There is a lot of technology and a lot of issues that are important to understand with voice/data integration. There’s also a lot of jargon and vocabulary. Pace yourself as we move through this section.
We値l start by looking at TDM versus packet-based networks. Then we値l cover the traditional telephony equipment. Voice quality issues are essential and we値l discuss these, along with the technologies that are making voice/data convergence a possibility.
Traditional Separate Networks
So let’s go back to looking at where most companies are today?
Many organizations operate multiple separate networks, because when they were created that was the best way to provide various types of communication services that were both affordable and at a level of quality acceptable to the user community.
For example, many organizations currently operate at least three wide-area networks, one for voice, one for SNA, and another for LAN-to-LAN data communications. This traffic can be very “bursty.”
The traditional model for voice transport has been time-division multiplexing (TDM), which employs dedicated circuits.
Dedicated TDM circuits are inefficient for the transport of “bursty” traffic such as LAN-to-LAN data. Let’s look at TDM in more detail so that you can understand why.
Traditional TDM Networking
TDM relies on the allocation of bandwidth on an end-to-end basis. For example, a pulse code modulated (PCM) voice channel requires 64 kbps to be allocated from end to end.
TDM wastes bandwidth, because bandwidth is allocated regardless of whether there is an actual phone conversation taking place.
So again, dedicated TDM circuits are inefficient for the transport of “bursty” traffic because:
– LAN traffic can typically be supported by TDM in the WAN only by allocating enough bandwidth to support the peak requirement of each connection or traffic type. The trade-off is between poor application response time and expensive bandwidth.
– Regardless of whether single or multiple networks are involved, bandwidth is wasted. TDM traffic is transmitted across time slots. Varying traffic types, mainly voice and data, take dedicated bandwidth, regardless of whether the time slot is idle or active. Bandwidth is not shared.
After: Integrated Multiservice Networks—Data/Voice/Video
With a multiservice network, all data is run over the same infrastructure. We no longer have three or four separate networks, some TDM, some packet. One packet-based network carries all the data. How does this work? Let’s look at packet-based networking.
As we have just seen, TDM networking allocates time slots through the network.
In contrast, packet-based networking is statistical, in that it relies on the laws of probability for servicing inbound traffic. A common trait of this type of networking is that the sum of the inbound bandwidth often exceeds the capacity of the trunk.
Data traffic by nature is very bursty. At any instant in time, the average amount of offered traffic may be well below the peak rate. Designing the network to more closely match the average offered traffic ensures that the trunk is more efficiently utilized.
However, this efficiency is not without its cost. In our effort to increase efficiency, we run the risk of a surge in offered traffic that exceeds our trunk.
In that case, there are two options: we can discard the traffic or buffer it. Buffering helps us reduce the potential of discarded data traffic, but increases the delay of the data. Large amounts of oversubscription and large amounts of buffering can result in long variable delays.