U.S. patent application number 13/705130 was filed with the patent office on 2014-06-05 for wellsite system and method for multiple carrier frequency, half duplex cable telemetry.
This patent application is currently assigned to SCHLUMBERGER TECHNOLOGY CORPORATION. The applicant listed for this patent is SCHLUMBERGER TECHNOLOGY CORPORATION. Invention is credited to Milos Milosevic, Alex Olson, Kun Wang, Nalin Weerasinghe.
Application Number | 20140152459 13/705130 |
Document ID | / |
Family ID | 49911099 |
Filed Date | 2014-06-05 |
United States Patent
Application |
20140152459 |
Kind Code |
A1 |
Olson; Alex ; et
al. |
June 5, 2014 |
Wellsite System and Method for Multiple Carrier Frequency, Half
Duplex Cable Telemetry
Abstract
Methods and systems for multiple carrier frequency, half duplex
cable telemetry for a wellsite. The methods involve generating a
first type of bi-directional message in a first propagation mode,
generating a second type of bi-directional message in the first
propagation mode and in a second propagation mode, transmitting
over a cable operatively coupling a surface modem and a downhole
modem the first and second types of bi-directional message
sequentially in a plurality of time periods across a single
frequency bandwidth, and separating each of the first and second
types of bi-directional message from a most subsequently
transmitted one of the first and second types of bi-directional
message by a quiet time sample during which no message is
transmitted.
Inventors: |
Olson; Alex; (Houston,
TX) ; Weerasinghe; Nalin; (Sagamihara-shi, JP)
; Wang; Kun; (Sugar Land, TX) ; Milosevic;
Milos; (Houston, TX) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
SCHLUMBERGER TECHNOLOGY CORPORATION |
Sugar Land |
TX |
US |
|
|
Assignee: |
SCHLUMBERGER TECHNOLOGY
CORPORATION
Sugar Land
TX
|
Family ID: |
49911099 |
Appl. No.: |
13/705130 |
Filed: |
December 4, 2012 |
Current U.S.
Class: |
340/854.9 |
Current CPC
Class: |
H04B 3/23 20130101; G01V
11/002 20130101; H04B 3/32 20130101 |
Class at
Publication: |
340/854.9 |
International
Class: |
G01V 3/18 20060101
G01V003/18 |
Claims
1. A method for multiple carrier frequency, half duplex cable
telemetry for a wellsite, comprising: generating a first type of
bi-directional message in a first propagation mode; generating a
second type of bi-directional message in the first propagation mode
and in a second propagation mode; transmitting the first and second
types of bi-directional message over a cable operatively coupling a
surface modem and a downhole modem sequentially in a plurality of
time periods across a single frequency bandwidth; and separating
each of the first and second types of bi-directional messages from
a most subsequently transmitted one of the first and second type of
bi-directional message by a quiet time sample during which no
message is transmitted.
2. The method according to claim 1, wherein the cable comprises one
of a heptacable, a monocable, a coaxial cable, a wired drill pipe
conductor, and a slickline.
3. The method according to claim 2, wherein the first propagation
mode comprises a T5 propagation mode.
4. The method according to claim 2, wherein the second propagation
mode comprises a T7 propagation mode.
5. The method according to claim 1, wherein the first type of
bi-directional message comprises downlink and the second type of
bi-directional message comprises uplink.
6. The method according to claim 1, wherein a length of the quiet
time sample is determined based on one or more of: a length of the
cable, a time interval sufficient for a cross-talk ECHO to
dissipate, a travel latency time sufficient to avoid overlap of the
first and second types of bi-directional message, and a variable
user input.
7. A system for multiple carrier frequency, half duplex cable
telemetry for a wellsite, comprising: a surface modem that
generates one or more downlink messages in a first propagation
mode; one or more downhole tools that obtain measurements relating
to at least one of borehole characteristics and formation
characteristics, the one or more downhole tools coupled by a
toolbus to a downhole modem, the downhole modem generating one or
more uplink messages in the first propagation mode and in a second
propagation mode; a cable electrically coupling the surface modem
and the downhole modem; wherein the downlink and uplink messages
are transmitted over the cable sequentially in a plurality of time
periods across a single frequency bandwidth, each of the downlink
and uplink messages is separated from a most subsequently
transmitted one of the downlink and uplink messages by a quiet time
sample during which no message is transmitted.
