HYBRID SWITCHING

Hybrid, as the name implies, is a combination of techniques. Various combinations of hybrid switching have appeared in data communications over the years, but have been relegated for the most part to CPE, outside the classical telephone world. In the broader context, IEEE 802.xx LLC (logical link control) and a combination of higher layer Ethernet packets mapped to Internet packets could be labeled a hybrid. However, continued innovation and development emerging after the 1984 deregulation of the long distance telephone business led to newer approaches to transmission and switching such as SONET/SDH, ATM, and other packet- or cell-based technologies. The impact of the Internet Society, in particular the Internet Engineering Task Force (IETF), on global communications standards has been significant to say the least. An example of the kind of innovation that is likely to survive and grow well in to the future is embodied in a technique called multi-protocol label switching (MPLS) sometimes truncated to simply label switching. MPLS is a combination of techniques derived from IP and ATM switching and protocols.

Other Emerging Switching and Routing Technologies

ASON (Automatic Switched Optical Network), GMPLS (Generalized Multi-Protocol Label Switching), and RPR (Resilient Packet Ring) are in active standards development and various stages of maturity at the time of this writing. A full description of each is far beyond the scope of this book. However, a brief description of each is noteworthy. ASON and GMPLS are often seen as a source of contention and competition between ITU and IETF standards development groups, but this is more an impression gained from reading press reports. ASON is a new architecture that will leverage existing SONET/SDH architecture by adding optical switching and control protocols. The motivation for ITU members and their suppliers lies mainly in reduction of capital investment at the core and operations cost associated with provisioning, while at the same time enabling paying customers to use the network in new and innovative ways. Recognize that this work is heavily biased by ITU due process in which the focus is working from well-defined detail requirements for architecture. The basic approach is to agree on the architecture and then work out a set of underlying control protocols, and don’t forget that the background and experience of the developers is heavily influenced by years of operating TDM networks controlled by CCS7/SS7 and DTMF signaling.

GMPLS is an initiative that grew out of effort directed at extending MPLS traffic-engineering techniques. Recall that MPLS as a concept is something of a hybrid or a combination of circuit and packet switching. Motivation for this work lies in the simple desire to more effectively and efficiently control the transmission or transport facilities—circuits in the network. How to do this? Simply invent a new protocol that tells the network to ‘‘do this’’ or ‘‘do that’’ based on a new bright idea, almost without regard to practical business considerations.

The main interest driving the work of both groups appears to be some form of standardized control of the multiplexing and transmission layers in the network. Significant work remains to be accomplished before stability and maturity, sufficient to entice suppliers to invest in design and development of real products capable of attracting service provider investment is achieved. Overall, the serial cycles of standards stabilization, product design, and, finally, deployment by service providers may take 2, 3, 5, or maybe 7 or 8 years. For now, the most appropriate action for the user community is simply monitoring events as they occur.

RPR (Resilient Packet Ring) is the most mature of the three. This work is currently under the moniker IEEE 802.17. Essentially, this is an extension of BLSR (bi-directional line switched ring) technology whereby TDM/PDH, or the more common TDM/SONET/SDH, is modified to replace TDM with Ethernet, or something akin to Ethernet packets. Such an architecture would permit more efficient use of the bandwidth on a given transmission facility because it would allow multiple classes of service on a single facility. Another way of looking at it is to consider that typical BLSR configuration is such that only half of the facility is carrying live traffic, while the other is idle waiting to pickup the traffic when the load-bearing side fails. If the TDM is replaced with Ethernet or similar packet switching techniques capable of carrying classed traffic and priority based packet switching, both sides of the BLSR become useful. If one side fails, lower priority traffic gets delayed or dropped while higher priority traffic moves.

More information on these emerging standards and technologies is as close as your favorite Internet search engine.