Introduction

Unlikely as it might seem, all the major physical elements of the public switched telephone network (PSTN) have remained pretty much unchanged since the divestiture of AT&T, even though the network has grown slowly but steadily over the years, and newer technology and service providers have been added.

When we look at the building blocks of the PSTN--the central office switches, the buildings that house them, and the network transmission facilities--the numbers stay fairly constant. For example, there are almost exactly 20,000 central offices (COs) in the PSTN today, a slight decline from the 20,043 counted in Insight's 1992 study of this market. This fact is not really surprising since the CO was first designed as a wire center--its major function is to serve as a point of concentration where access lines from individual subscribers are connected to other lines--and the overriding design consideration when establishing a CO is to physically
locate it near the geographic center of a cluster of subscribers. With only modest changes in population demographics, the number and locations of COs have changed little.

Yet underlying this seemingly static outward appearance, there is enormous inner turmoil. New forms of competition, user requirements, technology and regulation are forcing dramatic changes in the fundamental economics of the PSTN. Even as automation-enabled cost reductions caused significant cuts in telephone-company staffs, competition-induced price cuts forced provisioning changes and continual shifts in the allocation of investment among carriers.

Dispersion of Network Functions

Insight's analysis suggests that within the telco central offices and their associated networks, movements toward consolidation, dispersion and integration are all occurring simultaneously. Seemingly contradictory, these trends are in fact complementary, and are being driven primarily by cost-efficiency issues.

One of the clearest and most significant trends in the CO environment is the distribution of network functions to the periphery. This dispersion process is moving network intelligence toward both the user and to other locations away from the CO and further into the network.

Interface functions between end-user devices and the network, once performed almost exclusively at the telco CO, are being moved toward the user. This is exemplified at the locations of large users by DS1 and DS3 interfaces, which are built into customer equipment, such as PBXs and LAN-based communication servers. It appears on an even greater scale in the residential and small-business markets as digital-loop carrier (DLC) systems. In the rural markets this distribution phenomenon appears as the
introduction of remote switches providing users with all the intelligence of the larger remote host.

The process of distributing the switching functions is well under way in almost all telcos. Two levels of switching are distributed from the CO toward the subscriber in the distribution network, and two or more levels--satellite or remote switching and concentration--are distributed into the interoffice network.

For the past 20 years the DLC systems have incorporated a concentration capability as well as a multiplex capability. DLC systems are forecast to handle over 45 percent of all access lines by the year 2000. The concentration function of the DLC establishes connections to the central office for only those subscriber lines where there is activity. The typical level of concentration is 2:1, so that a 96-channel DLC could serve a total of 192 subscriber lines. While this is not a complete switching function, it is one of the elements that would otherwise be performed in a CO switch.

Remote switches controlled from the host switch in the CO are true switches in that they make complete connections between subscribers on the same remote switch. They are also of more recent vintage than most of the DLC systems and while the DLCs are basically intended to provide pair gain in the distribution plant, the remotes are intended to replace a complete switch, either as an upgrade of a small mechanical switch or as an alternative to a new CO in a newly developing subscriber area.

Mechanical, computer, digital, and remote switch technologies are all represented in the PSTN today, though the mechanical switches are being replaced rapidly. The computer switches are also being replaced, but at a slower rate. Of the more than 21,000 switches in the telco networks, less than 1,800 mechanica l switches remain; a reduction from over 4,000 mechanical switches three years ago.

The net effect of the distribution of intelligence further into the network is that the subscriber lines appearing at the telco CO (and essentially all telecom lines of all types do appear at these offices) are in the form of high-bandwidth signals, usually DS1 but increasingly DS3 and SONET. Such lines have long been the major type of interface on the trunk side of the CO, but the growth of high-bandwidth interfaces on the subscriber side is of relatively recent origin.

The Future of the Central Office

The increase in user demand for high-bandwidth interfaces and the continuing upgrades made at the local loop--which is occurring all across the PSTN--have some interesting implications for the CO given the new competitive landscape.

A key point is that almost two-thirds of all business access-line requirements fall in the midrange of establishment sizes: 8 to 256 employees. Conversely, less than 25 percent of all access-line requirements are derived from establishments in the 512-employee-or-larger range. The implication is that despite the marketing noise about losing critical market segments and the competitive threat posed by the CAPs, 75 percent of all business access-line requirements derive from establishments where the threat of bypass is minimal.

Thus Insight's view of the long term prospects for the CO--and for the LECs that own them--is positive. The CO is by design strategically located at the geographic center of a cluster of telecom users. Which ever course the unbundling of PSTN elements follow, most traffic must continue to pass through the CO.

The CO is not isolated from regulatory and evolutionary pressures; make no mistake about it, as intelligence moves outward to the end points and new data and video services increase the demand for bandwidth, the architecture of the CO will undergo substantial modification. Our analysis suggests that functionality within the CO will be distributed among a number of special-function devices all of which will be inter-connected by means of a high speed bus/ ring architecture.

Nearly all of the devices which have found their way into the COs over the last few years have been intelligent; that is, they are equipped with on-board computers and can be controlled from an external system via a high speed interface. One of the important implications of this is that with standardized interfaces such as SONET, these devices can be interconnected to each other within the CO. The resulting distributed processing network would enable the relatively easy addition of new devices and new functions. It is through such a distributed function system that ATM and other high-speed applications are most likely to be integrated into the telco network.

Looking ahead, Insight's research suggests that the latest round of CO switch replacements is almost over and a new phase will not begin for at least five years, and probably much longer.

The current evolutionary vector suggests that as newer functional devices are brought into the CO they will not only be interconnected with traditional voice switching fabric but with all of the CO facilities including the local distribution network, the signaling network, and the operational support systems. This arrangement not only ensures the continuation of total support long enjoyed by the end-customers, but it also facilitates the introduction of new capabilities without a major investment in new support
structures and personnel training.

Despite all of the major challenges on the technological, regulatory, competitive and marketing fronts, we came away from our research with an optimistic view of the place of the CO in the new telecommunications infrastructure. Our view is heavily premised on the likelihood of a shift in attention on the part of telco management back to business needs and opportunities within the local telecommunications-services market. Even in the face of major economic upheavals being wrought by the new regulatory environment, the telco central offices and their associated local distribution plants are ideally positioned to retain a dominant and profitable place in the telecommunications industry.

Back to Top