Data communication over power lines

Description of equivalent WO 2005011142 A2

DATA COMMUNICATION OVER POWER LINES

This application is being filed as a PCT International application in the name

of Hunt Technologies, Inc., a U.S. corporation, claiming priority to U.S. Patent

application's nos. 10/626,495 filed 24 July 2003, 10/626,496 filed 24 July 2003 and

10/626,465 filed 24 July 2003.

Technical Field This invention relates generally to data communication, and more

particularly to data communication over power lines.

Background

As is true with most companies, utility companies are striving to reduce

overhead costs, while providing more convenience to customers. For example,

electric companies are migrating from costly and time-consuming manual methods

of determining the amount of power consumed by customers of the power company.

Traditionally, a person periodically came to the customer's home, and requested

entry to read the consumer power usage from a power meter. This type of process

was costly, slow, and intrusive to their customers.

Newer systems provide some level of remote communication between an

endpoint such as an electrical meter and a central location. One such system is an

automated meter reading (AMR) system that utilizes a power line to establish a data

link between a concentrator and endpoint meter reading units positioned downstream

from the substation. The concentrator typically includes a transmitter for transmitting control information to the endpoint and a receiver for receiving data

such as watt-hour information from the endpoint. The endpoint includes a

transmitter, a receiver, and electronics or other circuitry for reading the meter. Other

remote meter reading and data communication systems that use modems, radio

frequency signals, or PLC signals can communicate with only one endpoint at a time

and thus have limited capacity.

These current systems have shortcomings. For example, the capacity of such

systems is limited because the concentrator (or other central processing system if

modems or RF are used) can receive signals from only one endpoint at a given time.

This limitation provides a bottleneck that limits the processing power and flexibility

of the system. Additionally, it limits the number of endpoints that the concentrator

can communicate within a 24-hour period and hence limits the number of endpoints

that can be connected downstream from any given concentrator.

The systems also have little scalability. This limitation is caused by two

factors including the limited number of endpoints that can be connected downstream

from a concentrator and by the manual programming required every time that an

endpoint is added to the system.

Other shortcomings of current AMR and other power line data

communication systems relate to reliability, flexibility, and scalability. For example,

the system needs to be manually programmed each time an endpoint is added. In

another example, if there is a power outage, automated meter reading systems

generally require polling of the endpoints to determine which ones are still

operational. This polling is slow and consumes processing and commumcation

resources. Furthermore, current systems generally do not have the capability of

reestablishing communication between an endpoint and an alternative concentrator if the communication link between the concentrator and the endpoint is disconnected

by intentionally taking the substation offline, through a power failure.

Summary In general terms, the present invention relates to a system for bi-directional

communication over a power distribution system having power distribution lines.

One aspect of the invention relates to a system for receiving and processing

signals received from a plurality of endpoints. Each endpoint includes an endpoint

transmitter in electrical communication with a power distribution line within a

power distribution system. The system comprises a power line coupler. A

substation receiver is in electrical communication with the power line coupler. A

substation circuit is in electrical communication with the substation transceiver. The

substation circuit is configured to simultaneously demodulate signals received from

the plurality of different endpoints.

Another aspect of the invention relates to a method of processing signals

received from a plurality of endpoints over power distribution lines. The method

comprises obtaining a plurality of signals from a power distribution line, each signal

corresponding to a different frequency bandwidth; and simultaneously demodulating

the plurality signals.

Description of the Drawings

Figure 1 is a block diagram illustrating one link of a power distribution

network over which data is communicated between a distribution substation and an

endpoint according to one possible embodiment of the present invention. Figure 2 is a block diagram illustrating a portion of a power distribution

network over which data is communicated between a distribution substation and an

endpoint according to one possible embodiment of the present invention.

Figure 3 is a schematic illustrating a substation controller and power line

coupler according to one possible embodiment of the present invention.

Figure 4 is a block diagram illustrating a hierarchy of data channels as they

are demodulated according to one possible embodiment of the present invention.

Figure 5 is a block diagram illustrating a circuit for processing data received

at a substation transceiver according to one possible embodiment of the present

invention.

Figures 6A-6D is a block diagram illustrating a digital signal processing

circuit for demultiplexing and demodulating a plurality of data channels according to

one possible embodiment of the present invention.

Figure 7 is a block diagram illustrating components of a substation

transceiver for transmitting a signal onto a power distribution network according to

one possible embodiment of the invention.

Figure 8 is a block diagram illustrating a digital signal processing circuit for

demodulating a transmission signal fed back into a substation transceiver according

to one possible embodiment of the present invention. Figure 9 is a flowchart illustrating operation of at least a portion of the

commands programmed into a computer illustrated in Figure 5.

Detailed Description

Various embodiments of the present invention will be described in detail

with reference to the drawings, wherein like reference numerals represent like parts and assemblies throughout the several views. Reference to various embodiments does not limit the scope of the invention, which is limited only by the scope of the claims attached hereto. Additionally, any examples set forth in this specification are not intended to be limiting and merely set forth some of the many possible embodiments for the claimed invention. Figure 1 is a block diagram of one link of an electric distribution system 100 distributing power between a distribution substation and a customer device at the power consumer's site. An electrical distribution system, or distribution plant as it is sometimes referred to, is that part of an electric power system that receives power from a power generator via high- voltage transmission lines, reduces or steps down the voltage, and then distributes the power to an endpoint at the premise of an energy customer. Within the electrical distribution system, distribution lines typically conduct electricity from the distribution substation to the endpoints. Distribution lines typically consist of underground cable, aerial cable, or overhead open- wire conductors carried on poles, or some combination of them. There maybe multiple layers of distribution substations connected in series between the power generation and the endpoint, wherein each consecutive distribution substation further steps down the voltage of the electricity being transmitted. Additionally, the power generators, distribution substations, and endpoints are commonly organized in a network where various generators supplying power can be taken on or offline and the distribution substation through which a particular endpoint receives its electricity can be changed, all without a loss or interruption of power. Distribution transformers are ordinarily connected in the distribution line between the distribution substation and the endpoint, which the distribution transformers serve to further step-down the voltage to a level that is used by consumers. These step-down transformers, often referred to as pole transformers, supply a consumer or group of consumers over a secondary circuit. Each consumer is connected to the secondary circuit through its service leads and meter. The distribution substation 102 shown in Figure 1 provides power to a customer device or endpoint 104 via a distribution line 106. The distribution line

106 may be coupled to one or more step-down transformers before reaching the customer premise. The distribution line 106 provides the power necessary to operate electrical devices, located at the endpoint 104, which is a device at the customer premise. Endpoints are discussed in United States patent application serial no. , which is entitled Endpoint Processing and Communication System

(which is identified by attorney docket no. 11838.54-US-01 and Express mailing label no. EN 155481401 US). For a variety of reasons, it may be desirable to communicate information

from the distribution substation 102 to one or more endpoints 104 at a particular customer premise. For example, it may be desirable to control or monitor a meter- reading device, which is installed at a customer premise to determine the power consumption at that customer premise. Additionally, control information could provide the ability to control or alter the operation of the meter-reading device and/or individual loads at the customer premise. Utility companies often provide a customer with a power rate discount if the customer agrees to allow for a temporary adjustment of their consumption. For example, a power company may provide a customer with a rate discount where the customer agrees to allow the power company to temporarily adjust or terminate their power consumption for certain nonessential power consuming devices, such as water heaters, swimming pool heaters, air conditioners, etc. during peak operation. This allows the utility company to limit the peak power consumption when necessary, hereinafter referred to as "load control."

Other more general information, which is not necessarily to "control"

customer devices, also can be provided via the power distribution lines. These

general information signals are transmitted in the same manner as signals intended to

control a customer device. Such general information signals include information to

display or store the price of power at the customer premise, the date and time, the

temperature or other information capable of being received and translated at the

customer premise. For example, the time displayed on an electronic device at the

customer premise could be periodically adjusted to display an accurate time as

transmitted by the utility station. Various embodiments of the apparatuses and methods disclosed herein

communicate control signals and general information signals to endpoints 104 via

the distribution line 106 to control customer devices and provide more general

information to the customer. Information from the customer device also may be sent

via the distribution line 106 to the distribution substation 102, thereby creating a

two-way control information communication link via the distribution line 106. The

aforementioned examples of control signal applications where control signals

(and/or general information signals) are provided by the distribution substation to an

endpoint 104 are merely representative of the various uses that such control signals provide. Therefore, the examples provided throughout the application are

illustrative in nature, as the invention is not limited to any particular control signal

use. In order to provide control information at the distribution substation 102, a

substation controller 108, which includes a substation transceiver, is used to drive the control signals along the distribution line 106 in the direction represented by the arrow 110. An endpoint transceiver 112 at the customer device 104 is configured to

recognize the control signals transmitted by the substation controller 108. Similarly, the substation controller 108 receives information, such as a power consumption

reading, from the endpoint transceiver 112 in the direction represented by arrow 118. The control information communications link 100 shown in Figure 1 therefore provides a full-duplex or bi-directional communication link between the

distribution substation 102 and the endpoint 104, which is typically located at a customer premise. Full duplex in this sense refers to simultaneous communications in both directions, although the information sent in one direction may travel at a speed different from that of the information provided in the opposite direction. This full-duplex communication link via the distribution line 106 provides for reliable transmission of control information, without the need for additional wiring, thereby minimizing cost and increasing data integrity. Referring now to Figure 2, a block diagram of a power distribution system