8. The cable telemetry system according to claim 7, wherein the
cable comprises one of a heptacable, a monocable, a coaxial cable,
a wired drill pipe conductor, and a slickline.
9. The cable telemetry system according to claim 8, wherein the
first propagation mode comprises T5 propagation mode; and wherein
the second propagation mode comprises T7 propagation mode.
10. The cable telemetry system according to claim 7, wherein a
length of the quiet time sample is determined based on one or more
of: a length of the cable, a time interval sufficient for a
cross-talk ECHO to dissipate, a travel latency time sufficient to
avoid overlap of two types of bi-directional message, and a
variable user input.
11. A system for multiple carrier frequency, half duplex cable
telemetry for a wellsite, comprising: a surface acquisition unit
operatively coupling to a surface modem that generates one or more
downlink messages in a first propagation mode; a downhole modem
that generates one or more uplink messages in the first propagation
mode and in a second propagation mode; a downhole toolstring
comprising one or more downhole sensing tools that obtain
measurements relating to at least one of borehole characteristics
and formation characteristics, the downhole toolstring operatively
coupled to the downhole modem via a toolbus; a cable electrically
coupling the surface modem and the downhole modem; wherein the
downlink and uplink messages are transmitted sequentially in a
plurality of time periods across a single frequency bandwidth, each
of the downlink and uplink messages is separated from a most
subsequently transmitted one of the downlink and uplink messages by
a quiet time sample during which no message is transmitted.
12. The wellsite system according to claim 11, wherein a length of
the quiet time sample is determined based on one or more of: a
length of the cable, a time interval sufficient for a cross-talk
ECHO to dissipate, a travel latency time sufficient to avoid
overlap of the two types of bi-directional message, and a variable
user input.
Description
BACKGROUND
[0001] The following descriptions and examples are not admitted to
be prior art by virtue of their inclusion in this section.
[0002] Hydrocarbon fluids, such as oil and natural gas, may be
obtained from a subterranean geologic formation, referred to as a
reservoir, by drilling a well that penetrates a hydrocarbon-bearing
formation. A variety of downhole tools may be used in various areas
of oil and natural gas services. In some cases, downhole tools may
be used in a well for surveying, drilling, and production of
hydrocarbons. The downhole tools may communicate with the surface
via various telemetry systems. In some cases, the downhole tools
may comprise one or more individual modules in operative
communication with one another, such as a master module and
multiple slave modules. Examples of communication systems are
provided in U.S. Pat. Nos. 6,628,992, 7,181,515, and
20020178295.
[0003] With the increased precision of downhole tools and sensors,
relatively shorter time may be available to send increasingly
larger amounts of data. In addition to new modules and assemblies
being developed for downhole use on a continuing basis, tool bus
systems may facilitate communication between older and newer
generation modules in order to obtain the maximum service life from
existing modules.
[0004] Applications of disclosed embodiments of the present
disclosure are not limited to these illustrated examples, different
industrial applications may benefit from implementations of the
following disclosure.
SUMMARY
[0005] In at least one aspect, the disclosure relates to a method
for multiple carrier frequency, half duplex cable telemetry. The
method can include generating a first type of bi-directional
message in a first propagation mode. The method can include
generating a second type of bi-directional message in the first
propagation mode and a second propagation mode. The method can
include transmitting the first and second types of bi-directional
messages over a cable operatively coupling a surface modem and a
downhole modem sequentially in a plurality of time periods across a
single frequency bandwidth. The method can include separating each
of the first and second types of bi-directional messages from the
most subsequently transmitted message by a quiet time sample during
which no message is transmitted.
[0006] In at least one aspect, the disclosure relates to a system
for multiple carrier frequency, half duplex cable telemetry. The
system can include a surface modem that generates one or more
downlink messages in a first propagation mode. The system can
include one or more downhole tools that obtain measurements
relating to one of borehole characteristics and formation
characteristics, the downhole tools coupled by a toolbus to a
downhole modem. The downhole modem generates one or more uplink
messages in the first propagation mode and a second propagation
mode. The system can include a cable electrically coupling the
surface modem and the downhole modem. The one or more downlink
messages and the one or more uplink messages are transmitted over
the cable sequentially in a plurality of time periods across a
single frequency bandwidth, each of the downlink and uplink
messages separated from the most subsequently transmitted message
by a quiet time sample during which no message is transmitted.
[0007] In at least one aspect, the disclosure relates to a wellsite
system for multiple carrier frequency, half duplex cable telemetry.
