Apparatus for decomposing an automatic cross connect system at a remote wiring hub

The present invention relates to providing connectivity to telephone services at a remote hub. Specifically, the present invention relates to the design and implementation of an analog cross-connect switch located between a subscriber line and a feeder in a remote hub.

Telephone service providers for telephone (eg, POTS and DSL) services use external plants that provide connectivity from service subscribers to the service provider's telephone office (CO). The most common medium used for external plants is the copper loop. Copper loops typically traverse from CO to the subscriber's home through manholes, wiring cabinets, pedestals, and utility poles before being terminated by a network interface device (NID) on the subscriber's premises.

Manholes and cabinets represent the main centers of wiring. Each cabinet or manhole (ie, remote hub) typically serves about 500 homes. Each home is wired using about 3 to 5 pairs of copper wire, depending on the service provider's practice. Therefore, about 1500 to 2500 subscriber lines can be terminated at such a remote hub. However, it is unlikely that all subscriber lines will be active. Therefore, for cost and practical reasons, fewer wires (ie, feeders) are laid between the remote hub and the central office.

Traditional remote hubs include two frames, one that terminates multiple subscriber lines from the home (subscriber premise) and one that terminates multiple feeders from the central office. Is done. The wiring connection between the two frames is made to provide end-to-end connectivity from the subscriber to the central office, thereby providing traditional telephone service. In current practice, such wiring connections are performed manually by dispatching field personnel to the hub.

In addition to traditional telephone services, service providers are now implementing digital subscriber line (DSL) access as a means of providing broadband access (eg, video and data) to subscribers. To achieve maximum rates, DSL service providers are increasingly using DSLAMs (digital subscriber line access) on remote hubs. Modem) is placed, which reduces the distance of the copper loop to the subscriber premises and increases the speed of the DSL service. However, the broadband access market is very competitive with many service providers competing in the same market using different technologies (eg, DSL, cable, fiber, etc.). This competitive market has resulted in high subscriber churn (ie, conversion) rates for such broadband services. All "cancellations" of DSL services usually involve the dispatch of field technicians to rewire the remote hub connection. Dispatches are costly and time consuming, and service providers want to reduce them as much as possible. One way is to deploy an automatic cross-connect (AXC) system that switches analog signals at these hubs. Such AXCs can be remotely controlled by engineers at the Network Operations Center.

When switching analog signals, the connection over the cross-connection must be able to carry a significant amount of current (eg 250-300mA). In addition, the remote hub must maintain connectivity in the event of a power outage, thereby guaranteeing emergency service calls (eg, 911 calls).

One of the prior art techniques for creating analog cross-connections that meet the above two requirements is to use electromechanical relay systems such as microelectromechanical systems (MEMS). Current MEMS technology allows the implementation of approximately 50 double-posts single-throw relays on an 80-pin chip. Each MEMS chip has a size of approximately 19mm (3/4 ) × 19mm (3/4) on a regular 279mm (11 ) × 457mm (18) board of an automatic cross-connection switch (AXC). It can accommodate about 150 chips and control circuits and internal connection circuits.

Even with MEMS technology, cost and space are key considerations in AXC placement at remote hubs. Service providers still face the question of whether to deploy a large system that offers higher coverage but higher cost or a smaller system that offers lower cost but lower coverage. Given these considerations, it is highly desirable to reduce the number of intersections and thus the number of MEMS cross-connecting chips in AXC, thereby reducing both the cost and spatial requirements of the AXC switch.<patcit num="1"><text>U.S. Patent Application No. __ (Reference No. LCNT / CHU 8-13-8)</text></patcit>

<p> Therefore, the inventor recognized the need for a device that would help reduce dispatch to remote hubs.</p>

<p> In one embodiment, an analog cross-connection (AXC) switch is placed to further reduce the number of hub cross-connections. AXC cross-connects subscriber lines to feeders under remote control, eliminating the need for dispatch on these lines.</p><p> The first step in reducing intersections is to place AXCs according to the penetration rate and churn rate of the line group (first line, second line, etc.). Each group of subscriber lines is evaluated separately and, for placement decisions, connect the subscriber line directly to the feeder, connect the subscriber line to an AXC switch, or terminate the subscriber line group at a hub. The terminated line is connected to the feeder via dispatch, if necessary.</p><p> In one embodiment, further reduction of intersections can be achieved by dividing the AXC into smaller but separate switching modules. This reduction comes at the expense of connecting multiple feeders with AXC.</p><p> Further reductions in normatively configured AXC switches can be provided by disassembling AXC into two subsystems. The disassembled system has far fewer intersections than the original system, which means it has a lower cost and requires less space. Lower cost and space savings are both critical features of remote hubs.</p><p> In another embodiment of the invention, one of the disassembled subsystems is replaced by a splitter, further reducing the number of intersections required. The trade-off between the two embodiments is the cost of the splitter vs. the cost of the AXC switch. Further reductions in the number of intersections of AXC switches can be provided by segmenting AXC switches. Partitioning can be performed on either the undisassembled AXC switch or the disassembled AXC switch. The method of partitioning can be practiced exemplary with respect to the disassembled embodiments to further reduce the number of intersections, but at the expense of more DSLAM ports or lower coverage rates.</p><p> The present invention is applicable to POTS services, DSL services (with or without line sharing), and hubs that provide integrated POTS / DSL solutions. Therefore, the present invention helps reduce spatial constraints at the hub by reducing the size of intersections and switches. The teachings of the present invention can be readily understood by examining the following detailed description with the accompanying drawings.</p><p> To facilitate understanding of the present invention, the same reference numerals are used where possible to specify the same elements that are common to a plurality of drawings. In addition, unless otherwise specified, the alphabetical subscript associated with the code represents an integer greater than 1.</p>

The present invention provides methods and devices for disassembling switches (ie, automatic cross-connection (AXC) switches) based on their static connection characteristics. The present invention can determine the optimal design of an AXC that must be located in a particular hub for POTS (plain old telephone system) services and digital subscriber line (DSL) services.

Further, the present invention decomposes one switch into a plurality of switching modules. By doing so, the number of intersections is greatly reduced. As a result, both the cost and size of the AXC are reduced, which is a critical factor in the placement of such switches in the hub.

The present invention presents AXC switch technology (especially robotics, MEMS) and, among other switch configurations, full matrix, 3-stage Clos (eg, SNB (strictly non-blocking) configurations, AR). It is independent of how each switching module is implemented, such as by using a (rearrangeable) configuration and a WSNB (wide-sense non-blocking) configuration. , POTS service, DSL service (with or without line sharing), and combinations thereof (ie, integrated service).

FIG. 1 shows a block diagram of an exemplary telephone network environment 100 suitable for supporting the present invention. Network 100 is exemplified as a telephone service provider (SP) network that provides POTS services and digital subscriber line (DSL) services to clients (customers). Network 100 includes multiple subscriber premises 102 that are selectively coupled to telephone exchange 170 and / or packet-switched network 150 (eg, the Internet) via a remote hub 120.₁From 102_n(Collectively 102 on the subscriber premises) is included.