200, which is substantially similar to the electrical distribution system 100 described above. In this exemplary embodiment, generating station 202 provides the bulk

power to downstream distribution substations 102 via high-power transmission lines

203. At least one of the distribution substations 102 includes the substation controller 108. As can be seen by the example of Figure 2, the substation controller 108 can simultaneously communicate data via the distribution lines 106 to multiple

endpoints 104 residing in multiple customer premises. The control information can pass through transformers 210, and ultimately to a particular endpoint 104 located at a customer premise. A plurality of endpoints 104 located at different customer premises may be serviced by a particular transformer 210. Furthermore, a single

customer premise such as site 212 may include a plurality of different customer

devices or endpoints 104. The transfer of control information from a substation

controller 108 to a great number of endpoint transceivers at different endpoints 104 is very useful and cost effective, hi various embodiments, one or more of the

distribution substations 102 may include the substation controller 108 for

communicating with endpoints 104 located downstream from the distribution substation 102. In other embodiments, the substation controller 108 is located at points that are upstream from multiple endpoints 104 other than a distribution

substation 102. Figure 3 illustrates the distribution substation 102 and one possible

embodiment of the connection of the substation controller 108 to the distribution

line 106. In this exemplary embodimenb the distribution line 106 interfaces with a

main transformer 300 that provides three-phase power (φA, ^B, and φC) and

includes three distribution line conductors 302, 304, and 306, one for conducting

each phase of the power. The first conductor 302 conducts φA, the second conductor

304 conducts φB, and the third conductor 306 conducts φC. A metering loop 307 has three metering lines 314, 316, and 318 that interface

with the three distribution line conductors 302, 304, and 306, respectively. The metering line 314 interfaces with the distribution line conductor 302 through a

current transformer 308, the metering line 316 interfaces with the distribution line conductor 304 through a current transformer 310, the metering line 318 interfaces with the distribution line conductor 306 through a current transformer 312. The substation controller 108 has three inputs 327, 329, and 331 for receiving data from the endpoint transceivers 112 and one output 337 for sending

data to the endpoint transceivers, and includes a substation processing unit 332 and an amplifier 336. The three inputs 327, 329, and 331 interface with the metering

loop 307 through current transformers 320, 322, and 324, respectively. Specifically,

a first input has a line 326 that is coupled to the metering line 314 through a cunent transformer 320, a second input has a line 328 that is coupled to the metering line 316 through a current transformer 322, and a third input has a line 330 that is

coupled to the metering line 318 through a current transformer 324. In this exemplary embodiment, signals transmitted by an endpoint transceiver 112 connected to the first distribution line conductor 302, are communicated over the distribution line conductor 302, the metering line 314, the input line 330, and

into the first input 327 of the substation controller 108. Signals transmitted from

endpoint transceivers 112 connected to the second and third distribution line

conductors 304 and 306 are fed to the substation controller 108 through inputs 329 and 331, respectively, following similar paths along distribution line conductors 304 and 306, respectively, metering lines 316 and 318, respectively, and input lines 328

and 330, respectively. The substation processing unit 332 has a single output that feeds signals for downstream communication to the endpoint transceivers 112. This downstream

signal from this single output is input to the amplifier 336. The amplifier 336 then outputs the downstream signal from the substation controller 108 to a power line coupler 337, which is formed by an impedance matching unit 338 and capacitors

340, 342, and 344.

The identical signal is then communicated from the impedance matching 338

onto each of the distribution line conductors 302, 304, and 306 through the

capacitors 340, 342, and 344, respectively. The impedance matching unit 338

matches the impedance between the distribution line conductors 302, 304, and 306

and the amplifier 336. The capacitors 340, 342, and 344 electrically isolate the

impedance matching unit from the distribution line 106.

In this exemplary embodiment, the substation controller 108 transmits its

command to all of the downstream endpoint transceivers 112. In an alternative

embodiment, the substation controller 108 can address a downstream signal to a

particular endpoint transceiver 112.

In yet another possible embodiment, the substation controller 108 can shift

the phase of the signal transmitted onto each of the distribution line conductors 302,

304, or 306 so that each conductor conducts a signal having a different phase. If the

signal bleeds from one distribution line conductor 302, 304, or 306 to another

distribution line conductor 302, 304, or 306, the signals tend to cancel each other

because the electricity conducted on each of the distribution line conductors 302,

304, or 306 is out of phase by about 120°. Shifting the phase of the signal

transmitted by the substation controller 108 reduces this cancellation when a signal

bleeds from one distribution line conductor 302, 304, or 306 to another distribution

line conductor 302, 304, or 306.

As explained in more detail herein, the substation processing unit

communicates with a central office 334 via a data network 346. hi various embodiments, the data network 346 is established using a suitable means for data

communication. Examples, include the Internet, an Intranet, a wide area network, a

local area network, satellite, microwave, and a modem interfacing with a plain old

telephone line (POTS) Additionally, other structures may have alternative structures and methods

for retrieving a data signal from the distribution line 106, for transmitting a data

signal onto the distribution line 106, and processing the signals in the substation

controller 108. For example, the substation controller 108 might be broken into a separate receiver and transmitter. Referring to Figure 4 and as explained in more detail herein, the endpoint

transceivers 112 disclosed in the exemplary embodiment modulate the signals that

they transmit to the substation controller 108 using f equency shift keying and

transmits its signal to the substation controller 108 located at the distribution

substation 102. In one possible embodiment of this modulation scheme, each

endpoint transceiver 112 sends its signal within a channel 402 having a

predetermined bandwidth of about 36 Hz, from about 970 Hz to about 1,006 Hz.

When demodulating the signals received over the 36 Hz channel 402, the substation

controller 108 separates the signals into about 72 sub-channels 404. It then separates

each of the sub-channels 404 into about 125 sub-sub-channels 406. Each of the sub-

sub channels 406 is assigned to a different downstream endpoint transceiver 112 and

corresponds to a signal having a bandwidth of about 4 Hz.

In the exemplary embodiment, each endpoint transceiver 112 is assigned

predetermined bandwidth of about 4 Hz, and each assigned predetermined

bandwidth within the main channel 402 is mutually exclusive from one another.

Given this configuration, the substation controller 108 has the capacity to receive signals from about 9,000 separate downstream endpoint transceivers 112.

Additionally, the predetermined bandwidth for each sub-sub-channel 406 includes a

base f equency to which an endpoint transceiver 112 is assigned and any frequency

to which the base frequency is shifted when the endpoint transceiver 112 modulates

a signal for transmission to the substation controller 108.

The distribution substation controller 108 demodulates each signal received

from separate endpoint transceivers 112 substantially simultaneously, which

provides significant advantages. For example, it increases the capacity of the system

because the substation controller 108 does not have to delay reception of one signal

from an endpoint transceiver 112 until the reception of the previous signal is

completed.

As further explained herein, each distribution line 106 has three phases and

one distribution line conductor 302, 304, and 306 for each phase. The interface for

each of the three distribution line conductors 302, 304, and 306 receives one channel

over each of the three distribution line conductors 302, 304, and 306. In one

possible embodiment, each channel has a bandwidth of about 36 Hz, from about 970

Hz to about 1,006 Hz. Accordingly, the substation controller 108 perfonns the

demodulation scheme illustrated in Figure 4 for each distribution line conductor 302,

304, and 306 of the distribution line 106. This configuration gives the substation

transceiver 106 the capacity to receive signals from up to about 9,000 endpoint

transceivers 112 on any one or combination of the three distribution line conductors

302, 304, and 306.

There are many other additional embodiments in addition to those described

herein. For example, an endpoint transceiver 112 can communicate with the substation controller 108 using any modulation scheme, including modulation

schemes other than frequency shift keying, that permit simultaneous or substantially

simultaneous demodulation of signals received from the endpoint transceivers 112.

Additionally, other bandwidths can be used within the apparatus and methods

disclosed herein. For example, the channel 402 might include a frequency

bandwidth other than 36 Hz and other than the range of 970 Hz to 1,006 Hz. Yet

other embodiment will have the capacity to receive signals from fewer than 9,000

endpoint transceivers 112 or more than 9,000 endpoint transceivers 112.

In alternative embodiments, the endpoint transceivers 112 communicate

using frequency bandwidths other than 4 mHz. For example, various embodiments

might use a frequency bandwidth of about 10 mHz or less, including frequencies of

about 2 mHz, 6 mHz, or 8 mHz. Yet other embodiments use frequency bandwidths

other than 2 mHz, 4 mHz, 6 mHz, 8 mHz, or 10 mHz. Still other embodiments

might use frequency bandwidths greater than 10 mHz. Referring to Figure 5, the exemplary embodiment of the substation

processing unit 332 includes four low pass filters 501, 503, 505, and 507, four

variable gain devices 508, 510, 512, 514, and 516, four high pass filters 509, 511,

513, and 515, a complex programmable logic device (CPLD) 530, four analog to

digital (AID) converters 518, 520, 522, and 524, first and second digital signal

processors (DSP) 536 and 538, two windowed watch-dog circuits 544 and 546, a

single board computer 540, flash memory 532, and a modem 542.

