VoFR, VoIP and ATM - Essay Example

Summary Frame relay, IP and ATM are the three technologies based on 'Packet Switching'. Packet switching is in contrast to the 'Public Switched Telephone Network' (PSTN) which is a 'Circuit Switching' based technology. In case of packet switching the data that flows through the networks is in the form of small packets whereas circuit switching technology offers to transmit analog signals. In the present age of information technology it has become a need of the day to integrate data and voice. Major companies having very large networks want to reduce the running cost of their networks. Instead of having two networks, one for data and one for voice, they want to carry voice over their data networks thus enjoying both facilities while running a single network. This need of companies has resulted in the development of packet switching technologies. The integrated networks came to be known as 'Voice over Frame Relay' (VoFR), 'Voice over IP' (VoIP) and 'Voice over ATM' (VoATM). As the technology improves, it becomes cheaper. Due to this not only big corporates but small business houses and even schools and colleges will adopt these technologies in the near future. In 'Frame Relay' and IP the data is represented in the form of variable sized packets or frames. In ATM data is chopped into small cells which results in increased switching rate of data through the network. Unlike circuit switching where different time intervals are allocated to different links, the packet switching dynamically allots bandwidth to various links based on their transmission activity. ATM has been designed to reduce delays whereas the other two, FR and IP, have the disadvantage of having delays over circuit switching technologies. However the use of access devices such as routers that use sophisticated schemes enables to overcome this limitation. These schemes include 'Prioritization', 'Fragmentation', 'Silence Suppression' and 'Voice Compression'. Introduction The network developers of both service providers and enterprise are interested in voice-data integration. For service providers the point of interest is the lower cost of production of packet voice. Presently the cost of packet voice network is 20 to 50 percent of an equivalent circuit based-voice network. Similarly for enterprise network developers the main attraction towards these technologies is the cost savings associated with toll bypass and tandem switching. Besides the cost of development, the maintenance cost is also les. Also, more efficient network control and management is achieved. 'Finally, packet based voice systems offer access to newly enhanced services such as unified messaging and application control. These, in turn, promise to increase the productivity of users and differentiate services' (cisco.com, 2005). The voice-data integration technologies have improved rapidly in recent years due to the eagerness shown by both developers and customers. The demand has increased as the customers are much more interested in investing in these technologies to enjoy integrated voice applications. The network developers have been fortunate enough in this case as there has been advancement in the areas such standards, technology and network performance. Overview of the technologies A general background of these technologies and the basic principles and features are given one by one in this section below:- Voice over Frame Relay (VoFR) Definition VoFR uses a router to carry voice traffic like telephone calls and faxes over a frame relay network. Deployment The most used packet voice technology in networks is VoFR generally used in corporates. Prioritization Frame relays use access devices which put 'Tags' on different applications setting priority to all of them depending on the sensitivity they show towards delay. The higher priority voice packets are allowed to move first and till that time the others are kept waiting. Fragmentation The data packets in this case are divided into small fragments which results in fast switching and fast switching ensures lesser delay of voice packets. Variable delay The variation in the time of arrival of packets is known as 'jitter'. This causes non smooth voice streams (Ahmed, Fatima, n.d.). In order to rectify this effect jitter buffers are used in frame relays networks. Voice compression 'Frame relay access is usually at 56/64 Kbps. Low bit rate algorithms such as ITU G.723.1 and G.729A are used to maintain high quality voice' (Ahmed, Fatima, n.d.). Silence suppression The telephone networks use half of the line at a time which means that while one person speaks the other one listens. The remaining time can be used for carrying other packets. This technique is used in VoFRs. Echo cancellation Improvement of voice quality is achieved by using echo cancellers which reduce the echo resulting from the reflection of caller's voice back towards him. Header size The header size in this case is 2 bytes. Voice over Internet Protocol (VoIP) Definition VoIP uses a single packet network line to transmit both voice and modem data at a time. Deployment VoIP uses an IP, i.e., internet network to pass voice along with the other data. Prioritization The prioritization technique used for VoIP is 'Quality Of Service' (QoS) scheme. 