Specifically, each subscriber premises 102 includes conventional analog and digital equipment (eg, telephones, computer devices, and the like) for subscribing to POTS and / or DSL services. Each subscriber premises 102 is coupled to a remote hub 120 via one or more subscriber lines 111, as described in detail below. For example, a subscriber receiving POTS services typically has one or more telephones 104 coupled to subscriber line 111 (ie, a "copper loop") via telephone line 107, and this subscription. The personal line 111 is coupled to the service provider's remote hub 120. Note that each subscriber line on the subscriber premises 102 will be associated with a unique telephone number when activated.

Human voice requires only a small amount of bandwidth. Most telephone networks allocate 4KHz bandwidth to support voice channels. However, copper loops can support much more bandwidth than 4KHz. By taking advantage of this extra bandwidth, copper loops can be used to carry high speed data through the use of modulators and demodulators (modems). Modems used in subscriber copper loops are referred to as digital subscriber line (DSL) modems.

Subscribers to the DSL service may use a computer device 106 (eg, a laptop, desktop, or other computer device capable of processing digital information) coupled to a DSL modem 108 over a DSL line 109. it can. The DSL modem 108 provides connectivity to the subscriber line 111 by modulating the packetized information into an analog signal for transfer over the subscriber line 111. There are numerous versions of DSL modems that represent the continued progress of this technology. Currently the predominant DSL modem is ADSL (Asymmetric DSL), which operates from 26KHz to 1.1MHz. With ADSL, the data rate in the downstream direction differs from that in the upstream direction because most consumers receive more data than they send when they access the Internet. The next generation of ADSL is generally called VDSL (ultra-high speed DSL). VDSL operates from 138KHz to 12MHz. VDSL has a maximum data rate of 51.84 Mbps for downstream traffic and 2.3 Mbps for upstream traffic.

Both ADSL and VDSL are designed to allow POTS to share the same line via frequency division multiplexing. In the description and implementation of the present invention, the DSL service is provided by only a single subscriber line 111. That is, only one DSL modem 108 is provided for each subscriber premises 102.

For subscribers receiving both POTS and DSL services, a splitter 110 with one low-pass filter (LPF) 112 and high-pass filter (HPF) 114 is installed on the subscriber's premises 102. As can be seen from FIG. 1, voice traffic from the telephone is coupled to the LPF 112 via the telephone line 107, and data traffic from the DSL modem 108 is coupled to the HPF 114 via the DSL line 109. Therefore, the presence of the splitter 110 addresses both types of services (POTS and DSL) over a single subscriber line 111.

The remote hub 120 is installed in a suitable location nearby (eg, manholes, utility poles, cabinets, among other locations) that provides service connectivity to multiple subscriber premises 102. Each hub 120 typically includes at least one crossover switch 122 and, optionally, a DSLAM (Digital Subscriber Line Access) where DSL services are provided, as shown in FIG. A modem, digital subscriber line access multiplexer) 130 can be included. The AXC switch 122 can facilitate the transfer of analog signals between subscriber premises 102 for both POTS and DSL services.

Specifically, the remote hub 120 has two termination frames 124 that terminate the subscriber line 111 from the subscriber premises 102 and the feeder line 123 from the central office 160.₁And 124₂(Collectively end frame 124) is included. Typically, the termination frame 124 can terminate and cross-connect thousands of subscriber and feeder lines.

A suitable wiring connection (ie, crossover point) 126 of the line is made between the two frames 124 to provide end-to-end connectivity from the subscriber premises 102 to the central office 160. At least some of the intersection points 126 are automatic intersection connection (AXC) switches 122._pIt can be provided via one or more switches such as, where p is an integer greater than 0. Subscriber premises 1 102₁And the first terminal frame 124₁Although only a single subscriber line is shown between, one of ordinary skill in the art will recognize that this single subscriber line 111 represents a signal line pair. In addition, subscriber premises 1 102₁Although only one subscriber line pair 111 is shown between and the AXC switch 122, one of ordinary skill in the art can combine a single AXC switch 122 to, for example, hundreds of subscriber line (pair) 111. I will admit. In fact, there are usually 3-5 pairs of subscriber lines to each subscriber premises (household) 102, so a single hub may terminate 1500 to 2500 subscriber lines 111. ..

Similarly, the second termination frame 124₂Although only a single feeder 123 is shown between and the telephone office 160, one of ordinary skill in the art will appreciate that this single feeder 123 also represents a signal line pair. Further, although only one feeder pair 123 is shown between the single AXC switch 122 and the central office 160, those skilled in the art will route the single AXC switch 122 to, for example, the central office 160 later. It will be admitted that it can be coupled to hundreds of feeders (pairs) 123.

When implementing the line sharing option (ie, the hub also facilitates DSL service), the remote hub 120's DSLAM 130 is an access point 126 between the subscriber line side 111 and the feeder side 123 of the AXC switch 122. Connected as. In one embodiment shown in FIG. 1, the DSLAM 130 includes a splitter 140, a DSL modem 132, and an optional router 134. Splitter 140 includes an LPF filter 142 that passes a low frequency voice (POTS) signal to the telephone exchange 170 for routing to another subscriber premises 102.

Specifically, the LPF 142 ensures that the spectrum between the phone 104 and the splitter 140 operates only from 0KHz to 4KHz. Splitter 140 also includes an HPF filter 144 that ensures that the spectrum between the DSL modem 108 and hub 120 splitter 140 operates only in the high frequency band. The HPF144 passes the high frequency DSL signal to the DSL modem 132, where the received DSL analog signal is demodulated in packet form (eg, IP packet) and forwarded to router 134 for routing. To. The router 134 transfers the packetized information to a specific destination node via a high-speed data line 129 associated with a packet switching network such as the Internet 150, an intranet, or a combination thereof.

Although the splitter 140 is shown exemplary to be implemented within the DSLAM 130, one of ordinary skill in the art will appreciate the splitter 140 as a separate filtering unit, exemplary at a location close to the DSLAM 130, within the hub 120. You will understand that it can be installed elsewhere.

The AXC switch 122 is a remote hub 120 and can be implemented in various ways. For example, given an intersection switch 122 with M input ports and N output ports, the simplest implementation is a full matrix of rectangular arrays of M × N intersections. For crossed connections with N inputs and N outputs, N²There are several intersections. This architecture is inefficient when a large number of subscriber lines and feeders are switched together. This architecture is also not scalable, as the number of intersections increases quadratic for N inputs and outputs. However, due to its extremely simple configuration, this architecture is typically used as a benchmark when comparing different switch sizes and architectures.

Instead, the Clos switching architecture for non-blocking switches requires far fewer intersections. As conventionally known in the art, Clos switches consist of multiple stages, the simplest of which has three stages. A multi-stage architecture can be developed using an exemplary 3-stage Clos network as a basic component. As the number of stages increases, more reductions can be achieved for large Ns. However, for each of the additional stages, there is a need for additional cost of wiring between stages as well as more complex control algorithms for routing through the successful stages of the switch. Each stage also adds attenuation to the signal, which imposes a practical limit on the number of stages that can be achieved. Clos networks are not the only switching architecture. Other switching architectures are possible, each with different characteristics.