The CPLD 530 is programmed to include a clock 526, a variable gain

controller 528, and four receive buffers 510, 512, 514, and 516 that are used to

buffer the signals received from the distribution line that are being input to the first

and second DSPs 536 and 538 for demodulation. As explained in more detail herein, the CPLD 530 may have additional buffers for various functions such as

buffering data being transmitted by the substation controller 108. One type of CPLD

that can be used is chip no. XC95144XL, which is manufactured by XILINX located in California, U.S.A. The clock 526 provides a clocking signal to each of the AID converters 518,

520, 522, and 524. The variable gain controller 528 provides a signal to each of the

variable gain devices 502, 504, 506, and 508. In one possible embodiment, the

variable gain devices 502, 504, 506, and 508 are op amps that output a signal having

an amplitude of about 2.5 Volts. The flash memory 532 stores the code for the first and second DSPs 536 and

538. The flash memory 532 also stores an error log that records error signals

generated by the first and second DSPs 536 and 538.

As explained in more detail herein, the first and second DSPs 514 and 516

demodulate the signals received over the distribution line 106 using frequency shift

keying. The second DSP 538 then outputs the demodulated data into the signal

board computer 540, which stores the data in the memory (not shown). One type of

chip that can be used for the DSPs is chip no. TMS3206711, which is manufactured

by Texas Instruments located in Texas, U.S.A. h operation, an endpoint transceiver 112 generates a signal that embodies

data and transmits that signal over the distribution line 106. For exemplary

purposes, the endpoint transceiver 112 is in electrical communication with and

transmits its signal over the distribution line conductor 302. The signal then

propagates through the current transformer 308, along metering loop line 314,

through current transformer 320, along input line 326, and into the first input 327. The signal is conditioned bypassing through the high pass filter 501, the first variable gain device 502, and the low pass filter 509. The variable gain device 502 is an amplifier that biases the signal a predetermined amount such as about 2.5 Volts. The high pass and low pass filters 501 and 509 isolate the signal and remove noise. The signal is converted from analog to digital by the first A/D converter 518

and input to the first buffer 510 programmed into the CPLD 530. The buffered

signal is input to the first DSP 536. The first DSP 536 performs a first portion of the operations to demodulate the signal and then passes the signal to the second DSP

538, which completes the demodulation process. The demodulated signal is input to the single board computer 540.

Signals transmitted on the transmission conductor 304 similarly propagate through the current transformer 310, along metering line 316, through cunent transformer 322, along input line 328 and input second input 329. The signals then propagate through the high pass filter 503, through the variable gain device 504,

through the low pass filter 511, through the second A/D converter 520 and input the

second buffer 512 programmed into the CPLD 530. The signal then is processed by

the first and second DSPs 536 and 538, and input to the single board computer 540. Signals transmitted on the distribution line conductor 306 also similarly propagate through the cunent transformer 312, along metering line 318, through

cunent transformer 324, along input line 330 and input second input 331. The

signals then propagate through the high pass filter 505, through a variable gain device 506, through the low pass filter 513, through an A/D converter 522 and input the second buffer 514 programmed into the CPLD 530. The signal then is processed by the first and second DSPs 536 and 538, and input to the single board computer

540.

The first and second DSPs 536 and 538 monitor each of the distribution line

conductors 302, 304, and 306 for each sub-sub-channel 406. Thus, if an endpoint

transceiver 112 is transmitting a signal on one of the sub-sub-conductors 302, 304,

or 306 and the signal bleeds to the other two conductors, the substation controller

108 will receive three separate signals within the frequency band of the sub-sub¬

channel 406. The single board computer 540 then determines which of the signals

has the greatest amplitude and discards data from the other two signals. An

advantage of this embodiment is that a technician installing the endpoint transceiver

112 can quickly connect it to any of the distribution line conductors 302, 304, or

306.

Signals transmitted by the substation controller 108 are fed back into the

substation controller 108 and recorded. This feedback provides a historical record of

the data actually transmitted onto the distribution line 106. Because the transmitted

signal propagates wherever power is distributed, it is also transmitted to a 120 volt

outlet in which a power supply (not shown) for the single board computer 540 is

plugged. A signal is picked up from the power supply, and is then processed similar

to the signals that are transmitted on the distribution line 106 by the endpoint

transceivers 112. More specifically, the signal is fed through the high pass filter

507, through the variable gain device 508, through the low pass filter 515, and then

through the fourth A/D converter 524. The signal is then buffered 516 by the fourth

buffer programmed in the CPLD 530 and input to the first and second DSPs 536 and

538 where it is demodulated. The first DSP 536 sends a watch-dog signal to the windowed watch-dog

timer 544 at a predetermined interval. If the windowed watch-dog timer 544 does

not receive a watch-dog signal within the predefined window, it generates and sends

a reset signal to the first DSP 536. In one possible embodiment, the window within

which the windowed watch-dog circuit 544 looks to receive a watch-dog signal is

from about 0.7 seconds to about 1.3 seconds. The second DSP 538 interfaces with

the windowed watch-dog timer 545 in a similar manner.

The single board computer 540 is a standard computer board 540 that

includes a programmable processor, memory, and various inputs and outputs. One

type of single board computer that can be used in Model # EBCTXPLUS-5222B,

which is manufactured by WinSystems located in Texas, U.S.A. The single board

computer 540 communicates with the second DSP 538 through a data bus 534. In

the exemplary embodiment, the data bus 534 is a universal asynchronous receive and

transmit (UART) data bus that communicates according to the RS-485 data protocol.

The second DSP 538 includes a command processor or decoder that decodes

commands received from the single board computer 540 and then either processes

the command itself or relays the command to the first DSP 536, clock 526, variable

gain device controller 528, or other hardware or firmware element for execution. Additionally, the single board computer 540 communicates to the outside

world through a network interface 542. Examples of possible network interfaces

include modems for communication over plain old telephone lines (POTS). Other

examples, include hardware and drivers to support communication using the

Internet, an Intranet, a wide area network (WAN), a local area network (LAN),

satellite, microwave, or any other type of network connection that can be used for communicating data to remote locations, whether the communication is over

hardwired lines or is wireless.

Although a certain hardware and software configuration is illustrated in this

exemplary embodiment for receiving and processing data signals there are many

possible alternative embodiments and the invention can be embodied in any

configuration of hardware and/or software that can receive, demodulate, and demultiplex the signals.

Referring to Figures 4 and 6A, the signal output from the first D/A converter

518 and buffered in the first buffer 510 conesponds to the¹ main channel 402, which

in the exemplary embodiment has a bandwidth of about 36 Hz (970 Hz to 1006 Hz).

The signal is input to the first DSP 536, where it passes through a low pass filter

600, through a 100:1 decimator 602 and then a high pass filter 604.

In one possible embodiment, the low pass filter 600 has a cutoff frequency of

about 1090 Hz and prevents aliasing or the enoneous interpretation of a high

frequency component as a lower frequency component as the signal is sampled. The

100:1 decimator 602 changes the sampling rate from about 250,000 samples per

second (sps) to about 2,500 sps. The 100:1 decimator 602 reduces the sampling rate

by accumulating 200 words. The high pass filter 604 has a cutoff frequency of about

906 Hz and filters all frequencies below 906 Hz, including the 60 Hz component

from the alternating cunent conducted by the transmission line 100.

The signal is then passed to first and second signal branches 606a and 606b.

Along the first signal branch 606a, the signal passes through a mixer 608a that

combines the signal with a cosine wave 610 having a predetermined frequency. In

exemplary embodiment disclosed herein the cosine wave has a frequency of about

960 Hz, which is line locked to the power distribution frequency of the 120/240 V input for the SPU 332. An advantage of line locking the frequency of the cosine wave is that it will vary with the frequency (about 60 Hz) of the alternating cunent

carried by the distribution line 106, which is the carrier wave for the signals transmitted by the endpoint transceivers 112. The signal then passes through a low pass filter 614a and then a 12:1 decimator 616a. In one possible embodimenb the

low pass filter 614a has a cutoff frequency of about 50 Hz and the 12:1 decimator 618a changes the sampling rate from about 2500 sps to about 208.33 sps. The signal passed to the second branch 606b is processed in a similar manner

passing through a mixer 608b, a low pass filter 614b, and a 12:1 decimator 618b. In the exemplary embodiment, the only difference for the processing in the first branch

606a is that the mixer combines the signal with a sine wave having a frequency of about 960 Hz. In the exemplary embodiment, the code for the 12:1 decimators 618a and 618b are split between the first and second DSPs 536 and 538. Accordingly, the

signals propagating along the first and second signal branches 606a and 606b are

passed from the first DSP 536 to the second DSP 538 as they are being processed by the 12:1 decimators 616a and 616b. In another possible embodiment, all of the demodulation steps illustrated in Figure 6A-6C are performed in a single digital signal processor so long as it has enough processing power and memory. Referring to Figures 4 and 6B, the signal output by the 12: 1 decimators 616a

and 616b are input into 72 parallel sub-channel signal branches, each parallel subchannel signal branch conesponding a separate sub-channel 404. As it passes through each of the parallel sub-channel signal branches, the signal transmitted along the first signal branch 606a passes through a bandpass filter 618a, through a mixer 622a, and into a subtractor 628. In the exemplary embodiment, the bandpass filter

has a lower cutoff frequency of about 10 Hz and an upper cutoff frequency of about

10.5. The mixer 622a mixes the signal with a cosine wave 624 having a

predetermined frequency of about 9.8 Hz, which is line locked to the power

distribution frequency of the 120/240V input for the SPU 332.