'The IP QoS protocol uses ToS (Type of Service) octet field of the IP header to classify traffic at the borders between customers and ISPs' (Ahmed, Fatima, n.d.). Fragmentation VoIP performs fragmentation in a similar manner as in the case of frame relays, the only difference being the size of IP header which is more in case of VoIP. Variable delay The technique of jitter buffer is used also in VoIP as in VoFR. Voice compression Voice compression is essential in case of VoIP as the traffic speed is very low. The voice compression techniques used are same as in VoFR. Silence suppression VoIP uses silence suppression to save about 60% bandwidth. Echo cancellation VoIP uses echo cancellation to eliminate echo. Header size The size of IP header is large, about 20 bytes. Voice over Asynchronous Transfer Mode (VoATM) Definition VoATM transmits voice traffic over an ATM network by utilizing an ATM switch. Deployment ATM is a multi service, high speed, scalable technology (Ahmed, Fatima, n.d.). Instead of this it does not find many applications as the cost of installation and services in this case is very high. Prioritization ATM uses QoS parameters to set prioritization to the voice packets. Fragmentation ATM also uses fragmentation. The data packets are fragmented into small fixed-sized cells of 53 bytes. Variable delay ATM implements 'Dynamic Bandwidth Circuit Emulation Services' (DBCES) which do not send the voice in the form of constant bit stream of cells but the transmission takes place only at an active voice call resulting in reduction of delays. Voice compression In case of pure ATM networks compression does not take place whereas it is essential if a hybrid (ATM & FR) network is used. Silence suppression Only hybrid networks need silence suppression. Echo cancellation ATM employs the service of same kind of echo canceller for cancellation of echo as in VoFR and VoIP. Header size 5 byte headers are used in case of ATM cells. Figure 1: Equipment that supports FRF.12 fragmentation can break up large data packets to ensure that voice traffic is not subject to delay (figure courtesy of the Frame Relay Forum [FRF]). (Source: novell.com) Routing A number of channel configurations are possible in packet voice systems. PCM-Packet In this routing a PCM time slot is linked with a network data stream. Data is taken from a PCM time slot, encoded, packetized and placed in the network data stream. While decoding, packetized data is taken from the data stream, decoded and placed on to the PCM time slot. PCM-PCM In this case a PCM time slot is linked with another PCM time slot. Data from first PCM time slot is encoded and placed on second PCM time slot. Decoding involves taking of data from the second PCM time slot, decoding it and placing it on the first one. N PCM-Packet This routing involves the linking of N PCM time slots of circuit data (rather than voice) with network data stream. The encoding process of this link takes data from N separate PCM time slots, multiplexes and packetizes it and places the packetized data in a network data stream (Adaptive Digital Technologies, Inc., 2001). During decoding process, packetized data from network data stream is demultiplexed and placed onto N PCM time slots. Packet-Packet In this case a network data stream is linked directly with a second network data stream. DSP carries on the translation of one packet type to another packet type. This translation can occur in a gateway that is connecting two users whose protocols are not compatible (Adaptive Digital Technologies, Inc., 2001). Signaling VoATM PVCs are created in case of both signalling and voice transport. Firstly, from end station to end station, a signalling message is carried over the signalling PVC. Secondly, a PVC is selected between the end systems as a result of coordination so that voice communication is carried between the end stations. VoFR Historically, Frame Relay call setup has been proprietary by vendor (cisco.com, 2005). This means interoperability of products from different vendors is not achieved. Frame Relay Forum FRF.11 creates a standard for call setup, coding types, and packet formats for VoFR, which defines the basis for vendor interoperability. Quality Whichever of the three technologies we may be using, they all have the capability to transmit voice of high quality. There are a number of factors on which the voice quality depends. One of the problems faced in this regard is due to overloading of the network. The network may be having not enough bandwidth to support a large number of voice calls simultaneously. This may happen during peak hours when maximum numbers of calls are made. So it is necessary that one has sufficient bandwidth available in the network. Another reason for this may be the non availability of Quality of Service (QoS) mechanisms such as jitter buffers which minimise jitter in delivering voice packets. However most of the service providers do offer such mechanisms and thus the reason for low quality of voice (if any) is mainly due to lesser availability of bandwidth. The use of VoATM is better than VoFR as far as delays are considered. This is because ATM has been designed to support all types of traffic including voice and is thus having intrinsic QoS mechanisms to deal with voice delays. 'For example ATM defines a Constant Bit Rate (CBR) service that specifies a peak cell rate, a cell loss ratio, and a cell delay variation (A cell is the unit of information, or packet, that ATM networks carry)'(Kennard, Linda, 1999). ATM CBR service offers to transmit voice with very low delay and jitter. Contrary to this VoFR does not provide with QoS mechanisms. So in this case the equipments being used in the network must support traffic prioritization and QoS techniques in order to enable low delay and jitter of voice. Besides this one has to go for the equipments supporting FRF standards FRF.11 and FRF.12. Cost The most appealing thing regarding packet voice is lesser cost of production. In fact, the first thing that companies look for before installing packet voice technologies is the percentage reduction in the usage cost and then they look for quality. Surely, since we are having a single network operating for both data and voice transmissions, there is a lot of money being saved. But for the companies already having a circuit switched technology employed what matters most is the reduction in overall call costs. Packet voice is at present 20 to 50 percent less expensive than the circuit switched voice. However the companies have to invest a lot before utilising the services. So the reduction of charges above 30-40 percent is worth considering. Figure 2: More than 50 percent of normal speech is silence (figure courtesy of the Frame Relay Forum [FRF]). Regardless of which packet voice technology one chooses, one can expect good quality and long-term savings. However, nothing is fail proof or free (Source: novell.com) Technology Recent technological advances have resulted in improved voice-data integration. Digital signal processor (DSP) technology has