X.25 Protocol - Introduction
As stated at the end of the previous chapter, X.25 is one of the oldest standard protocols for virtual circuit packet switching. The use of X.25 started in the 1970s, in response to the need to provide users with wide area network (WAN) connectivity across public data networks (PDNs). It was felt that improved equipment compatibility and lower cost would increase subscription to PDNs. The X.25 protocol was the most popular one among others that were developed at that time.
X.25 was developed by the common carriers (i.e, telecommunications providers), rather than any single commercial enterprise. Because of that, the specifications of the protocol were well-suited to various system types. One of the big advantages of X.25 was its international character - it was standardized by the International Telecommunications Union (ITU), an agency of the United Nations, making X.25 a global standard.
When X.25 was developed, communications links were slow, analog, and "noisy", which means that other electrical signals interfere with the desired signals, causing a high bit error rate. Because of these conditions, the protocol provides end-to-end error detection and correction as well as error detection and correction between the DTE and the DCE. Error correction is provided by retransmission of damaged or missing packets. Though there is more overhead and processing in provision of multi-layered error detection and correction, when the bit error rate is high, it is better to do that than to have a damaged packet make its way all the way from source to sink, only to find out that it is damaged, and have to return to "square one".
To qualify what I mean by "slow", a 64kbps link was considered fast in the early years of X.25. The maximum speed link in the last version of the standard is 2 Mbps. We're talking about a WAN technology that is intended to serve as an infrastructure for public data networking. The wide area links in the Internet today are several orders of magnitude faster.
It is the increased speed and quality of communications links - speed and quality that was achieved over the decades with each new generation of communications technology - that has made X.25 all but obsolete. It was replaced, in turn, by several generations of packet switching technologies, which will be looked at in this packet switching portal (or in linked portals). However, we must remember that X.25 will always serve as the basis for comparison and as the foundational packet switching technology for all later and future packet switching technologies. It incorporates many protocol techniques that recur in a variety of protocols, thereby serving as a source of protocol concepts.
As with all packet switched technologies, in order to transmit the data over the network, the data is divided into packets. In addition to the data (the payload), each packet contains information regarding the setup and managing the data flow. Standardized addresses are used to identify the communicating DTEs.
The X.25 protocol specifies three layers: physical, data-link, and network, which are commonly known as the three lowest layers of the OSI Reference Model. The scope of each layter is depicted in Figure 8.
Figure 8: X.25 Layers
The purpose of defining a standard is to enable compatibility of multi-vendor hardware and software systems. What we see in Figure 8 is that the X.25 standards define the interfaces between the DTE and the DCE, between the DTE and the local switch (which in much of the literature is also referred to as a DCE), and an end-to-end (DTE to DTE) protocol. What is not defined is a standard for the behavior of the switches among themselves, represented by a cloud in the diagram. Thefore, there is no guaranteed interoperability between different vendors' switches. X.75 is a standard for interoperation between different X.25 networks, and it can be used as a standard between switching nodes within a particular network, but this is not required.
This idea - that a given vendor operates according to its own proprietary protocols - for example, to choose the paths within the X.25 cloud - is very consistent with the mindset of the era in which X.25 was developed. It was the era in which vendors each defined its own communications protocols and expected a customer to buy one vendor's equipment. Allowing different DTEs to interface to an X.25 network, which was likely homogeneous as far as the switch manufacturer, was the concept of an "open" system at the time. The TCP/IP suite of protocols was revolutionary in that not only did it allow heterogeous end systems (the DTEs of the internetwork) to communicate with a common set of networking and application protocols; all the intermediate systems (the routers) also have to abide by a standard protocol, namely IP, and the IP routing protocols are standard. The result of such a policy is that a network provider is not tied to any one vendor. There are vendors (in fact a very successful one) who developed proprietary IP routing protocols, but the goal of the Internet suite of protocols is interoperability at every level and across the network.
Well, I am digressing. We're here to talk about X.25 and not IP, but by comparing X.25 to IP, we can better understand the characteristics of each, and for those of you who are familiar with IP, you can appreciate the scope of the X.25 protocol and the greater scope of IP. In the coming chapters of the X.25 tutorial, we will look at each of the X.25 layers, including some packet formats and an animated demo. We'll conclude the X.25 tutorial with a brief look at how it is hanging on today.