Similarly, the signal transmitted along the second signal branch 606b passes

through a bandpass filter 618b, through a mixer 622b, and into the subtractor 628.

The primary difference from the first signal branch 606a is that the mixer 622b

combines the signal with a sine wave having a frequency of about 9.8 Hz, which is

line locked to the power distribution frequency of the 120/240 V input for the SPU

332. The subtractor 628 then subtracts the signal processed along the second signal

branch 606b from the signal processed along the first signal branch 606a. The

signal output from the subtractor 628 is passed through a low pass filter 630 and a

71:1 decimator 632. The low pass filter 630 has a cutoff frequency of about 0.7 Hz

and further isolates the data signal. The 71:1 decimator further reduces the sampling

rate from about 208.33 sps to about 2.93 sps.

In the exemplary embodiment, the bandpass filter 618b has a lower cutoff

frequency of about 10 Hz and an upper cutoff frequency of about 10.5. The mixer

622b mixes the signal with a sine wave 626 having a predetermined frequency of

about 9.8 Hz, which is line locked to the power distribution frequency of the

120/240V input for the SPU 332.

Each of the parallel sub-channel signal braches is substantially the same as

the illustrated block diagram for sub-channel 1. The primary difference is that the

frequency input to the mixer is incremented by about 0.5 Hz for each successive sub-

channel 404. The frequency input to the mixer is line locked to the power distribution frequency of the 120/240V input for the SPU 332. Similarly, the maximum and minimum cutoff frequencies for the bandpass filters 618 are increased by about 0.5 Hz for each successive sub-channel 404. Thus, for example, the second sub-channel has a mixer frequency of about 10.3 Hz and the bandpass filter has cutoff frequencies of about 10.5 Hz and about 11.0 Hz. In this exemplary embodiment, the seventy-second sub-channel has a mixer frequency of about 45.3

Hz and the bandpass filter has cutoff frequencies of about 45.5 Hz and about 50.0

Hz.

Refening to Figures 4 and 6C, the signal output by the 71 :1 decimator 632 for each parallel sub-channel signal branch is input into 125 parallel sub-sub-channel signal branches. Each parallel sub-sub-channel signal branch conesponds to a separate sub-sub-channel 406. Given the architecture in this exemplary embodiment, there are about 9000 parallel sub-sub channel signal branches.

The signal input to the first sub-sub-channel signal branch passes through a high pass filter 634, a mixer 636, and a low pass filter 642. In the exemplary embodiment, the high pass filter 634 has a cut off frequency of about 0.198 Hz, and the mixer 652 combines the data signal with a cosine wave 640 having a frequency of about 0.184 Hz, which is line locked to the power distribution frequency of the

120/240V input for the SPU 332. After passing through the low pass filter 642, the data signal is output to a first signal branch 644 and a second signal branch 646. The signal passed to the first signal branch 644 is passed through a 44:1 decimator 648, which decreases the

^■ sampling rate from about 2.93 sps to about 0.0667 sps.

Referring to Figures 6C and 6D, the decimated signal passed into a heuristically conelated detector 652, which estimates the phase of the signal received from the endpoint transceiver 112. The signal propagates onto two signal

paths 667a and 667b. On the first signal branch, the signal is input to five multipliers 668a-668e. The first multiplier 668a mixes the signal with a cosine wave having a phase calculated according cos(0.01583333(t-0)). The second multiplier 228b mixes the signal with cosine wave having a phase calculated according to cos(0.0158333(t-T)) where t is time and T is the sample rate of the signal output by the 44:1 decimator 648. The third multiplier 668c mixes the signal with cosine wave having a phase calculated according to cos(0.0158333(t-2T)) where t is time and T is the sample rate. The fourth multiplier 668d mixes the signal with cosine wave having a phase calculated according to cos(0.0158333(t-3T)) where t is time and T is the sample rate. The fifth multiplier 668e mixes the signal with cosine wave having a phase calculated according to cos(0.0158333(t-4T)) where t is time and T is the sample rate. In the exemplary embodiment, the samples output by the multipliers 668a-668e are accumulated by the accumulators 672a- 672e, respectively, for about a 20-minute period, which results in the collection of

about 80 samples. The sample from each of the accumulators 672a-672e is input to a maximum value function 676. The maximum value function 676 then determines

which of the five samples received from the accumulators 672a-672e has the greatest value and inputs that greatest value to the subtractor 680. The maximum value function 676 discards the remaining samples.

Similarly, the signal propagating along the second signal branch 667b is

input to multipliers 670a-670e. The first multiplier 670a mixes the signal with a cosine wave calculated according to cos(0.0175000(t-0)). The second multiplier 670b mixes the signal with a cosine wave calculated according to cos(0.0175000(t- T)), where T is the sampling rate of the signal output by the 44:1 decimator 648. The third multiplier 670c mixes the signal with a cosine wave calculated according to cos(0.0175000(t-2T)), where T is the sampling rate of the signal output by the 44: 1 decimator 648. The fourth multiplier 670b mixes the signal with a cosine wave calculated according to cos(0.0175000(t-3T)), where T is the sampling rate of the

signal output by the 44: 1 decimator 648. The fifth multiplier 670b mixes the signal with a cosine wave calculated according to cos(0.0175000(t-4T)), where T is the sampling rate of the signal output by the 44: 1 decimator 648. In the exemplary embodiment, the samples output by the multipliers 670a-670e are accumulated by

the accumulators 674a-674e, respectively, for about a 20-minute period, which results in the collection of about 80 samples. The sample from each of the accumulators 674a-674e is input to a maximum value function 678. The maximum value function 678 then determines which of the five samples received from the accumulators 674a-674e has the greatest value and inputs that greatest value to the

subtractor 680. The maximum value function 678 discards the remaining samples. The subtractor then determines the difference between the samples received from the maximum value functions 676 and 678 and outputs the difference to the sheer 658. Alternative embodiments of the heuristically correlated detector 65 are

possible. For example, it can estimate the phases of the signal received by the endpoint transceiver more accurately, by mixing the signal propagating along each of the signal branches with more than five cosine functions having the phase of the function increases in smaller increments. For example, it could increment the phase by about 0.5T rather than about IT. The heuristically conelated detector also could estimate the phases of the signal less accurately by mixing signals with fewer cosine functions having the phase function for each cosine function increased in greater

increments.

Referring back to Figure 6C, the sheer 658 limits each sample received from

the heuristically correlated detector between two predetermined amplitude

boundaries, which increases the signal-to-noise ratio. In one possible embodiment,

the upper boundary is a voltage level of about +1 volts and the lower boundary is

voltage level of about -1 volts, hi the exemplary embodiment, the sheer 658 reads

the sliced samples at about 20-minute intervals, although other embodiments will

read the sliced samples at other intervals. A bit latch 660 is clocked and synchronized with the sheer 658 so that it is

set by the last bit at the same interval at which the sheer 658 reads the sliced sample.

Again, in the exemplary embodiment, the interval is about 20 minutes although other

embodiments might have other intervals. Again other embodiments might if the last

bit has a positive value, the bit latch 660 sets a logical "1". If the last bit has a

negative value, the bit latch 660 sets a logical "0". The bit output by the latch 660 is

then input to the single board computer 540, which stores the data in memory.

The signal passed to the second signal branch 646 passes through a bandpass

filter 661, which in the exemplary embodiment has a pass band from about 15.08333

mHz to about 18.25 mHz. The signal is then passed through an absolute value

function 662 that rectifies the signal and through a low pass filter 664, and through a

scaling function 666. In the exemplary embodiment, the low pass filter 664 that has

a cut off frequency that is adjustable in the range from about 250 μHz to about 1.0

mHz. The cut off frequency is adjusted until the low pass filter 664 smoothes the

signal being processed and the second DSP 538 outputs a signal strength word that accurately identifies the strength of the signal being received from the endpoint transceiver 112. The scaling function 666 scales the amplitude of the signal within the bandwidth defined by the low pass filter 664. hi the exemplary embodiment, it scales the amplitude of the signal by a factor of about 9.587xl0^"6 (Gain/100), where Gain is the gain of the variable gain device 502, 504, or 506 that provides an input for the signal being processed by the scaling function 666. The scaled signal is a binary signal strength word that conesponds to the strength or amplitude of the signal received from the endpoint transceiver 112. The single board computer 540

can selectively obtain the signal strength word. The signal output by the low pass filter 664 is also input to a fault detection signal branch 681 that deteraiines whether the substation controller 108 fails to

receive a signal from the endpoint transceiver 112 on that sub-sub channel 406 or whether the signal that it is receiving is too noisy. The signal passes through a standard deviation function 682. The signal has a D.C. bias, and the standard

deviation function 682 calculates the standard deviation of the signal's amplitude, which varies around the D.C. bias. The signal output by the standard deviation function is input to a 44:1

decimator, which reduces the sampling rate from about 2.93 sps to about 0.0667 sps. The signal then passes through a low pass filter 686 that has a cut off frequency that

is adjustable between about 1 μHz to about 100 μHz. The cut off frequency is

adjusted until the low pass filter 686 smoothes the signal being processed and the

second DSP 538 outputs a signal strength bit and a noise word that accurately indicates lost and noisy sub-sub-channels 406, respectively, with a desired degree of accuracy.