resulted in processing of analog signal in digital domain. This involves the use of small powerful chips with very high processing speeds which sample, digitize and then compress the voice. At present a single chip can mature about 4 calls by itself simultaneously. These advancements in technology have reduced the cost of installation and maintenance by a great margin. Similarly there has been advancement in voice codec (coder/decoder) technology which has resulted in improvement of voice quality. Generally it is believed that a decrease in the bandwidth causes the voice quality to be reduced linearly. Algorithms used in modern day codecs have ensured a good voice quality even when using a fraction of bandwidth employed earlier. Interoperability There is a need for interoperability between the three packet voice technologies, mostly in case of a corporate network. A corporation using a frame relay or ATM for transmitting voice and data may require to communicate with remote branches not having frame relay or ATM infrastructure employed. For this they can use VoIP, thus resulting in inter working of different technologies. The other users employing this sort of interoperability are telecommuters and salespersons working from home, and resellers needing to access information. Inter working between Frame Relay and IP is essential as maximum companies have frame relays installed as compared to ATMs. This is however not as easy as algorithms of different standards for voice compression are used by Frame Relay and IP (ITU G.729 and ITU G.723A, respectively). A difference also exists in signalling methods. The FRF.4 recommendation for switched virtual circuits is not widely used as the basis for Frame Relay voice switching (Biran, Gil, n.d.). Actually, no standard is there for voice switching in Frame Relay networks. Whatever is the case Frame relay does not inter work with VoIP voice switching based on the H.323 protocol stack. In large corporate networks where we have many remote branches having pre installed Frame Relay service and VoFR, a high speed ATM service is needed which will be able to support the huge amount of traffic at company headquarters. It is evident that in this case an inter working or interoperability standard is required between Frame Relay and ATM that will carry on the required functioning particularly speaking there should be inter working between VoFR and VoATM. There is another reason for the need of inter working between these technologies. That is to migrate to different technologies, if and when needed, without having to waste the initial investment made in the existing devices. Future As the advancements in voice-data integration technologies continues some new applications have come to the surface. The packet voice technologies will slowly and slowly replace PBXs completely instead of providing simple transport and switching functions for PBXs (cisco.com, 2005). The future packet voice based products can be classified into two general types:- Un PBX In this case a PC based server contains both trunk gateway ports and analog telephone ports (cisco.com, 2005). Some special softwares are available running on NT operating systems which can fully control the functionality of the telephones and can function as analog telephones themselves. As many as 48 telephones can be normally made to operate through systems. LAN PBX The products coming under this category are those based on LAN telephony all the way to desktop. Some of these products offer to install software which acts as LAN telephony whereas in other cases a simple telephone equipment is allowed to plug into the network. The major hurdle in the path of replacement of PBXs with packet voice technologies are the reliability and scalability problems. These issues must be addressed on high priority basis. In general call control models that are designed to reduce complexity of server provide for better scalability. Conclusion We are in that age where data traffic is exceeding the voice traffic. This means more data networks can be seen in the future than voice networks. Due to this there is a need to transmit voice over data networks rather than vice versa. The introduction of packet voice is a step forward in this direction. Data-voice integration has already entered into the phase where from it is expected to do wonders in the field of voice transmission and with the pace of technological advancement the day is not far when packet switching will completely take over from the circuit switching technology. In comparison VoFR, VoIP and VoATM all have some advantages and disadvantages over each other. The need of the hour is to choose the benefits of each and make them to interoperate. Whereas VoIP is beneficial in transmitting calls to remote locations far away from the main office, the other two have their own benefits. This however does not mean that traditional systems such as PBXs will instantly vanish but it will take time. It is interesting to see that traditional PBX developers are upgrading their products to become packet enabled. To provide toll-bypass capability, PBX vendors have started to use H.323 VoIP cards to set the PBXs to manage H.323 clients as well. The PBX is seen as developing into a voice server. It will be decided by time as to which solution is better to use but it will surely add to the list of choice of the customers. References Adaptive Digital Technologies, Inc. (11 October, 2001). G.PAK Packet Voice DSP System. Retrieved April 05, 2006 from < www.adaptivedigital.com/pdf/gpak_white_paper1.pdf > Ahmed, Fatima (n.d.). A Comparison of Voice Technologies (VoIP, VoFR, and VoATM). Retrieved March 30, 2006 from < www.developer.com/voice/article.php/3112781 > Biran, Gil (n.d.). Voice over Frame Relay, IP and ATM. Retrieved March 30, 2006 from < http://www.protocols.com/papers/voe.htm > Cisco.com (February 15, 2005). Voice/Data Integration Technologies. Retrieved March 30, 2006 from < http://www.cisco.com/univercd/cc/td/doc/cisintwk/ito_doc/voicdata.htm > Kennard, Linda (01 Aug 1999). Packet Voice. Retrieved March 30, 2006 from < http://support.novell.com/techcenter/articles/nc1999_08a.html > Read More