communication standard according to ITU-T recommendation is H.323 (1995-1996). H.323 standard consists of the components, protocols, and procedures that provide communications multimedia over packet-based network. Form of packet-based network that can be passed between other Internet networks, the Internet Packet Exchange (IPX)-based, Local Area Network (LAN) and Wide Area Network (WAN). H.323 can be used for services - multimedia services such as Voice communications (IP telephony), video communication by voice (video telephony), and combined voice, video and data The purpose of the design and development of H.323 is to enable interoperability with the type of Other multimedia terminals. Terminal with H.323 standard can communicate with the terminal On N-ISDN H.320, H.321 terminal on the ATM, and H.324 terminals on the Public Switched Telephone Network (PSTN). H.323 terminal enables real-time two-way communication form voice, video and data.
3.1 H.323 Architecture
H.323 standard consists of 4 physical components which are used when connecting communication multimedia point-to-point and point-to-multipoint on several kinds of networks:
A. Terminal
B. Gateway
C. Gatekeeper
D. Multipoint Control Unit (MCU)
A. Terminal
B. Gateway
C. Gatekeeper
D. Multipoint Control Unit (MCU)
3.1 H.323 Architecture
H.323 standard consists of 4 physical components which are used when connecting communication
• Terminal, Used in real-time multimedia communications in both directions. H.323
can be a personal computer (PC) or other equipment that can stand alone
run multimedia applications.
• Gateway is used to connect two different networks between the network
H.323 and non-H.323 networks, for example, the gateway can connect and
provide communications between H.233 terminals with the telephone network, for example: PSTN. In connecting two different forms of tissue by translate protocols for call setup and release and deliver information between networks connected to the gateway. However, the gateway does not needed for communication between two H.323 terminals.
• Gatekeeper can be considered as the brain in H.323 network as it is a point important to the H.323 network. multimedia point-to-point and point-to-multipoint on several kinds of networks:
can be a personal computer (PC) or other equipment that can stand alone
run multimedia applications.
• Gateway is used to connect two different networks between the network
H.323 and non-H.323 networks, for example, the gateway can connect and
provide communications between H.233 terminals with the telephone network, for example: PSTN. In connecting two different forms of tissue by translate protocols for call setup and release and deliver information between networks connected to the gateway. However, the gateway does not needed for communication between two H.323 terminals.
• Gatekeeper can be considered as the brain in H.323 network as it is a point important to the H.323 network. multimedia point-to-point and point-to-multipoint on several kinds of networks:
A. Terminal
B. Gateway
C. Gatekeeper
D. Multipoint Control Unit (MCU)
• MCU is used to service the conference three or more H.323 terminals. All terminals
who wish to participate in conferences to build relationships with the MCU a set of materials for conferences, negotiations between the terminals for ensuring the audio or video coder / decoder (CODEC). According to the H.323 standard, a MCU consists of a Multipoint Controller (MC) and a Multipoint Processor (MP). MC handles H.245 negotiations (involving signaling) between terminals – terminal to decide audio and video processing capabilities. MC also controls and determines a series of audio and video will be multicast. MC does not deal with directly series media. This task was given to the MP who did mix, switches, and process audio, video, or bits - bits of data. Gatekeepers, gateways, and MCU is a logically separate component in the H.323 standard but can be implemented as a physical instrument.
who wish to participate in conferences to build relationships with the MCU a set of materials for conferences, negotiations between the terminals for ensuring the audio or video coder / decoder (CODEC). According to the H.323 standard, a MCU consists of a Multipoint Controller (MC) and a Multipoint Processor (MP). MC handles H.245 negotiations (involving signaling) between terminals – terminal to decide audio and video processing capabilities. MC also controls and determines a series of audio and video will be multicast. MC does not deal with directly series media. This task was given to the MP who did mix, switches, and process audio, video, or bits - bits of data. Gatekeepers, gateways, and MCU is a logically separate component in the H.323 standard but can be implemented as a physical instrument.
3 2 H.323 Protocol
In H.323 there are several protocols in the delivery of data to support for data sent real-time. Below are described several protocols on the network and transport layer.
3.2.1 RTP (Real-Time Protocol)
Is a protocol designed to compensate jitter and desequencing that occurs in IP network. RTP can be used for some kind of real-time data streams such as data voice and video data. RTP contains information on the types of data to send, timestamps are used to adjust the time sounded like a voice conversation as spoken, and sequences numbers are used for sorting data packets and detect lost packets
3.2.2 RTCP (Real-Time Control Protocol)
It is a protocol that is usually used in conjunction with RTP. RTCP is used to transmit control packets every terminal participating in conversations used as information for the quality of transmission on the network. It has two important components of RTCP packets, the first is that the sender report contains information on the number of submitted data, check the timestamp in the RTP header and ensure that data is accurate with its timestamp. The second element is the receiver's report sent by the recipient of the call. Receiver report contains information on the number of packets lost during a conversation session, displaying the last timestamp and delay since the delivery last sender report.