The signal is then input to a high-limit scaling-function 688 and a low-limit scaling function 690. The high- limit scaling function 688 scales the amplitude of the signal by the factor 240*(Gam/100), where Gain is the gain of the variable gain

device 502, 504, or 506 that provides an input for the signal being processed by the

scaling function 688. The constant (240 in the exemplary embodiment) used in the

scaling function 688 is set so that the output of the scaling function 688 is about 6 sigma. The output of the scaling function 688 is an upper limit. Similarly, the low- limit scaling function 690 scales the amplitude of the signal by the factor

120*(Gain/100), where Gain is the gain of the variable gain device 502, 504, or 506 that provides an input for the signal being processed by the scaling function 690.

The constant (120 in the exemplary embodiment) used in the scaling function 690 is set so that the output of the scaling function 690 is about 3 sigma. The output of the scaling function 690 is a lower limit.

The upper and lower limits are input to LOS detection logic 692, which

compares the unsealed signal strength word output by the low pass filter 664 to the

upper and lower limits output by the scaling functions 688 and 690. The LOS detection logic 692 outputs a signal strength bit indicating a lost channel if the

strength of the signal received on the sub-sub-channel 406, as output by the low pass filter 664, first goes above the upper limit and then below the lower limit or if the

strength of the signal never goes above the upper limit. The LOS detection logic 692 outputs a signal strength bit indicating a signal is being received on the sub-subchannel 406 if the strength of the signal as output by the low pass filter 664 goes above the upper limit and stays above the upper limit. The LOS detection logic 692 also outputs a signal strength bit indicating a signal is being received on the sub-subchannel 406 if the strength of the signal as output by the low pass filter 664 goes above the upper limit and then falls below the upper limit but not below the lower

limit. The single board computer 540 selectively obtains the bit output by the LOS detection logic 692.

The output of the low pass filter 686 is also input to a bit latch 694, which latches the binary word output by the low pass filter 686. The value of this word is the standard deviation and conesponds to the amount of noise present in the signal

received on the sub-sub-channel 406. The single board computer 540 selectively obtains the noise word latched by the bit latch 694. The single board computer 540 stores in memory the signal strength word received from the scaling function 666, the signal strength bit output by the LOS detection logic 692, and the noise word output by the bit latch 694. If the single

board computer 540 does not receive the signal strength bit from the LOS detection

logic 692 and it determines that there should be an endpoint transceiver 112 sending a signal on that particular sub-sub-channel 406, it can generate an error. In one possible embodiment, the single board computer 540 can then send a command to

find the missing endpoint transceiver 112. In another possible embodiment, the

single board computer 540 can send the signal strength bit or other data item to the central office 334 either indicating a possible failure in the distribution plant or identifying the endpoint transceiver 112 as failed so that it can be repaired or replaced. Additionally, one can use the signal strength word output by the scaling

function 666 for diagnostic purposes. For example, if there is a signal strength word,

but the scaled value is not within the predetennined range, one might adjust the

biasing of the variable gain device 502, 504, or 506. In yet another possible

embodimenb the single board computer 540 automatically instructs the variable gain

device controller 528 in the CPLD 530 to adjust the bias of the variable gain device

502.

Other possible embodiments might include different combinations of bits or

binary words output by the LOS detection logic 692. hi one possible embodiment,

the LOS detection logic 692 outputs only a single signal strength bit. In other

embodiments, the LOS detection logic 692 outputs binary words or other

combinations of bits and/or flags. Yet another possible embodiment does not

include the fault detection signal branch 681. This embodiment outputs only the

signal strength word from the scaling function 666 and the single board computer

540 uses this information to determine the presence or absence of a signal from the

endpoint transceiver 112. Other possible embodiments do not include the separate

signal branch 646 at all.

Again, although an exemplary embodiment of digital signal processing logic

is illustrated and described herein, many other embodiments for demodulating

received signals are possible. For example, the demodulating and demultiplexing

can be performed using any combination of hardware and/or software and any type

of multiplexing/demultiplexing and modulation/demodulation schemes that can

simultaneously process signals received from multiple endpoint transceivers 112.

Each of the parallel sub-sub-channel signal braches is substantially the same

as the illustrated block diagram for sub-sub-channel 1. The primary difference is that the frequency input to the mixer is incremented by about 0.004 Hz for each

successive sub-sub-channel 406. The frequency input to the mixer is line locked to

the power distribution frequency of the 120/240V input for the SPU 332. Similarly,

the maximum and minimum cutoff frequencies for the high pass filters 634 are

increased by about 0.004 Hz for each successive sub-sub-channel 406. Thus, for

example, the second sub-sub-channel has a mixer frequency of about 0.188 Hz and

the high pass filter has a cutoff frequency of about .202 Hz. hi this exemplary

embodiment, the one hundred twenty-fifth sub-sub-channel has a mixer frequency of

about 0.680 Hz and the high pass filter has cutoff frequency of about .694 Hz. Referring to Figure 7, when transmitting a signal to the endpoint transceivers

112, the single board computer 540 generates a data packet that contains a message.

The data packet is then buffered by an input buffer 700 and a transmit buffer 702,

both of which are 60-second buffers that hold the data for one minute.

The input and the transmit buffers 700 and 702 fonn a ping-pong buffering

scheme. The data passes first through the input buffer 700 and then through the

transmit buffer 702. When the data is received at the transmit buffer 702, it sends a

message to the input buffer 700 to delete the data from the input buffer register.

This buffering scheme prevents the loss of data.

The data is held in the transmit buffer 702 for about 60 seconds and then

input to the second DSP 538 where it is modulated using a single channel, binary

frequency shift keying scheme. Specifically, the data passes through a frequency

change function 704. The frequency change function 704 determines whether the

frequency shifts from one frequency to another. If the frequency shifts, the

frequency change function 704 determines that the data bit is a logical "1". If the

frequency does not shift, the frequency change function 704 determines that the data bit is a logical "0". In the exemplary embodiment, the frequency change function

determines whether the data bit is a logical "1" or "0" depending on whether the

frequency shifts between frequencies of about 555 Hz and about 585 Hz.

The data is output from the second DSP 538 and input to a digital to analog

(D/A) converter 710, which outputs a modulated transmit signal to a low pass filter

712. The low pass filter 712 has a cut-off frequency of about 1.2 KHz and smoothes

the modulated transmit signal. The modulated transmit signal then passes through a

variable gain device 714 that is electrically isolated by capacitors 716 and 718. The

variable gain device 714 is biased by the variable gain device controller 528 of the

CPLD 530. The variable gain device 714 is biased to set the modulated transmit

signal at a predetermined voltage level, hi one possible embodiment, the variable

gain device 714 sets the modulated transmit signal to a voltage level of about 1.4 Volts RMS.

The modulated transmit signal from the variable gain device 714 is then

amplified, h one exemplary embodiment, the modulated transmit signal is

amplified by a power audio amplifier 720 capable of amplifying a signal by about

400 Watts or more. One type of power audio amplifier that can be used is model

KI, which is manufactured by Crown Audio, Inc., of Indiana, U.S.A. After it is

amplified, the modulated transmit signal is passed through an impedance matching

unit 722, and then is simultaneously transmitted onto each of the distribution line

conductors 302, 304, and 306 through the isolation capacitors 724, 726, and 728,

respectively. The modulated transmit signal simultaneously propagates along each

of the transmission conductors 302, 304, and 306 to each of the endpoint

transceivers 112. To prevent insertion loss and distortion of the modulated transmit signal, the

impedance matching unit 722 matches the impedance between the output of the

audio power amplifier 720 and the distribution line conductors 302, 304, and 306.

One skilled in the art will appreciate that the impedance matching unit 722 includes

an inductance-capacitance network (not shown).

Although a certain hardware and software configuration for transmitting and

processing data signals is illustrated in this exemplary embodiment, there are many

possible alternative embodiments and the invention can be embodied in any

configuration on hardware and/or software that can transmit and modulate the

transmission signals.

As discussed above, and referring to Figures 5 and 8, the modulated transmit

signal propagates to the 120/240V input for the SPU 332. The voltage level is taken

from the power supply, conditioned, input to the fourth A/D converter 524, buffered

516 in the fourth buffer programmed into the CPLD 530, and then input to the first

DSP 536.