In addition, there are three types of non-blocking switches. The first type is the SNB switch. With respect to the SNB switch, the user can always establish a new connection without disturbing the current connection, regardless of the current state of the switch. The second type of switch is an AR (replaceable) switch. With respect to AR switches, users can always establish new connections. However, the user may have to relocate some of the current connections to handle the new request. The third type of switch is the WSNB switch. Related to each WSNB is a routing algorithm, which is used to establish a connection through that switch. For WSNB switches, the user can always establish a new connection without disturbing the current connection, as long as the previous connection was routed using the specified algorithm. These non-blocking type switches require fewer intersections than full matrix switches when the number of input (or output) ports exceeds 36 when implemented within a 3-stage Clos switch.

The AXC switch 122 can be formed by robotics, relays, and in preferred embodiments microelectromechanical systems (MEMS), as conventionally known in the art. In any embodiment, the AXC 122 must support current levels of 250 to 300mA, which determines the minimum size of MEMS chip wiring and miniaturized relays. However, the invention is independent of the technology and implementation of the AXC module (Full Matrix, 3-stage Clos SNB, 3-stage Clos AR, etc.).

The service provider's telephone office (CO) 160 includes a controller 162 and a telephone exchange 170. The telephone exchange routes the voice signals associated with the POTS service between the subscriber premises 102 in a conventionally known manner.

Specifically, the controller 162 is connected to the AXC switch 122 on the remote hub 120. In one embodiment, one of the feeders 123 can be used to provide connectivity to hub 120. An alternative would be to use an integrated router on the DSLAM. Through controller 162, a support representative at CO 160 can cross-connect a subscriber line connected to the AXC switch 122 to a feeder connected to the same AXC. This capability will allow SPs to significantly reduce the number of dispatches to manually interconnect lines at the remote hub 120.

Another function performed by controller 162 is one or more AXC switches on a particular hub, including the size of the AXC and a particular set of subscriber and feeder lines that must be connected to the AXC switch 122. Recommend the optimal configuration of 122 to the service provider. This recommendation is based on a number of other factors, including service subscription and churn rates at the hub and the cost of dispatch, among other considerations.

For example, hubs located far from the dispatch office usually have high dispatch costs. Therefore, the placement of the AXC 122 on the hub 120 is more attractive. This second feature is an offline feature that does not require connectivity to AXC. For a detailed understanding of connectivity optimizations in controller 162 and hubs, U.S. Patent Application No. __ transferred to the same assignee as this application, the entire contents of which are incorporated herein by reference. Please refer to the specification (reference number LCNT / CHU 8-13-8).

In the United States, the number of subscriber lines 111 entering the subscriber premises 102 is between three and five. In future examples and discussions, it is assumed that each subscriber premises 102 has four subscriber lines 111 (ie, since each subscriber line represents a signal line pair, there are actually eight signal lines). The number of feeder lines can be the same as, less than, or more than the number of subscriber lines 111 connected to the hub 120. However, as a practical matter to save resources and costs, the number of feeder lines 123 is usually less than the number of subscriber lines 111 on the hub 120. For example, about 1200 feeders 123 are typically suitable for supporting about 2000 subscriber lines 111 per hub 120.

For an exemplary profile with 500 subscriber premises and 4 subscriber lines per subscriber premises, in one embodiment, the direct solution to provide connectivity is to place an AXC 122 of size 2000 x 1200. Here, 2000 × 1200 represents the number of subscriber lines and the number of feeder lines, respectively. Note that the full matrix embodiment of this switch requires 2.4M intersections, while the above 2000x1200 switch 3-stage AR Clos embodiment results in approximately 217K intersections. To implement 217K intersections using current technology, MEMS technology supports about 50 intersections per MEMS chip, with each board holding about 150 chips. This results in 7.5K intersections per substrate. Therefore, about 29 boards (217K / 7.5K) mount a 2000x1200 switch with 217K intersections. Such a large number of cross-connected boards exceed the size / space constraints of a conventional remote hub 120. The focus of the present invention is to identify techniques for designing AXCs so that the number of intersections is dramatically reduced.

The present invention will first be described by showing the case where the service provider provides only the POTS service and at least one AXC switch 122 is deployed to eliminate the dispatch service provision for the POTS service. The present invention will then be described in the case where the DSL service is also provided by the service provider.

Note that the four subscriber lines for each household 102 have different characteristics. The first line has a very high penetration rate, but a very low churn rate, as most households usually subscribe to one POTS line. For example, the first line is usually disconnected when a subscriber moves, but the next resident subscribes to the service when they move. On the other hand, the fourth line usually has a very low penetration rate, but a very high churn rate.

As an example, the characteristics of a service provider's line can be identified as shown in Table 1 below. Note that the values shown in Table 1 are for illustration purposes only.<tables num="1"><img file="JP5111749B2_D0001.tif" /></tables>

FIG. 2 is a graph representation 200 showing the wiring policy at the remote hub 120. Graph 200 includes ordinate 202 (high, medium, and low) representing penetration and abscissa 204 (high and low) representing churn. The policy for dealing with a particular line depends on the characteristics of that line. Specifically, for a line with a high penetration rate, the subscriber line is directly connected to the feeder line regardless of the churn rate. This policy is applicable to subscriber line 1 in the above example.

For lines with low penetration rates, regardless of churn rate, the policy is to terminate the subscriber line at the remote hub 120 until needed. When needed, a dispatch will be sent to connect the subscriber line to the feeder. The fourth line in the above example fits this characteristic. In the above example, there are 500 fourth lines. On average, about 50 (10%) of this 4th line is active. These 50 lines bring about 15 dispatches (50 x 30%) each year. This policy provides that it is not economical to eliminate this 15 dispatches by using the AXC 122.

Lines with average penetration and average or high churn are candidates for connection to AXC Switch 122. The final decision is based on economic factors, among other factors, such as dispatch costs, AXC equipment costs, and income generated from quick delivery. The second and third lines in this example fall into this category.

Lines with average or low penetration rates as well as low churn rates are treated as fourth lines. That is, this subscriber line is terminated at the remote hub 120. Send dispatches to remote hubs to connect subscriber lines to feeders when needed. Note that the graph in Figure 2 is described in detail in US Patent Application No. __ (reference number LCNT / CHU 8-13-8).

FIG. 3 is a block diagram of an exemplary wiring policy for a hub. Specifically, four exemplary groups of subscriber lines 111 are the first termination frame 124.₁It is illustrated in combination with. Continuing the above example of 500 subscriber premises supported by hub 120, where each subscriber premises has 4 subscriber lines, a total of 2000 subscriber lines are in the first frame 124.₁Combined with. Similarly, the four groups of each feeder line 123 are the second termination frame 124.₂Combined with. Note that the number of feeder lines 123 may be less than the number of subscriber lines. Group 1 G1 302₁Is related to the first of the four subscriber lines to each premises 102. Therefore, the first group G1 302₁Has 500 first subscriber lines. Similarly, Group 2 G2 302₂Is related to the second of the four subscriber lines to each premises 102. Therefore, the second group G2

302₂Also has 500 second subscriber lines, and so on. Group 1 G1 302₁Is illustrated as having, by way of example, having 500 subscriber lines directly coupled to a feeder. In addition, subscriber line group G4 302₄Is the first terminal frame 124₁It is terminated with. Group 4 G4 subscriber lines are connected to feeders when needed. This wiring is performed manually by dispatching field personnel.