3.2.2 RSVP (Resource Reservation Protocol)
RSVP works in the transport layer. Used to provide bandwidth for voice data submitted do not experience delays or damage when it reaches the destination address unicast or multicast. RSVP is an addition to the VoIP signaling protocol that affect the QoS. RSVP works by sending requests on each node in a network that is used for shipping data stream and at each node RSVP makes resource reservation for transmitting data. Resource reservation in a node is done by running two modules, namely admission control and policy control. Admission control is used to determine whether a node has a resource enough to meet the required QoS. Policy control is used to determine whether user who has administrative permissions (administrative permissions) to make reservation. When errors in one application module, an error will occur where the request is not RSVP will be met. When both modules are running well, then RSVP will form the parameters packet classifier and packet scheduler. Packer Clasiffier determine the QoS class for each packet data used to determine the path used for transmission of data packets based on class and packet scheduler function for setting the interface (interface) of each node for package delivery in accordance with the desired QoS.
4. Voice Data Compression Standard
ITU-T (International Telecommunication Union - Telecommunication Sector) makes some
standard for voice coding is recommended for VoIP implementation. Several standard which is often known, among others:
4.1 G.711
Before you know more what it was previously given little G.711 brief overview function
of compression. A good video channel without the compression will take a bandwidth of about 9Mbps. A voice channel (audio) is good without the compression will take a bandwidth of about 64Kbps. With the compression techniques, we can save a video channel to about 30Kbps and voice channels into 6Kbps (half-duplex), meaning that an Internet channel that is not too quickly can actually be used to distribute video and audio simultaneously. Certainly to the needs of two-way conference takes double the bandwidth, meaning at least once we must use a 64Kbps channel to the Internet. That way voice / audio will take bandwidth is much less in the appeal sending pictures / video.
PCM converts analog signals into digital form by sampling the analog signal is 8000 times per second and encoded in numeric code. The distance between samples is 125 μ sec. Analog signal in a conversation is assumed to 300 Hz frequency - 3400 Hz. Signal sampled then converted into discrete form. This discrete signal is represented by the code
adjusted to the amplitude of the signal sample. PCM format using 8 bits for encoding. Obtained by multiplying the transmission rate of 8000 samples / second with 8 bits / sample, resulting in 64 000 bits / second. 64 kbps bit rate is a transmission standard for single channel digital telephone.
adjusted to the amplitude of the signal sample. PCM format using 8 bits for encoding. Obtained by multiplying the transmission rate of 8000 samples / second with 8 bits / sample, resulting in 64 000 bits / second. 64 kbps bit rate is a transmission standard for single channel digital telephone.
Conversations in the form of analog signals through the PSTN network compression and coding experience into digital signals by PCM G.711 before entering the VoIP gateway. In the VoIP gateway, in the terminal, there is an audio codec framing process (the formation of an IP datagram frames compressed) from digitized sound signal (the PCM G.711) and also doing reconstruction at the receiver. Frame - a frame that is the package - this last information packet transmitted over IP networks with a packet network communication standard - based. Standard G.711 is a compression technique that is inefficient, because it will eat the bandwidth of 64Kbps to channel talks. To be used bandwidtrh not large and did not rule out sound quality, then the solution used to compresed and used in G.723.1 standard.
4.2 G.723.1
G.723.1 voice signal coding is a type of coding is recommended for voice terminals low bit rate multimedia. G.723.1 dual rate speech coder that can be switched the limit 5.3 kbit / s and 6.3 kbit / s. By having a dual rate speech coder is then G.723.1 have flexibility in adapting to the information contained by the voice signal. G.723.1 is equipped with facilities to enhance the sound signal synthesis. In section G.723.1 encoder is equipped with a formant perceptual weighting filter and harmonic noise shaping filter at the decoder while his G.723.1 has a pitch and formant postfilter so that the reconstructed sound signal to be very similar to the original. Excitation signal for low bit rate encoded with Algebraic Code Excited Linear Prediction (ACELP) while for the high rate encoded using Maximum Likelihood Multipulse quantization (MP-MLQ). Rate a higher yield and better quality. Input for G.723.1 is a digital sound signal in-sampling with sampling frequency of 8000 Hz and quantized with PCM 16 bit. G.723.1 is an algorithmic delay of 37.5 msec (frame length plus lookahead), the processing delay is determined by the processor is working count G.723.1 algorithm. By using DSP priosesor the processing delay cans be minimized. In Addition, voice data compression is recommended ITU G.726, an ADPCM voice coding technique with the results of coding at 40, 32, 24, and 16 kbps. Also Usually Used on the sending of data packets on a public phone or equipment That PBX supports ADPCM. G.728, a CELP voice coding technique with results 16 kbps encoding. G.729 is a CELP type voice encoding with compression results in 8kbps.