Within the first DSP 536, the transmit signal is passed through a 50:1

decimator, which converts the sample rate of the signal from about 500,000 samples per second (sps) to about 10,000 sps. The signal then propagates onto two signal

branches 801a and 801b. Within the first signal branch 801a, the signal first passes

through a bandpass filter 802a having a first predefined pass band, h one possible

embodiment, the bandpass filter 802a has a lower cutoff frequency of about 575 Hz

and an upper cutoff frequency of about 595 Hz. It isolates signals having a

frequency of about 585 Hz. hi other embodiments, the bandpass filter 802a has

different bandpasses. The signal then flows through an absolute value function 804a, which rectifies the signal, and then through a low pass filter 806a, which has a cutoff

frequency of about 5 Hz. The low pass filter 806a smoothes the rectified signal. The rectified signal passes through a 500: 1 decimator 808a, which changes the sample rate from about 10,000 sps to about 20 sps. The signal is then passed from the first DSP 536 to the second DSP 538.

The signal passed along the second signal branch 801b is processed in a

substantially similar method. It flows through a bandpass filter 802b, an absolute value function 804b, a low pass filter 806b, and then a 500:1 decimator 808a. The primary difference from the first signal branch 801a is that the bandpass filter 802b has a second predefined pass band. In one possible embodiment, the bandpass filter

802b has a lower cutoff frequency of about 545 Hz and an upper cutoff frequency of about 565 Hz. It isolates a signal having a frequency of about 555 Hz. hi other embodiments, the bandpass filter 802b has different bandpasses. The signal output from each signal branch 801a and 801b is input to a max

signal function 810, which selects the signal that has the largest amplitude and then

passes the larger of the two signals through a low pass filter 812. The low pass filter

812 further smoothes the signal so that a steady-state signal strength can be determined. The low pass filter 812 has a cutoff frequency of about 100 mHz. The

filtered signal then is passed through a scaling function 814 that biases the voltage level of the signal by a predetermined factor. In one possible embodiment, the predetermined factor, which equals l.lxlO^"4*(Gain/100), where Gain is the gain of

the variable gain device 508. The scaled signal is input to both a data latch 816 and a loss-of-signal (LOS) detection logic 818. In one possible embodiment, the data latch 816 operates in a substantially similar manner as the data latch 668. The LOS detection logic 818 compares signal strength to a predetermined value and outputs a signal strength bit indicating whether the signal strength is below the predetermined value.

Additionally, the data latch 816 and the LOS detection logic 818 interface with the single board computer 540 in a substantially similar manner as the data latch 668

and the LOS detection logic 692. The primary difference is that the single board

computer 540 uses the output of the data latch 816 and the LOS detection logic 818

to determine whether the substation controller 108 is properly transmitting a signal onto the distribution line 106 as opposed to detennining whether a signal is being properly received from the endpoint transceiver 112. The data latch 816 and LOS detection logic 818 can have similar alternative embodiments as the data latch 668

and the LOS detection logic 692, respectively. The signal output from each signal branch 801a and 801b also is input to a bit detection algorithm 820, which determines whether the data bit is a logical "1" or

"0". In one possible embodiment, the bit detection algorithm 820, which is programmed into the second DSP 538, receives the signal output from each signal

branch 801a and 801b. The bit detection algorithm then calculates which signal has a greater amplitude or value. If the signal that has the greater amplitude changes, the bit detection algorithm 820 determines that the frequency changed and conesponds to a logical "1" and outputs a high or "1" bit. If the signal that has the greater

amplitude does not change, the bit detection algorithm 820 determines that there is not change in the frequency, which conesponds to a logical "0" and outputs a low or "0" bit. The bits output by the bit detection algorithm 820 are input to a buffer 822.

In one possible embodiment, the buffer 822 is a 60-second buffer that accumulates

data in one-minute increments and then outputs the data to the flash memory 532

where it is stored. In one possible embodiment, the flash memory 532 stores the data

transmitted by the substation controller 108 for a predetermined period of time, such

as 4.5 days. As new data is stored in the flash memory 532, it overwrites oldest data.

The data is retrieved from the flash memory 532 upon receiving a fetch command

from the single board computer 540. In one possible embodiment, the single board

computer 540 generates a fetch command upon receiving instructions from a user. Although and exemplary embodiment of digital signal processing logic is

illustrated and described herein, many other embodiments for modulating a data

signal for transmission are possible. For example, modulation can be performed

using any combination of hardware and/or software and any type of modulation

scheme that is compatible with the endpoint transceivers 112.

Refening to Figure 9, the single board computer 540, which has a preemptive

operating system, communicates with a control server 934 via a data communication

link 936 such as an Ethernet network. In the exemplary embodiment, the controller

server 934 is centrally located in the central office 334 and communicates with

single board computers located at several different substations. The control server

934 generates commands and forwards them to the single board computer 540. The

single board computer 540 then assembles the commands into data packets, and the

substation confroller 108 relays the commands to the endpoint transceivers 112. The control server 934 also sends commands to the single board computer 540

instructing the computer 540 to upload data to the control server 934.

The code executed by the single board computer 540 is organized into tasks.

In the exemplary embodimenb there are three tasks that control communication with

the data communication link 936. These tasks are the Com Serv task 938, the Com

Client task 940, and the Net Time task 932.

The control server 934 communicates with the single board computer 540

using a metalanguage such as XML. The Com Serv task 938 is a command

processor that receives commands and data from the control servers 934. The

commands and data that it receives are in XML. The Com Serv task 938 then parses

the commands into a predetermined data structure that is understandable by the

operating code executed by the single board computer 540. hi the exemplary

embodiment, the Com Serv task 938 receives and processes all of the commands that

communicated from the control server 934 to the single board computer 540. Additionally, the Com Serv task 938 processes data for uploading to the

control server 934 by converting it to XML and then transmitting it to the control

server 934 over the data communication link 936. With the exception of certain data

processed and transmitted by the Com Client task 940, all of the data transmitted to

the Control server 934 is processed and transmitted by the Com Serv task 938. In the exemplary embodiment, the Com Client task 940 converts certain data

to XML and then transmits it to the control server 934. In the illustrated

embodiment, the Com Client task 940 processes and transmits messages and data

that are not specifically requested by the control server 934 or otherwise scheduled

for period transmission to the control server 934. An example includes a message that the status of a sub-sub-channel 406 was changed to Lost, indicating that

substation controller 108 failed to receive a signal from the endpoint transceiver 112

assigned or mapped to that sub-sub-channel 406. In one possible embodimenb the

Com Client task 940 reports the Lost status to the control server 934 if the status of the sub-sub-channel is not restored in a predetermined amount of time, hi another

embodimenb the Com Client task 940 reports the Lost status to the control server

934 as soon as the Lost status of the sub-sub-channel 406 is generated, which

permits quick identification of failures, including the type and location of the enor. It also permits the quick dispatch of technicians and/or engineers for repair. The Com Client task 940 also process and transmits messages and data

requested by the control server 934 in certain circumstances, hi the exemplary

embodiment, the Com Client server 940 processes and transmits data that takes more

than a significant amount of time to generate when requested from the control server

934. An example of such a command includes a request by the control server to

rebuild the table (discussed herein with respect to the CHAT.DAT data file 916)

allocating each of the endpoint transceivers 112 to a particular sub-sub channel 406.

In this exemplary embodiment, a significant amount of time lapses between the time

the single board computer 540 requests information or an action and the time the

single board computer 540 responds to the request or acknowledges the action. The code in the single board computer 540 defines on a command-by-

command basis, whether the Com Serv task 938 or the Com Client task 940

processes and transmits data to the control server 934. In this embodiment, a table

stores a list of commands executable by the single board computer 540 and identifies

either the Com Serv task 938 or the Com Client task 940 as the task responsible for sending the requested information or acknowledgement of the requested action to the

control server 934.

The Net Time task 932 retrieves the current time from a global time server

that provides an accurate time source. The global time server retrieves the cunent

time based on Universal Time Coordinated (UTC) from a source such as the Global

Positioning System (GPS) or a DCF77 signal, which is a radio clock signal generated by an atomic clock, h the exemplary embodiment, the global time server

is located remotely 102 and is connected to the single board computer 540 via the

data communication link 936. Additionally, the single board computer 540 is

programmed to adjust the UTC for the time zone served by the distribution

substation 102 and for daylight savings time, if time is adjusted for daylight savings

time in the geographic region served by the distribution substation 102. hi an

alternative embodiment, the global time server is programmed to adjust the UTC for

the time zone served by the distribution substation 102 and for daylight savings time,

if time is adjusted for daylight savings time in the geographic region served by the

distribution substation 102.

In the exemplary embodiment, the global time server retrieves the UTC in a

predetermined interval such as once per day. In turn, the Net Time task 932 retrieves

the time, which is adjusted as appropriate, in a predetermined interval such as once

per day. The Net Time task 932 has a clock function that tracks the time and is

recalibrated to the time retrieved from the global time server. In other embodiments, the global time server retrieves and adjusts the time more or less frequently than

once per day and the Net Time task 932 retrieves the time from the global time

server more or less frequently than once per day. For example, the global time

server can retrieve the UTC in intervals such as 5 minutes. Similarly, the Net Time task 932 can retrieve the time from the global time server in intervals such as 5

minutes.