In this example, it is assumed that the placement of AXC is justified for the 3rd and 4th line groups. AXC (122)₁And 122₂) Are 500 x 327 and 500 x 137, respectively.

The size of AXC (122) is determined through the important concept of coverage rate. The coverage rate (for a particular value on the feeder) is the percentage of the churn rate that can be processed by the AXC switch without the need to start dispatching. Therefore, the percentage value when dispatch is required is the coverage rate minus 1 (1-x%, where x 0).

Consider the above example shown in Figure 3. Second subscriber line 302₂All 500 of AXC 122₁Suppose it is desirable to connect to. Therefore, AXC 122₁A determination must be made regarding the appropriate number of feeders connected to. Larger AXCs are required when a large number of feeders are connected, which increases the cost of the system. On the other hand, only a few feeders are AXC 122₁When connected to, the number of second subscriber lines may exceed the total number of feeders in hub 120 and may require dispatch to provide service.

The number of feeders required to support the desired coverage rate can be calculated from the penetration rate. Suppose there is a group of N subscriber lines with the same characteristics with a penetration rate of "p". According to the example in Table 1 above, subscriber line 2 has an exemplary penetration rate of 60%. The number of lines N of this second group is 500 subscriber lines (N = 500). This second group 302₂The number of lines that are exemplarily considered active with respect to follows a binomial distribution, as shown in Table 2.<tables num="2"><img file="JP5111749B2_D0002.tif" /></tables>

This binomial distribution can be approximated by a normal distribution, and a table of coverage rate vs. number of feeders can be easily constructed as shown in Table 3. Note that the value identified by the asterisk (*) is the value obtained by using a traditional normal distribution table.<tables num="3"><img file="JP5111749B2_D0003.tif" /></tables>

Therefore, if an AXC of size 500 x 330 is deployed to support a second line, the switch eliminates 99.5% of dispatches in service delivery. AXC with a size of 500 x 335 eliminates 99.9% of dispatch. Note that there are a total of about 30 dispatches with a 10% churn rate each year. For all practical purposes, all dispatches are eliminated. Similar exemplary values are shown in Table 4 for the third line (eg, group 302 in Figure 3).₃) Can be determined.<tables num="4"><img file="JP5111749B2_D0004.tif" /></tables>

Continuing with the example shown above, the first subscriber line group 302₁Has a high penetration rate and is directly connected to the feeder. 4th subscriber line group 302₄Has a low penetration rate and is exemplifiedly terminated at a remote hub. In addition, the second and third subscriber line groups 302₂And 302₃Are each AXC 122₁And 122₂It is a candidate for connection to AXC switch 122 such as.

As shown in Figure 3, by selectively arranging individual line groups as needed, the number of intersections is a single AXC (eg, connecting all subscriber lines and feeders). Reduced compared to full matrix) placement. A comparison of the number of intersections is illustrated in Table 5 below.<tables num="5"><img file="JP5111749B2_D0005.tif" /></tables>

By comparison, a single AXC that connects all subscriber lines to the feeder is 2400K, 381K, and 217K for the Full Matrix switch, Clos 3-stage SNB switch, and Clos 3-stage AR switch, respectively. Has an intersection of. It should be noted that the configuration of Figure 3 is shown and discussed for illustration purposes only, and those skilled in the art will appreciate that economic analysis (eg, cost / profit analysis) can provide other hub configurations. For example, an economic analysis shows that the second line (second group G2) 302₂AXC 122₁Must be connected with the third line 302₃May indicate that it must be terminated at a remote hub. In addition, other configurations are possible (eg, second and third subscriber lines 302).₂And 302₃Note that both are terminated at the remote hub).

Assuming a coverage rate of 99%, the second line group 302 in this example₂An AXC with a size of 500 x 327 is required to support. Note that in some applications this switch may be too large or expensive to implement. Another technique for further reducing switch size is through segmentation. That is, the selected AXC switch 122 can be further subdivided to reduce the number of intersections required. However, the number of feeders connected to AXC will increase. An exemplary 500 subscriber line switch can be optionally subdivided into smaller groups, such as four groups, each of which has 125 lines.

In order to maintain the dispatch at the desired 1% ratio, each segment must have a dispatch rate of 0.25% or a coverage rate of 99.75%. Therefore, 91 feeders must be used for each section. This results in four AXC switches, each with a size of 125x91. Using the full matrix embodiment as a reference, the number of intersections is 500 x 327 = 163.5K when not classified, but 4 x 125 x 91 = 500 x 91 = 45.9K when classified. , This is about 1/3 of the original size.

FIG. 4 shows the AXC switch 122 classified according to the principle of the present invention. Specifically, the AXC switch 122, illustrated in FIG. 3, having a size of 500 x 327.₁Is four compartments 402₁From 402₄Each compartment has a subscriber line-to-feed line size of 125 x 91. Note that 91 feeders were calculated based on the fact that there are 4 compartments. Specifically, each parcel must have a dispatch rate of 0.25% in order for the total to have a dispatch rate of 1%. As a result, each parcel should have a coverage rate of 99.75% instead of the usual 99.0%. For 99.75%, about 2.72 times σ (standard deviation) was used to calculate the number of feeders in a manner similar to that mentioned above.

Therefore, there is a considerable savings on the number of intersections. However, AXC must be designed with division in mind to enable cost savings. In addition, the arrangement of the four compartments increased the number of feeders from 327 to 364, which may not be acceptable in situations where the feeders are very limited. Moreover, the management of partitioned nets can be slightly more complex. Therefore, there are considerable trade-offs as to whether the division is useful or not. However, traditional business case type analysis on the net present value or break-even period can be used to determine whether to categorize the switch.

The present invention has been described so far when POTS services are provided by service providers on hubs. The present invention also addresses when the DSL service is available on the subscriber premises. Note that POTS services usually have a low churn rate, while DSL services usually have a high churn rate. Specifically, DSL services have a moderate penetration rate, but a high churn rate of about 20% each year. Therefore, the AXC switch, together with the present invention, is useful to help reduce the number of dispatches of DSL services.

Continuing with the above example, in addition to the previous assumption of 500 serviced homes, 4 subscriber lines per home, and 1200 feeders, additional assumptions are needed when DSL services are offered. is there. The first assumption is that there is at most one DSL subscriber per household (ie, subscriber premises). This is a reasonable assumption, as the DSL modem 108 on the subscriber's premises will eventually be connected to router 134. Specifically, all of the customer terminal equipment is connected to this router and thus to the DSL service via the local area network.

