Chapter 8: Managing the TDM Infrastructure

---

8.1 Introduction

The TDM infrastructure spans both public and private networks and has been in widespread use for several decades in support of voice and data applications. This infrastructure is based on dividing bandwidth into time slots that constitute channels that convey information from source to destination. Despite TDM’s high reliability, it has several limitations, the most serious of which is that it is very inefficient in the use of available bandwidth. Channels are assigned to various devices, but during periods when these devices are idle, the bandwidth goes unused. While this technique is good for real-time applications because bandwidth is always available when needed, it inflates the cost of networks, since they must be provisioned with more bandwidth than is really necessary.

Different techniques have been developed over the years in an attempt to make TDM-based networks operate more efficiently. With statistical time division multiplex (STDM) technology, for example, devices share the available bandwidth. If a device is idle, its channel is dynamically reassigned to another device. The drawback to STDM, however, is that real-time applications like voice may have to wait for a channel to become available. This introduces delay, which can make communication difficult. The variability of delay can also make voice communication impractical.

Digital cross-connect systems (DCSs) represent another way to make TDM-based networks operate more efficiently via their groom-and-fill functions. These devices are deployed in the network to fill in the empty time slots of a T-carrier facility with traffic from other locations so that the available bandwidth can be used more efficiently. The DCS also performs a “groom” function, which splits out individual channels from the T-carrier facility, directing them to their appropriate destinations. The drawback to the DCS is that it must be programmed to carry out the groom-and-fill functions. This means that the channels are set up end-to-end and stay fixed between locations until the DCS is manually programmed to change them or until a stored program is implemented to change them based on such parameters as time of day or specific events. The functions of a DCS are implemented in a static manner rather than in the dynamic manner of a true switch.

The TDM infrastructure—characterized by traditional multiplexers, digital cross-connects, and switches—is being replaced by a more efficient and economical fast-packet infrastructure. Even voice is moving in this direction. Instead of buying new PBXs or investing in expensive upgrades, companies are increasingly implementing VoIP, VoFR, and voice-over ATM. Not only are the routers, integrated access devices (IADs), and packet switches needed for a fast-packet infrastructure less expensive than using a PBX, they also use the available bandwidth very efficiently without sacrificing voice quality or skimping on call-handling features.

Despite the migration of the TDM infrastructure to a fast-packet infrastructure, the complete transformation will be long and gradual. There is still plenty of useful life left in the TDM infrastructure, and it is premature to dismiss it. Furthermore, the two infrastructures are entirely interoperable. While this interoperability is necessary for the migration to fast packet, it also has the effect of preserving existing investments in legacy TDM infrastructures. Even from a management perspective, the TDM and fast-packet domains have been integrated via the SNMP. While many legacy TDM devices lack built-in support for SNMP, the use of proxy agents makes them manageable with SNMP.

---