The wellsite system can include a surface acquisition unit
including a surface modem that generates one or more downlink
messages in a first propagation mode. The wellsite system can
include a downhole modem that generates one or more uplink messages
in the first propagation mode and a second propagation mode. The
wellsite system can include a downhole tool string including one or
more downhole sensing tools that obtain measurements relating to
one of borehole characteristics and formation characteristics. The
downhole tool string may be operatively coupled to the downhole
modem via a toolbus. The wellsite system can include a cable
electrically coupling the surface modem and the downhole modem. The
one or more downlink messages and the one or more uplink messages
may be transmitted sequentially in a plurality of time periods
across a single frequency bandwidth, each of the downlink and
uplink messages separated from the most subsequently transmitted
message by a quiet time sample during which no message is
transmitted.
[0008] This summary is provided to introduce a selection of
concepts that are further described below in the detailed
description. This summary is not intended to identify key or
essential features of the claimed subject matter, nor is it
intended to be used as an aid in limiting the scope of the claimed
subject matter.
BRIEF DESCRIPTION OF THE DRAWINGS
[0009] Embodiments of systems, apparatuses, and methods for
multi-carrier frequency, half duplex cable telemetry are described
with reference to the following figures. Like numbers are used
throughout the figures to reference like features and
components.
[0010] FIG. 1 is a schematic representation illustrating a wellsite
with a borehole traversing a subsurface formation and having a
system for multiple carrier frequency, half duplex cable telemetry
in accordance with an embodiment of the present disclosure.
[0011] FIG. 2 shows a block diagram illustrating an example system
for multiple carrier frequency, half duplex cable telemetry in
accordance with an embodiment of the present disclosure.
[0012] FIGS. 3-1 through 3-3 are schematic diagrams illustrating
end views of a wireline heptacable in accordance with an embodiment
of the present disclosure.
[0013] FIG. 4 is a graph illustrating uplink and downlink spectrum
allocation transmitted in a multi-carrier frequency, full duplex
telemetry architecture.
[0014] FIG. 5 is a graph illustrating uplink and downlink spectrum
allocation transmitted in a multi-carrier frequency, half duplex
telemetry architecture in accordance with an embodiment of the
present disclosure.
[0015] FIG. 6 shows a flow chart illustrating a method of
multi-carrier frequency, half duplex cable telemetry in accordance
with at least one embodiment of the present disclosure.
[0016] FIG. 7 is a schematic diagram illustrating end views of
various propagation modes applied in wireline communication along a
heptacable in accordance with an embodiment of the present
disclosure.
[0017] FIG. 8 is a schematic block diagram of a cable telemetry
system illustrating the passage of signals therethrough in
accordance with an embodiment of the present disclosure.
[0018] FIG. 9 is a schematic diagram of a cable telemetry system
illustrating uplink and downlink therethrough in accordance with an
embodiment of the present disclosure.
[0019] FIG. 10 is a schematic diagram illustrating a cumulative
signal in accordance with an embodiment of the present
disclosure.
[0020] FIG. 11 is a schematic diagram illustrating cumulative
cross-talk sources on a wireline heptacable in accordance with an
embodiment of the present disclosure.
[0021] FIG. 12 is a schematic diagram illustrating a time-domain
diagram of uplink and downlink frames in a half duplex architecture
in accordance with an embodiment of the present disclosure.
[0022] FIGS. 13-1 and 13-2 are graphs illustrating oscilloscope
captures of uplink frames and downlink frame in accordance with an
embodiment of the present disclosure.
[0023] FIGS. 14-1 and 14-2 are graphs illustrating a shortened and
an unshortened ECHO impulse response after and before TEQ
respectively in accordance with an embodiment of the present
disclosure.
[0024] FIG. 15 is a graph illustrating cable delay on guard
interval length in accordance with an embodiment of the present
disclosure.
[0025] FIG. 16 is a graph illustrating a process to manage
superpacket data flow in accordance with an embodiment of the
present disclosure.
DETAILED DESCRIPTION
[0026] In the following description, numerous details are set forth
to provide an understanding of the present disclosure. However, it
will be understood by those skilled in the art that the present
disclosure may be practiced without these details and that numerous
variations or modifications from the described embodiments are
possible.
[0027] The disclosure relates to cable telemetry for a wellsite.
The cable telemetry may enable data reception from tools and send
data commands to downhole tools via, for example, a wireline
heptacable. "Cable telemetry" refers generally to communication
between an uphole modem and a downhole modem over a cable.