The second assumption is that the penetration rate of DSL is 50% per household. This assumption is based on marketing expectations that 50% of households expect to use some form of DSL service.

The third assumption is that the DSLAM 130 is located on the remote hub 120. Placing the DSLAM 130 on the remote hub 120 reduces the distance between the DSLAM 130 and the customer premises 102. By reducing the distance between these two points (DSLAM and subscriber premises), DSL services can operate near maximum speed, which allows service providers to offer more services (such as video). Will be.

Another assumption is that DSL line sharing options are supported. Specifically, service providers can provide both voice (POTS) and digital (DSL) services over a single subscriber line. Note that these assumptions are for illustration purposes only and the present invention applies without these assumptions.

FIG. 5 is a block diagram of an exemplary switch suitable for providing a DSL service with a line sharing option. An exemplary switch 122 is illustrated as coupling a group of subscriber lines 302 to a first plurality of feeder lines 502. In addition, the DSLAM 130 is coupled to a second plurality of feeders 506 . The DSLAM 130 includes a high-speed data line 129 that can be coupled to a router or switch (not shown) and multiple feedback lines 504 that are fed back as input lines to the AXC 122.

Specifically, since each household has at most one DSL line, only one subscriber line from each household (line group 302) needs to be connected to the AXC 122. If the customer subscribes only to the POTS service for this subscriber line, the subscriber line (in line group 302) is cross-connected to one of the feeders in feeder group 502. When the subscriber line is used for the DSL service with or without the POTS service, the subscriber is cross-connected to the line of group 506. The line of group 506 is connected to the DSLAM 130. If you are not subscribed to the POTS service, the DSL signal will be terminated on the DSLAM 130. When the line sharing option is used for a particular line (ie, the POTS service is also provided on that line), the DSLAM

The 130 splitter 140 (FIG. 1) separates the POTS signal from the multiplexed signal sent from the subscriber premises 102 and transfers the POTS signal to the AXC 122 using line group 504. The AXC 122 cross-connects its line (of group 504) to the feeder of group 502.

As mentioned above for step 514 of method 500 of the POTS service, the size of the various line groups is first determined. An exemplary group 302 includes 500 lines, one line per household. The first step is to determine the number of DSL ports required for a given DSL coverage rate. The techniques described above for coverage rates for POTS services can also be applied to DSL services. For a penetration rate of 50%, the average value (m) of 500 households is 250. The standard deviation (σ) is about 11.2 (ie,<maths num="1"><img file="JP5111749B2_D0006.tif" /></maths> 11.2). For a DSL coverage rate of 99.9%, the number of DSL ports is equal to about 285 (ie, (m + 3.1σ) 285 DSL ports, where m = 250). Therefore, 285 feeders from the AXC 122 to the DSLAM 130 are required.

After that, a determination is made regarding the number of feeder lines of the group 502 that must be connected to the AXC 122. Group 502 represents connectivity for POTS services. In order to calculate the number of feeder lines in group 502, the normal POTS penetration rate of the subscriber lines in group 302 is required. This is the first line group with high penetration (eg group 302)₁), The optimum number of feeders in group 502 is equal to the number of subscriber lines, which is 500 feeders in this example.

The size of line group 504 depends on the prevalence of line sharing options among DSL subscribers. The simplest design is to assume this rate is 100%. In this case, the line group 504 has the same size as the line group 506, that is, 285 lines. This configuration has all the conditions (for example, line group 302, among other configurations).₁The second line group 302₂This configuration is called a normative configuration (that is, a normalized configuration) because it works with changing to, reducing the line sharing option penetration rate. That is, the size of the line group 504 is the same as the line group 506 unlike the non-normative configuration, and in the non-normative configuration, the line group 504 is different from the line group 506.

Therefore, for the normative configuration, the number of upstream ports on the AXC 122 is 500 + 285 = 785 ports. Similarly, the number of downstream ports is 785. The system size of this AXC is 785 x 785. The final step is to perform a cost / profit analysis to see if the placement of AXC can be economically justified.

If the DSL service does not support the line sharing option, line group 504 is not needed and AXC has a reduced size of 500 x 785. The normative configuration (785 x 785) is a bit larger, but in other scenarios (eg, second line group 302).₂, 3rd line group 302₃Etc.), so it represents the most flexible configuration.

Note that you must have the same number of feeders connected to AXC, equal to the number of subscriber lines connected. However, to do so, the service provider first asks for line group 302 (eg, first line group 302).₁Or second line group 302₂) Must be determined. The SP also needs to determine POTS, DSL penetration and line sharing options. The details of the form for determining the size will be described below.

Note that SP can also provide VoIP (voice over IP) over DSL. When DSL subscribers use VoIP services over the second line instead of DSL services, the line sharing option will likely be much less prevalent and the reduction in AXC switch size will be more apparent.

The classification technique is also applicable to DSL services (ie, AXC shown in Figure 5). However, the number of DSLAM ports required increases depending on the division. Normally, the size of the DSLAM 130 is severely limited by the remote hub 120 due to hub space constraints. Therefore, the division may not always be practical in many cases.

DSL services are usually provided with POTS services. You can merge the AXCs of both solutions. The merged AXC is the AXC whose size is the maximum of the two components. Consider the POTS example in Figure 3. If DSL service is also provided, a more efficient configuration is AXC 122₁Is to be merged with the normative structure of the DSL shown in Figure 6.

FIG. 6 is a block diagram showing a logical wiring arrangement that integrates POTS and DSL services in an exemplary hub according to the principles of the present invention. That is, an integrated solution is shown in Figure 6. Figure 6 shows the second line of Figure 3, 500 x 327 AXC 122.₁Is the same as in Figure 3, except that it is replaced by the 785 x 785 size AXC switch shown in Figure 5.

Specifically, Group 1302₁The exemplary 500 first subscriber lines are directly coupled to the 500 feeders. Group 2 302₂An exemplary 500 second subscriber line, AXC switch 122 with a size of 785 x 785₁Coupled to, this is the switch 122 from the output port of the DSLAM 130, with 500 second subscriber lines where this switch is coupled to 500 feeders.₁It means accommodating 285 DSL lines that are fed back to the input port of. Group 3 302₃An exemplary 500 third subscriber line is an AXC switch 122 with a size of 500 x 137.₂This means that 500 second subscriber lines will be coupled to 137 feeders. Finally, Group 4 302₄The exemplary 500 fourth subscriber lines are terminated at hub 120 and coupled to feeders by dispatch as needed.

In most cases, the normative AXC configuration for DSLs is a larger switch when compared to AXCs for POTS services. The end result is that the normative configuration acts as an AXC for one of the line groups (usually the first or second line). Therefore, it is advantageous to further reduce the number of normative switch intersections.

Further optimization can be provided by reducing the size of the normative switch. In one embodiment of the invention, the AXC switch 122 of the remote hub 120 can be disassembled into two subsystems. The disassembled system has far fewer intersections than the original system, which results in lower costs and requires less space. The disassembly of the AXC switch is illustrated and illustrated with reference to FIGS. 7-12 with respect to the exemplary conditions, assumptions, and examples provided exemplary.

