September 09, 2005
Carrying voice over frame relay, IP, and ATM - Part 1
Voice can be transferred over frame relay, IP, and ATM. The rate of growth of data networks has been greater than that of voice traffic, primarily due to the increased availability of broadband. This has led to the phenomenon of voice being sent more regularly on data networks. Voice over Frame Relay, VoFR, has been developed from the Frame Relays that were developed in the 1990's but were not suitable for carrying voice.
The growth of the Internet fuelled the demand for networks that would enable carriage of voice data in an inexpensive manner. Frame Relay, IP, and ATM are examples of packet networks designed for carrying voice and data, thereby satisfying the market need for a universal and cheap convergent technology. The next step is to improve the integration standards for the ubiquitous delivery of voice over Frame Relay, IP, and ATM.
Frame Relay, IP and ATM differ from PSTN, which is a circuit switching technology and has been specifically designed to carry voice. ATM differs from Frame Relay and IP in the sense that it breaks the data into small cells, this helps to speed up the data switching process in the network. Cell switching technologies can perform dynamic allocation of bandwidth depending upon their activities, this is known as statistical multiplexing. There is no reservation of bandwidth for a particular use, at any given time a bandwidth may be allotted as per the requirements of the network. As opposed to to cell switching networks, in a traditional circuit switching network, a line is dedicated to a call for the duration of the call and even when the call is on hold. This has the advantage of a consistent voice transmission but the bandwidth utilization is inefficient as the dedicated line cannot be used for other data transmissions even in the absence of voice transmissions.
The packet switching networks were originally developed to handle traffic that moved in bursts. This means that packet switching networks are by default not as efficient as circuit switching networks as far as handling voice is concerned. The shared nature of a packet switching network implies that the delay in voice packet transmissions is erratic and not as less as desired for achieving a satisfactory quality of voice. Voice transmission should ideally be an accurate representation of the speaker's tone, inflection, etc. Delay in the delivery of the voice packets can lead to a communication gap, which may be further accentuated by packets being dropped along the network path.
Delay in the network can be reduced by increasing the bandwidth but it is an expensive alternative and the benefits of a shared bandwidth cannot be availed. Ideally, traffic congestion should be managed at the customer's end and at the backbone by prioritizing the flow of traffic. This had led to the development of smart access equipment for implementing procedures that help in reducing packet loss due to network congestion.
Voice over Frame Relay (VoFR): Frame Relay finds application in data networks in companies as it offers flexible bandwidth, easy accessibility, and a mature technology that supports a variety of traffic. Frame Relay offers the advantage of a predictable performance, it runs on the principle of Permanent Virtual Connections (PVCs) and is well suited for star configurations and closed user groups. Voice Frame Relay access devices (VFRADs) connect the router, SNA controller, and the PBX to the Frame Relay Network. This helps in achieving the integration of voice into the data network. MAXcess is a VFRAD that helps to surmount the difficulties in transmitting voice data over the Frame Relay without having to increase the bandwidth. It does so by employing the techniques described below:
Prioritization: VFRADs mark applications as per their reactions to delay. Voice and other time-sensitive data like SNA is given higher priority. Since voice transmissions are short and do not require much bandwidth, they do not have a detrimental effect on data traffic; in fact, they can be sent alongside other information like emails, graphics, etc that travels on the network. Different QoS packages are provided by the Frame Relay service providers. Clients prefer to purchase the best QoS for voice and SNA traffic; Non-real time variable frame and a slightly lower QoS, it is preferred for a LAN network and Internet connections. VFRAD can also be set to recognize traffic that can be dropped in case of network congestion. It achieves this by utilizing the Discard Eligibility bit.
Fragmentation: VFRADs have the capability of fragmenting data packets to allow voice data higher priority in terms of transmission even if it leads to stopping the other transmissions. However, increasing fragmentation leads to reduced bandwidth efficiency due to an increase in the number of data frames. Applications such as RAD FR+ make it possible to send complete data frames and fragmenting occurs only if the voice data reaches a switch in the midst of a data transmission.
--
Did you enjoy this post?
« VoIP protocols - Part 3 | Main | Security Concern of VoIP »