The single board computer 540 is also programmed with a plurality of tasks for executing a variety of functions and stored data files. I-n the exemplary

embodiment, these tasks include a CmdExe task 900, a Payload task 908, a Work

task 912, a Scheduler task 930, CmdEpMon task 902, a CmdEpdWr task 904, a

CmdLog task 910, and a CmdFind task 906. Data files include a CHAT.DAT file

914, a FLND.DAT file 916, an ENDPT.DAT file 920, a CHANBITS.DAT file 918,

and a SPU.LOG file 928. The CmdExe task 900 executes all of the tasks received and processed by the

Com Serv task 938. The CmdExe task 900 executes commands requesting data to

be uploaded to the control server 934. The CmdExe task 900 also either executes (or

initiates execution of commands by other tasks) requesting a change in the operating

parameters of the single board computer 540. Examples include reporting lost

endpoint transceiver 112 to the control server 934, initiating commands to find a

new or a lost endpoint transceiver 112, commands to reassign an endpoint

transceiver 112 from one sub-sub-channel 406 to another sub-sub-channel 406, and

settings for operating parameters of the first and/or second DSPs 536 and 538. The Payload task 908 interfaces with the CPLD 530 and assembles each

command intended for the endpoint transceivers 112 into a data packet that is

decoded by the endpoint transceivers 112. The Payload task 908 outputs the data

packet to the CPLD 530, which buffers the data 700 and 702. The data packet is

then transmitted to the endpoint transceivers 112 as discussed herein in more detail. The Work task 912 interfaces with the CPLD 530 and the first and second DSPs 536 and 538. It sends instructions to the variable gain controller 528 to change operating parameters of the first and second digital signal processor 536 and

538, such as the gain of the various variable gain devices 502, 504, 506, 508, and

714. It also sends instructions to retrieve log files from the flash memory 532, the updated time base, the signal strength bits from the scaling function 666 and the data latch 816, the signal strength bits corresponding to signal strength from the LOS detection logics 692 and 818, and the noise word from the bit latch 694.

The CmdEpdWr task 904 interfaces with and retrieves data from the second

DSP 538 that the endpoint transceivers 112 transmitted, hi the exemplary

embodiment, it retrieves the data bit from the bit latch 660 about once every 20 minutes. When retrieving the data bit, the CmdEpdWr task 904 actually retrieves a bit from the bit latch 660 for the signal received from each of the three distribution

line conductors 302, 304, and 306. It determines which of the distribution line

conductors 302, 304, or 306 has the strongest signal strength and disregards the bits detected on the other two distribution line conductors. In the exemplary embodiment, the CmdEpdWr task 904 determines which distribution line conductor

302, 304, or 306 has the strongest signal by comparing the signal strength word

output by the scaling function 666 along the signal path 646 of the second DSP 538. After the data bit from the strongest distribution line conductor 302, 304, or

306 is selected, the CmdEpdWr task 904 performs an enor conection algorithm. In the exemplary embodiment, the CmdEpdWr task 904 executes a Bose Chaudhuri Hocquengham (BCH) code to conect enors in the data bits selected from the bit latch 660. After enor conection is complete, the data bit is stored in the data

fileENDPT.DAT 918.

The CHANBITS.DAT data file 918 stores a table that has a record for each

of the sub-sub-channels 406. hi the exemplary embodiment, each record in the table

has an entry identifying the sub-sub-channel 406 to which it conesponds, the serial

number of the endpoint transceiver 112 assigned to the sub-sub-channel 406, and

fields to store 63 data bits. Each endpoint transceiver 112 is assigned a unique

identification (I.D.) such as a serial number that can be formed with numbers, letters, and/or any other appropriate characters or signature. The data bits received from

each of the endpoint transceivers 112 are stored in the table in the record

conesponding to the sub-sub-channel 406 over which the data bit was received.

After 63 data bits are accumulated for a particular sub-sub-channel 406 and

stored in the table, the CmdEpdWr task 904 assembles the 63 data bits into a data

packet that includes the serial number of the endpoint transceiver 112 that

transmitted the data bits and stores the data packet in the ENDPT.DAT data file 920.

After the data packet is formed and stored in the ENDPT.DAT data file 920, the

CmdEpdWr task 904 clears the data bits from the table in the CHANBITS.DAT data

file 918. In one possible embodiment, the ENDPT.DAT 920 stores 30 data packets

for each sub-sub-channel 406. In the exemplary embodiment, each data packet

conesponds to about 24 hours worth of data collection, and thus, the ENDPT.DAT

920 data file stores 30 days worth of data for each sub-sub channel 406.

The CmdFind task 906 and the CmdEpMon task 902 cooperate to find and

monitor endpoint transceivers 112 that are connected to one of the distribution line

conductors 302, 304, or 306. The CmdFind task 906 makes the initial communication with each endpoint transceiver 112 to assign it a base frequency for

data transmission to the substation controller 108. The base frequency is within the bandwidth of an open or unused sub-sun channel 406.

When a new endpoint transceiver 112 is installed, the control server 934 is

programmed with the serial number for the new endpoint transceiver 112. The

command server 934 then sends the serial number and a find command to find the endpoint transceiver 112 to the CmdExe task 900 via the data link 936 and the Com

Serv task 938. The CmdExe task 900 passes the serial number and the find

command to the CmdFind task 906, which identifies an open sub-sub-channel 406 and maps the serial number to a base frequency within the open sub-sub-channel 406. The CmdFind task 906 relays the serial number, assigned or mapped base frequency, and find command to the Payload task 908, which assembles this data into a FIND ENDPOINT and passes the FIND ENDPOINT data packet to the

substation controller 108 as described in more detail herein. The substation controller 108 then transmits the FIND ENDPOINT data

packet downstream on the distribution line conductors 302, 304, and 306 and it is processed by the all of the endpoint transceivers 112. The transceivers process the

FIND ENDPOINT data packet. If the endpoint transceiver 112 has the serial number in the FIND ENDPOINT data packet, it transmits an acknowledge signal to the substation controller 108 indicating that it has received its assigned sub-sub-channel 406 and begins transmitting data to the substation controller 108. If the endpoint transceiver 112 does not have the serial number in the FIND ENDPOINT data packet, it discards the FLND ENDPOINT data packet and takes no further action with respect to the data packet. The CmdEpMon task 902 monitors the status of each sub-sub-channel 406.

It receives the signal strength word output by the scaling function 666, the LOS bit

output by the LOS detection logic 692, and the noise word output by bit latch 694.

Depending on the output of the LOS detection logic 692 and the bit latch 694, the

CmdEpMon task 902 changes the status of the sub-sub-channel 406 that is being

monitored as described below in conjunction with the CHAT.DAT 916 data file.

The CmdFind task 906 and the CmdEpMon task 902 interface with two data

files, FIND.DAT 914 and CHAT.DAT 916.

The FIND.DAT data file 914 stores information related to the process of

finding an endpoint transceiver 112. For example, when ever the CmdFind task 906

attempts to find and endpoint transceiver 112, it stores in the FIND.DAT data file

914 an ID. of the find command, or any other command that is being sent in the

FIND ENDPOINT data packet, the serial number of the endpoint transceiver 112 it

attempting to find, and the frequency it assigned to the endpoint transceiver 112.

The CmdFind task 906 also stores in FIND.DAT 914 a time stamp indicating when

transmission of the FIND ENDPOL T data packet was initiated and a time stamp

indicating when the substation controller 108 received a reply or acknowledgement

from the endpoint transceiver 112 that it was attempting to find.

The CHAT.DAT data file stores a database that has a record for each sub-

sub-channel 406. Each record includes a field identifying the base frequency for the

sub-sub-channels 406, the serial number of the endpoint transceiver 112 assigned to

that sub-sub-channels 406 (i.e., base frequency with in the sub-sub-channel 406),

and the status of the sub-sub-channel 406 (i.e., base frequency with in the sub-sub¬

channel 406). In the exemplary embodiment, the CHAT.DAT file has 9000 records, each conesponding to one of the 9000 sub-sub-channels 406. In one possible

embodiment, the status assigned to a sub-sub-channel 406 selected from the

following: Unused, Temporary, Found, Permanent, Blocked, Lost, Restored, and

Waiting. The Unused status is assigned to a sub-sub-channel 406 if it is not assigned

to an endpoint transceiver 112. When the CmdFind task 906 receives a find

command from the control server 934, it identifies a sub-sub-channel 406 having an

Unused status and assigns the base frequency within the identified unused sub-sub¬

channel 406 to the new endpoint transceiver 112. The CmdFind task 906 then inputs

the serial number, find command, and assigned base frequency to the Payload task

908 for processing and down stream transmission.

Upon receiving an acknowledgement from the endpoint transceiver 112, the

CmdFind task 906 changes the status of the sub-sub-channel 406 from Unused to

Temporary. The CmdExe task 900 then reports the found status of the new endpoint

transceiver 112 to the control server 934. Upon reporting the Found status of the

endpoint transceiver 112, the CmdFind task 906 changes the status of the assigned

sub-sub-channel 406 from Found to Permanent.

