Chapter 8: ISDN & Transaction Processing

Understanding Packet Switching

What all of these applications share is the almost universal availability of twisted-pair copper telephone connections, and the exceptionally low costs of reaching and using the X.25 network.

X.25 packet switching was defined by CCITT (now ITU-TSS) in the early 1970s, and is an accepted worldwide communications standard. Data to be transmitted is divided into small "packets," and routed through the network.

Packetizing is done by a Packet Assembler/Disassembler (or PAD) at the sending end. At the receiving end, another PAD accepts the hundreds of packets that may be involved, and forwards them as a continuous message to their destination.

Since packet switching was originally developed for accurate data transmission through what were often noisy and interference-prone analog lines, X.25 also performs very high-level error-checking and error-correction. If any packet is not received correctly, based on elaborate parity checking schemes, the receiving network location signals for retransmission until the packet comes through correctly.

The result is exceptional accuracy through lines that were - and in many areas of the world still are - less than perfect for data transmission. The price of this accuracy, however, is a D-channel transmission speed limited to 9.6Kbps.

"D channel speeds are fairly mundane by today's standards," said TeleSystems Marketing Applications' John Mazalewski, an ISDN consultant serving New York and New England. "Yet the primary need for most POS applications isn't speed but low cost, and end-to-end accuracy. In these uses, packet switching is as important today as it ever was in the past."