7A and 7B collectively show a block diagram showing the disassembly of the exemplary switch 122 of FIG. 5 according to the first embodiment of the present invention. As can be seen in Figure 7A, a 785 x 785 switch is coupled between a group of 500 subscriber lines 302 and a group of 500 feeder lines 502. Further, the switch 122 has, exemplary, a group 506 of 285 DSLAM lines coupled to the DSLAM 130, and a group 504 of 285 lines is fed back to the switch 122.

With reference to Figure 7B, this exemplary switch 122 disassembly is two structurally identical switch subsystems, one on each side of the DSLAM 130.₁And 732₂It results in a disassembled switch 730 with (collectively switch subsystem 732). First subsystem 732₁Is located between the subscriber line 302 and the DSLAM 130 and is the first diagonal (or Y) switch 734.₁And 1st AXC module 736₁including. Second subsystem 732₂Is located between the feeder line 502 and the DSLAM 130, as well as the second diagonal (or Y) switch 734.₂And 2nd AXC module 736₂including.

FIG. 8 is a graphical representation of the diagonal switch 734 suitable for use in the present invention. The graphical representation of diagonal switch 734 includes N input lines 802 and 2N output lines 804. Specifically, each input line 802 is at each crossing point 806, with two pre-specified output lines 804.₁And 804₂Connected to any one of. For example, first input line 802₁Is the first crossing junction 806₁And the first output 804₁₁And second output 804₂₁Connected to. Similarly, the second input line 802₂Is the second crossing junction 806₂And the first output 804₁₂And second output 804₂₂Connected to, and so on. Therefore, a diagonal switch with N inputs also has N intersections, one per line. None of the other input lines except those specified cannot be cross-connected to other output lines. Note that term inputs and outputs are only descriptive terms. While switching the analog signal, the two signal lines are switched at the same time. In addition, one of ordinary skill in the art will recognize that current can flow in either direction. Therefore, the diagonal switch acts as a much simpler cross-connection.

With reference to the exemplary disassembled switch 730 shown in Figure 7B, the subscriber line for group 302 is the first diagonal switch 734.₁It is connected to the. 1st Diagonal Switch 734 for subscribers who subscribe to only POTS services (or not subscribe to any services)₁Peer the line through line group 742 (ie, second) diagonal switch 734₂Cross-connect to. Note that only a single line 742 is shown between the diagonal switches 734, but this line 742 represents multiple cross-connect lines, which in number equals the amount of subscriber lines. For example, if the subscriber line group 302 has 500 subscriber lines, the group 742 has 500 crossovers and the group 502 has 500 feeders.

2nd diagonal switch 734₂Cross-connects line 742 to the group 502 feeder. Therefore, groups 302, 742, and 502 are all the same size. There is a one-to-one fixed mapping of wires between these three groups. That is, the i-line of group 302 is always mapped to the i-line of group 742, and the i-line of this group 742 is mapped to the i-line of group 502.

First diagonal switch 734 when a customer subscribes to a DSL service (for example, line i)₁1st AXC module 736 via line group 744₁Cross-connect to. 1st AXC module 736₁Cross-connects this line to one of the available ports on the DSLAM 130 through line group 738. A DSLAM 130 splitter (not shown) separates the POST signal and sends the POST signal through line group 740 to the second AXC module 736.₂Transfer to. 2nd AXC module 736₂Sends this signal to the second diagonal switch 734₂Switch to the corresponding line (line i) of line group 746 to. 2nd diagonal switch 734₂Switches this signal to the corresponding feeder (line i) of line group 502.

The size of the AXC module 736 of the disassembled switch 730 is 500 x 285, respectively, in this example. The number of intersections for each AXC module 736 is approximately 142.5K for the full matrix, 45.5K for the 3-stage Clos SNB, and approximately 27K for the 3-stage Clos AR switch. Each diagonal switch gives an additional 500 intersections and there are two such subsystems. All together, the total number of intersections is about 286K for the full matrix switch, 92K for the 3-stage Clos ANB, and 55K for the 3-stage Clos AR switch. Therefore, the savings are significant (53.6%) in the full matrix embodiment, modest (32.8%) in the 3-stage Clos SNB, and slightly less (21.4%) in the 3-stage Clos AR.

The mathematical foundation of decomposition is shown in Figure 7B. Figure 7B further shows that the state of the cross-connection can be represented by the matrix 702. The rows of matrix 702 represent the lines on one side of the matrix, and the columns represent the lines on the other side. For convenience of explanation, the subscriber line side is referred to as an input, and the feeder line side is referred to as an output.

The connectivity of normative configurations has certain special characteristics. Specifically, the DSL signal does not pass through the DSLAM twice, so the input line from the DSLAM 130 is not connected to the output line to the DSLAM 130. Since the number of subscriber lines is equal to the number of feeder lines, a one-to-one mapping can be established between the subscriber lines and the feeder lines. The POTS signal of the subscriber line can always be transferred to the designated feeder.

Given these properties, the state matrix M 702 of AXC 122 has the special form shown in Figure 7B. The matrix M 702 contains four submatrixes 706, 708, 710, and 712. The submatrix 712 that connects the input line from the DSLAM 130 to the output line of the DSLAM 130 is designated as the 0 matrix because these two groups of lines are never connected. The submatrix 706 or J that connects the subscriber line to the feeder is a matrix whose item is 0 when it is not on the main diagonal (ie, the intersection connection point 806 in FIG. 8). However, the item on the main diagonal (806 in Figure 8) can be either 1 or 0 because the subscriber maps to the specified feeder (i-line vs. i-line). There is. If the customer is not a DSL subscriber, that subscriber is connected to the feeder and this item is 1. If the customer subscribes to the DSL, the line will be connected to the DSLAM port and this item will be 0. In addition, a submatrix Q 710 connecting the subscriber line 302 to the DSLAM 130 and a submatrix P connecting the DSLAM 130 to the feeder line 502. 708 is expanded as a complete matrix.

The decomposition described above for Figures 7A and 7B is based on this special matrix structure 702. First subsystem of this disassembly 732₁Is formed by matrix J1 (a copy of submatrix J) 706 and submatrix Q 710. In addition, the second subsystem 732₂Is formed by the submatrixes J2 and P. J1 and J2 are copies of the submatrix J706 and are always equal to each other. Note that the matrices Q and Q are also relevant, and that the submatrix Q 710 is always equal to the transposed matrix of the submatrix P 708. Therefore, the disassembly (ie, the disassembled switch 730) is based on the special connectivity characteristics of the normative configuration of the AXC switch 122.

FIG. 9 is a block diagram showing an exploded view of the exemplary switch 122 of FIG. 5 according to a second embodiment of the present invention. The second embodiment uses a splitter to disassemble the first diagonal switch 734 on Figure 7B.₁And the second subsystem 732₂Effectively replaces to form the disassembled switch 930.