The status of a sub-sub-channel 406 is changed to Blocked to prevent an

endpoint transceiver 112 from being assigned that sub-sub-channel 406. A sub-sub-

channel 406 can have a Blocked status for a variety of reasons. An example includes

detection of a noisy sub-sub-channel 406. If the single board computer 540 runs

diagnostics and determines that a sub-sub-channel 406 is too noisy or if the Control

Server 934 sends an instruction to the substation computer to block out the noisy

sub-sub-station 406. In the exemplary embodiment, the single board computer 540 compares the value of the noise word received from the bit latch 694 on the second

DSP 538 to a predetermined value, and determines that the channel is too noisy if

the value of the noise word is greater than the predetermined value. The noise word

corresponds to the level of noise on the sub-sub-channel 406. If a Blocked status is assigned to a sub-sub-channel 406 that is currently

being used by an endpoint transceiver 112, the CmdFind task 906 will select a new

sub-sub-channel 406 that has an Unused status and instruct the Payload task 908 to

generate and initiate transmission of a FIND ENDPOINT data packet. The serial

number for the endpoint transceiver 112 is then stored in the record for the newly

assigned sub-sub-channel 406 and the status sequences through the Found and

Permanent states as described above.

The status of a particular sub-sub-channel 406 is changed to Lost if the

CmdEpMon task 902 detennines that there is a loss of power at the frequencies in

that sub-sub-channel 406. In the exemplary embodiment, the Work Task 912

receives the signal strength bit from the LOS detection logic 692 and provides the

signal strength bit to the CmdEpMon task 902. If the signal strength bit indicates

that the endpoint transceiver 112 assigned to that sub-sub-channel 406 is lost, the

CmdEpMon task 902 changes the status of the sub-sub-channel 406 to Lost and the

CmdExe task 900 reports the Lost status for the sub-sub-channel 406 to the control

server 934. The Control Server 934 then typically instructs the CmdFind task 906 to

find the lost endpoint transceiver 112 assigned to the sub-sub-channel 406. hi an

alternative embodiment, the CmdFind task 906 automatically attempts to find a lost

endpoint transceiver 112. When a lost endpoint transceiver 112 is found again, the status of the

assigned sub-sub-channel 406 is changed from Lost to Restored. The status remains

Restored while the Restored status is reported to the Control Server 934 by the

CmdExe task 900. The status of the sub-sub-channel 406 is changed from Restored

to Permanent upon reporting of the Restored status to the control server 934.

The status of a sub-sub-channel 406 is changed to Wait when the CmdFind

task 906 is instructed to find an endpoint transceiver 112 and the current state of the

sub-sub-channel 406 is cunently Temporary, Found, Permanent, or Restored. When

the status of the sub-sub-channel 406 is Wait, the CmdFind task 906 enters a wait

state for a predetermined period of time and then instructs the Payload task 908 to

generate and transmit a find packet after the predetermined period of time lapses.

Because the signal received from the endpoint transceivers 112 are at low frequency,

they take a while to decay. Given the frequencies used in the exemplary

embodiment, the signal transmitted by the endpoint transceiver 112 can take up to 20

minutes to decay, and the predetermined period of time in which the CmdFind task l

906 sits in the wait state is about 20 minutes or longer.

When an endpoint transceiver 112 is found after the wait state of the

CmdFind task 906 lapses, the status of the sub-sub-channel 406 is changed to Found

while the Found status is being reported to the Control Server 934 and then the

status is changed to Permanent after reporting is complete.

The Scheduler task 930 schedules all actions or events that are executed by

the single board computer 540 at a predetermined time or in periodic intervals. The

Scheduler task 930 schedules and initiates execution of a variety of commands

including commands executed internal to the substation processing unit 332. For example, the Scheduler task 930 instructs the CmdEpdWr 904 to retrieve a data bit

from the bit latch 660 about once every 20 minutes.

The Scheduler task 930 also schedules the transmission of data packets to the

endpoint transceivers 112. For example, the Scheduler task 930 instructs the

Payload task 908 to send a TIME data packet to the endpoint fransceivers 112 about

once every five minutes. The payload for the TIME data packet includes the current

date and time as determined by the Net Time task 932, which periodically retrieves

the cunent date and time from the time server and then tracks the current time

between download from the time server. If no other data is scheduled for

transmission to the endpoint transceivers 112 on the scheduled one-minute intervals,

the Scheduler Task 930 instructs the Payload Task to send a TIME data packet by

default.

An advantage of frequently sending Time data packets to the endpoint

transceivers 112 is that it facilitates synchronous operation, including data

collection, of all the endpoint transceivers that are downstream from the substation

controller 108. It also facilitates synchronous communication between the

substation controller 108 and the endpoint transceivers 112.

In another example and as explained in more detail with respect to the

CmdFind task 906, when a new endpoint transceiver 112 is to be installed or a lost

endpoint transceiver 112 is being restored, the Scheduler task 930 repeatedly

instructs the Payload task 908 to send a FLND ENDPOINT data packet until the new

endpoint transceiver 112 acknowledges assignment of a sub-sub-channel 406. The

endpoint transceiver 112 acknowledges assignment of a sub-sub-channel 406 by

transmitting a data packet that contains an acknowledgment to the substation controller 108 and generating an acknowledgement signal for observation at the endpoint transceiver 112 itself. This repeated transmission of a FIND ENDPOINT data packet is momentarily interrupted when the interval lapses for another scheduled transmission of a data packet, such as a TIME data packet. In one possible embodimenb the Scheduler task 930 instructs the Payload task to send a data packet to the endpoint transceivers 112 once every minute. An advantage of sending data packets in such short intervals facilitates rapid

deployment of newly installed endpoint transceivers 112. As a result, an installer

can receive quick confirmation that the endpoint transceiver 112 was successfully installed without waiting for more than a couple of minutes. The CmdLog task 910 records each command executed by the single board computer and stores the command in the SPU.LOG text file 928. The CmdLog task

910 also executes or initiates execution of various diagnostics of the single board computer 540 and the first and second DSPs 536 and 538, and records the results of

the diagnostics in the SPU.LOG text file 928. The architecture of the code executed by the single board computer 540 has many advantages. For example, it provides a scalable system because endpoint

transceivers 112 can be quickly added to the system by simply installing the endpoint transceiver 112 and entering the serial number of the endpoint transceiver

112 into the control server 934. It also provides flexibility for reassigning endpoint transceivers 112 from the SPU 332 located at one distribution substation 102 to another SPU 332 located at a different distribution substation 102. For example, if

service for an endpoint 104 is switched from one distribution substation 102 to a different distribution substation 102, the confrol server 934 instructs the single board computer 540 located at the different distribution substation 102 to execute a find

command for the endpoint transceiver(s) 112 located at the endpoint 104.

Another example occurs when a distribution substation 102 goes offline and

service for all of the endpoints 104 downstream from the offline distribution

substation 102 are switched to different distribution substations 102. In this

example, the control server 934 sends find commands for each of the endpoint

fransceivers 112 that were originally downstream from the offline distribution

substation 102 to the various distribution substations 102 that are still online. The

single board computer 540 at the online distribution substations 102 executes the

find commands. The single board computer 540 at each of the online distribution

substations 102 then registers the downstream endpoint transceivers 112 for which it

executed the find command and from which it received an acknowledgment.

Additionally, the single board computer 540 and the control server 934

cooperate to identify and locate faults in the distribution plant, hi one possible

embodiment, for example, when the status of a sub-sub-channel is Lost and reported

to the control server 934, the control server generates an enor report so that a service

technician can repair or replace the endpoint transceiver 112 assigned to the Lost

sub-sub-channel 406. An advantage of automatically reporting the Lost status of

sub-sub-channels 406 is that the utility can identify failures as quickly as they are

reported. Another advantage is that the energy utility can be made aware of the Lost

status (whether it due to a failed endpoint transceiver 112 or a loss of power) before

a failure is even reported by a customer and can more quickly take action to repair the failure than if they had to rely on reports from customers. Several sub-sub-channels being reported as Lost indicates that there is a failure in the distribution system itself. In another possible embodiment, if more

than one sub-sub-channel 406 has a Lost status, the control server 934 identifies the

common points in the distribution system that are upstream from the endpoint

transceivers 112 assigned to the Lost sub-sub-channels. Identifying the common

points of the distribution system in the manner helps to identify the potential points of the distribution system that failed resulting in a loss of power at the endpoints.

Again, because the single board computer 540 reports the Lost status of the sub-sub¬

channels 406, the control server 934 can quickly and automatically identify a power

failure and isolate possible points in the distribution system where the enor could

have occuned. As a result, technicians and/or engineers can be dispatched to repair

the failure.

In yet another possible embodiment, when some sub-sub-channels 406 have

a lost status and other sub-sub-channels 406 do not have a lost status, the control

server 934 identifies those points in the distribution system that are upstream from

the endpoint transceivers 112 assigned to the Lost sub-sub-channels 406 and not

upstream from the endpoint transceivers assigned to the sub-sub-channels 406 that

are not Lost. Identifying those points between the endpoint transceivers 112

assigned to the Lost sub-sub-channels 406 and the endpoint transceivers 112

assigned to the sub-sub-channels 406 that are not Lost helps the energy utility more

accurately isolate areas of the distribution system where the failure could be located.

The various embodiments described above are provided by way of

illustration only and should not be construed to limit the invention. Those skilled in

the art will readily recognize various modifications and changes that may be made to

the present invention without following the example embodiments and applications illustrated and described herein, and without departing from the true spirit and scope

of the present invention, which is set forth in the following claims.