Specifically, the subscriber lines of the selected group 302 (eg, 500 subscriber lines) are connected to the splitter 950. The splitter 950 includes a low-pass filter (LPF) 952 and a high-pass filter (HPF) 954, as described above for the splitters 110 and 140 in FIG. The splitter 950 is located outside the DSLAM 130 rather than being implemented within the DSLAM as described above for Figures 5 and 7B.

Splitter 950 dedicates LPF 952 and HPF 954 to each subscriber line in Group 302. The low pass signal, including the POTS signal from the subscriber premises 102, is transferred to the CO 160 via the feeder line 123. High-pass signals, including DSL signals, are transferred to the AXC 736 via one of the lines 734. The AXC switch 736 cross-connects this signal to an unused port on the DSLAM 130. Therefore, this second embodiment effectively replaces half of the system with a splitter 950.

There are multiple trade-offs between the second resolution (ie, using a splitter) and the first decomposition (ie, using two subsystems). One consideration is the cost difference between the splitter 950 and the AXC module 736. Specifically, splitters are much cheaper to purchase and implement AXC modules. Specifically, in many cases, the DSLAM 130 has already implemented an HPF, while the LPF has already been implemented on a Class 5 voice switch 170 with a CO 160. In this case, the splitter simply splits the signal into two branches, resulting in lower cost.

Another consideration is that the splitter-based configuration (Figure 9) does not support unbundling of copper loops to competitive local exchange carriers (CLECs). When CLEC operates a line, CLEC wants to provide both POTS and DSL services to its subscribers using the same line, so high-pass signals and low-pass signals in the same copper loop. Expect to have both signals. As a result, a dispatch is required to bypass the splitter in this case. This may be acceptable for some service providers, especially in countries where unbundling is not a legal requirement.

FIG. 10 is a block diagram showing the division of the disassembled switches 730 of FIGS. 7A and 7B according to the principles of the present invention. Figure 10 also performs a division for the disassembled embodiment of the switch to further reduce the size of the disassembled AXC switches, such as the disassembled switch 730 in FIG. 7 and the disassembled switch 930 in FIG. It is shown. It should be noted that although the partitioning and disassembling techniques described herein are independent of each other, both techniques can be performed in any order to further reduce the size of the switch.

Referring to FIG. 7B, each subsystem 732 (disassembled switch) can be further reduced by dividing the 500 subscriber lines into four smaller groups of 125 subscriber lines each. Continuing with the current example, using a coverage rate of 99% (or a dispatch rate of 1%), each parcel must have a coverage rate of 0.25% or 99.75%. In this example, each compartment has a mean of 62.5 and a standard deviation (σ) of 5.6. Therefore, a total of 80 DSL ports are required for each partition.

With reference to Figure 10, the decomposition solution (without splitter) is divided into four compartments. Specifically, first subsystem 732₁(Figure 7B) is the four diagonal switches 1034.₁₁From 1034₁₄It is divided into. Similarly, the second subsystem 732₂Has four diagonal switches 1034₂₁From 1034₂₄It is divided into. Each of the diagonal switches 1034 is 125 x 125 in size.

1st AXC switch 736₁The four AXC modules (in Figure 7B) are the four AXC modules 1036.₁₁From 1036₁₄It is divided into. Similarly, the second AXC switch 736₂4 AXC modules 1036₂₁From 1036₂₄It is divided into. Each of the AXC modules 1036 is 125 x 80 in size.

Group 1 302 of 125 subscriber lines with high penetration rate₁(For example, POTS service) is a diagonal switch 1034₁₁, Line 1042₁(Representing 125 lines to connect), and diagonal switch 1034₂₁(125 feeders 502₁125 feeder lines 502 via a line path including)₁Is directly combined with. Similar connectivity, second and third subscriber line groups 302₂And 302₃Provided for 125 subscriber lines. Finally, Group 4 302 with 125 subscriber lines, which also has a high penetration rate.₄(For example, POTS service), but diagonal switch 1034₁₄, Line 1042₄(Also represents 125 lines to connect), and diagonal switch 1034₂₄(125 feeders 502₄125 feeder lines 502 via a line path including)₄Is directly combined with.

For DSL services, each diagonal switch 1034 associated with the subscriber line, whether provided on a common access line or a separate line with the POTS service._1XCorresponds to the line 1044_XThrough each AXC module 1036_1XCombined with (X is an integer greater than 0). For example, diagonal switch 1034₁₁Line 1044₁Via AXC module 1036₁₁Combined with the diagonal switch 1034₁₂Line 1044₂Via AXC module 1036₁₂And so on. Each line 1044_XIs a diagonal switch 1034_1XFrom AXC module 1036_1XRepresents 125 lines to.

Each AXC module 1036_1XIs a line connection 1038_XCombined to the DSLAM 130 via. Each line connection 1038_XIs the AXC module 1036_1XRepresents 80 lines from to the DSLAM 130. Diagonal switch 1034_2XAnd AXC switch 1036_2XA similar arrangement of the second subsystem 732₂Is formed about.

The total number of intersections in this disassembled and partitioned system is 81K for the full matrix, 51K for the 3-stage Clos SNB switch, and 31K for the 3-stage Clos AR switch. This is comparable to the 286K, 92K, and 55K solutions that are unclassified but not decomposed, and the 168K, 71.6K, and 37.2K solutions that are classified but not decomposed.

That is, for comparison, a segmented but undecomposed configuration with the same parameters as those shown herein consists of four subscribers of 125 lines, each connected to a size 205 x 205 AXC. Provide a line group. For such partitioned but undecomposed configurations, the number of intersections in each section is 42K for the full matrix, 17.9K for the 3-stage Clos SNB switch, and 9.3K for the 3-stage Clos AR switch. With a total of four compartments, the total number of intersections in a partitioned but undisassembled system is approximately 168K for the full matrix, 71.6K for the 3-stage Clos SNB switch, and 37.2K for the 3-stage Clos AR switch. .. Therefore, the trade-off for reducing intersections is increasing the number of DSL ports. The alternative would allow the same number of ports, but with a slightly lower coverage rate.

The classification can also be applied to cases where the splitter is used during disassembly. Referring to FIG. 9, the splitter 950 is a diagonal switch 1034.₁And second subsystem 732₂Placed during disassembly instead of. The intersections in this disassembled configuration can be further reduced by the solution shown in Figure 11.

FIG. 11 is a block diagram showing the division of the disassembled switches of FIG. 9 according to the principle of the present invention. Continuing the same example of 500 subscriber lines, 4 subscriber lines per subscriber premises, and 1 subscriber line used for DSL service per subscriber premises, it has a size of 500 x 285. The exemplary disassembled switch is divided into four AXC switches, each with a size of 125x80. The splitter LPF 952 couples the low pass signals associated with the popular POTS service to the 500 subscriber lines of Group 302 to the 500 feeders 502, as described above with respect to FIG.

DSL service is a four-segment AXC module 1160₁From 1160₄Divided between, each module 1160 is coupled to the HPF 954 of the splitter 950 via a high-pass line group 1156 containing 125 lines. Each AXC module 1160 is further coupled to a DSLAM 130 via a line group 1162 containing 80 lines. For example, the HPF 954 has a line group of 1156.₁AXC module 1160 via₁Combined with the AXC module 1160₁In addition, line group 1162₁Combined to the DSLAM 130 via. The same connectivity pattern is the AXC module 1160₂From 1160₄Provided about. The DSLAM supplies the high-speed data line output 129 to the packet switching network 150 as described above with respect to FIG. Therefore, both embodiments of the disassembled switch can be further subdivided to reduce the number of intersections per switch. However, such intersection reductions come at the expense of increased DSLAM port numbers or reduced coverage rates, as mentioned above.

A comparison of all embodiments of the invention with respect to the number of intersections is summarized in the following table. Keep in mind that there are other factors that can influence your judgment. However, the number of intersections is important because it affects both the cost and space of AXC, which is a critical factor in this application. The numbers shown in Table 6 correspond to the exemplary 785 × 785 AXC of the normative configuration exemplified herein.<tables num="6"><img file="JP5111749B2_D0007.tif" /></tables>

12A and 12B collectively show block diagrams showing the decomposition of non-normative switches according to the third embodiment of the present invention. As can be seen from Figure 12A, the number of feeders connected to AXC need not be equal to the number of connected subscriber lines to support DSL services.

However, when deploying a non-normative AXC, the service provider must first determine the characteristics of the subscriber line group to be considered. The SP also needs to determine the penetration rate of POTS and DSL and the line sharing options. Using the same example as described above, assume that POTS and DSL penetration rates are 60% and 50%, respectively. Assume that 80% of all DSL subscribers use the line sharing option. This assumption results in: 40% of the lines support both POTS and DSL services (ie, the line sharing option is used), 20% of the lines only support POTS services, and of the lines 10% support only DSL services and 30% of lines are inactive (ie, not subscribed to either service).

Assuming a coverage rate of 99.5% and referring to Figure 12A, the size of line group 502 is 327 lines and the size of group 506 is 285 lines. Further, the size of the line group 504 is 230 lines. Therefore, the size of the AXC switch is 730 x 611.

This non-normative configuration of the connectivity matrix does not have the simple form shown in Figure 7B. Specifically, the matrix J 706 is not a diagonal matrix. However, the lower right 712 of this matrix is still 0. Therefore, decomposition is still possible, but less efficient, as shown in Figure 12B.

Referring to FIG. 12B, diagonal switch 1234 receives exemplary 500 subscriber lines of group 302, exemplary. The first output of diagonal switch 1234 goes to first AXC switch 1236, which has a size of 500 x 285. The first output is sent via line group 1244, which line group 1244 is formed by 500 lines as described above with respect to FIG. 7B. The output of the 1st AXC 1236 is coupled to the DSLAM 130 via a line group 1238 formed by 285 lines. The DSLAM 130 outputs a DSL signal to the second AXC switch 1210 via a line group 1240 formed by 230 lines. The second AXC switch 1210 has a size of 352 x 339. In addition, the second output of diagonal switch 1234 is also coupled to second AXC switch 1210 via line group 1242 formed by 122 lines. Therefore, the output of the second switch is 339, which is the CO. Combined with 160 voice exchanges 170. Please understand that the size of the switch and the number of lines are provided for illustration purposes only. See U.S. Patent Application No. __ (reference number LCNT / CHU 8-13-8) for a detailed understanding of how to determine the number of input and output lines for each AXC.

In general, the development of normative constructs is simpler than non-normative constructs and has an efficient decomposition. However, the use of non-normative forms is advantageous in cases where there are severe restrictions on the remote hub feeders, either because the non-normative configuration uses fewer feeders or because the line sharing option is less prevalent. There is a possibility of becoming.

Note that the normative configuration is based on the assumption that at most one DSL line is provided per household. In addition, it is assumed that the line will be modeled after the first line of the household, which has a high penetration rate of POTS and line sharing options.

The connectivity of normative configurations has certain special characteristics. Based on this property, the present invention specifies how the remote hub AXC can be disassembled into two subsystems. The disassembled system has fewer intersections than the original system, which means it has a lower cost and requires less space. Both are critical in remote hub applications. It is envisioned that the present invention will be applied as an AXC switch embodiment.

In another embodiment of the invention, one of the subsystems is replaced by a splitter, further reducing the number of intersections required. The trade-off between the two embodiments is the cost of the splitter vs. the cost of AXC. In addition, the splitter-based solution does not support unbundling of copper loops to CLEC. However, dispatch is needed to deal with the CLEC offer.

The method of division can be carried out with both normative and non-normative embodiments by further reducing the number of intersections. However, the division is obtained at the expense of more DSLAM ports or lower coverage rates.

Although the present invention has been described through remote hub application examples, the same principles can be applied to analog cross-connection applications at other locations, such as telephone offices, basements of high-rise buildings, etc. it can.

The above description merely illustrates the principles of the present invention. Therefore, it should be appreciated that one of ordinary skill in the art can practice the principles of the present invention and devise various arrangements within its gist and scope, although not explicitly described or illustrated herein. Furthermore, further, the words of all the examples and conditions presented are useful, primarily to assist the reader in understanding the principles of the invention and the concepts that the inventor contributes to aid the art. It is expressly intended to be for purpose only and must be construed as having no restrictions on such specifically stated examples and conditions. Moreover, all statements herein specifying the principles, aspects, and embodiments of the present invention and its particular examples are intended to include both structural and functional equivalents thereof. ..

<figref num="1">FIG. 6 is a block diagram showing an exemplary telephone network environment suitable for supporting the present invention.</figref><figref num="2">A graphical representation of the wiring policy at a remote hub.</figref><figref num="3">It is a block diagram which shows the logical wiring arrangement in an exemplary hub by the principle of this invention.</figref><figref num="4">It is a figure which shows the AXC switch classified according to the principle of this invention.</figref><figref num="5">It is a block diagram which shows an exemplary switch suitable for providing a DSL service which has a line sharing option.</figref><figref num="6">It is a block diagram which shows the logical wiring arrangement which integrates POTS service and DSL service in an exemplary hub by the principle of this invention.</figref><figref num="7A">It is a block diagram that collectively shows the decomposition of the exemplary switch of FIG. 5 according to the first embodiment of the present invention.</figref><figref num="7B">It is a block diagram that collectively shows the decomposition of the exemplary switch of FIG. 5 according to the first embodiment of the present invention.</figref><figref num="8">It is a graph representation which shows the diagonal switch suitable for use in this invention.</figref><figref num="9">It is a block diagram which shows the decomposition of the exemplary switch of FIG. 5 by 2nd Embodiment of this invention.</figref><figref num="10">It is a block diagram which shows the division of the disassembled switch of FIGS. 7A and 7B by the principle of this invention.</figref><figref num="11">It is a block diagram which shows the division of the disassembled switch of FIG. 9 by the principle of this invention.</figref><figref num="12A">It is a block diagram which collectively shows the decomposition of the non-normative switch according to the 3rd Embodiment of this invention.</figref><figref num="12B">It is a block diagram which collectively shows the decomposition of the non-normative switch according to the 3rd Embodiment of this invention.</figref>