Multi-Carrier Frequency, Half Duplex Cable Telemetry Overview
[0028] Referring to FIG. 1, an example wireline logging operation
is illustrated with respect to the wellsite system 100 employed in
a wellbore 102 traversing a subsurface formation 104. A downhole
telemetry cartridge 110 is connected to a toolstring 116. In a
well-logging operation, a plurality of tools (e.g., 230, 230', etc.
of FIG. 2) may be connected in the toolstring 116. The tools of the
toolstring 116 communicate with the downhole telemetry circuits of
downhole telemetry cartridge 110 via a bi-directional electrical
interface.
[0029] In some embodiments, the tools of the toolstring 116 may be
connected to the telemetry cartridge 110 over a common data bus. In
some embodiments, each tool of the toolstring 116 may be
individually, directly connected to the telemetry cartridge 110. In
one embodiment, the telemetry cartridge 110 may be a separate unit,
which is mechanically and electrically connected to the tools in
the toolstring 116. In one embodiment, the telemetry cartridge 110
may be integrated into a housing of one of the well-logging tools
116.
[0030] The telemetry cartridge 110 is operatively coupled to a
wireline cable 114. The tools of the toolstring 116, including the
telemetry cartridge 110, may be lowered into the wellbore 102 on
the wireline cable 114.
[0031] A surface data acquisition computer 118 is located at the
surface end of the wireline cable 114. The surface data acquisition
computer 118 includes or couples to an uphole telemetry unit 112.
The data acquisition computer 118 may provide control of the
components in the toolstring 116 and process and store the data
acquired downhole. The acquisition computer 118 may communicate
with the uphole telemetry unit 112 via a bi-directional electrical
interface.
[0032] The uphole telemetry unit 112 may modulate downlink commands
from the acquisition computer 118 for transmission down the cable
114 to the toolstring 116, and demodulates uplink data from the
toolstring 116 for processing and storage by the surface data
acquisition computer 118.
[0033] The downhole telemetry cartridge 110 contains circuitry to
modulate uplink data from the tools of the toolstring 116 for
transmission up the wireline cable 114 to the surface data
acquisition computer 118 and to demodulate downlink commands or
data from the surface data acquisition computer 118 for the tools
of the toolstring 116.
[0034] A more detailed schematic view of an example cable telemetry
system 200 is shown in FIG. 2. The cable telemetry system 200 shown
includes a surface acquisition module/surface modem (DTM) 220
having a telemetry interface module (TIM) 222, which can be located
at the surface as a portion of or operatively coupled to the
surface data acquisition front end 119 (a component of surface data
acquisition computer 118 of FIG. 1). The front end 119 may be, for
example, eWAFE.TM. commercially available from SCHLUMBERGER.TM.
(see: www.slb.com).
[0035] The surface data acquisition front end 119 is coupled to the
wireline cable 114, and a downhole modem (DTC) 226 (as a portion of
the downhole telemetry cartridge 110 at the head of the toolstring
116 of FIG. 1). The tool string 116 includes a number of downhole
tools, 230, 230', 230'', 230''', etc. The downhole tools 230, 230',
etc., each containing a respective interface package, 232, 232',
232'', 232''', etc., through which they are in communication with
the DTC 226 via a tool bus 228. The downhole tools 230, 230',
230'', 230''', etc. may also have tool node controllers 233, 233',
233'', 233''', etc., respectively.
[0036] The cable telemetry system 200 may handle data flows in
opposite directions (i.e., from the tools 230, 230', etc.) via the
respective node and the tool bus 228. The flow extends to the DTC
226 to the DTM 220 over the cable 114 ("uplink"), and the reverse
direction from the DTM 220 to the DTC 226 and tools 230, 230',
etc., over the same path ("downlink"). The cable telemetry system
200 provides a communication path from the tools, 230, 230', etc.,
to the DTM 220 of the data acquisition computer 118 so that data
acquired by sensors 231, 231', 231'', 231''', etc. of the downhole
tools 230, 230', etc. can be processed and analyzed at the surface,
as well as communication between tools 230, 230', etc.
[0037] Each individual tool (230, 230', etc.) may include a node
command buffer (not shown) at the interface package 232, 232',
etc., as well as a logic controller of its own (not shown). The
surface acquisition front-end unit 119 may also include various
additional components, such as a power module 221, a depth and
tension module 223, and a flow controller software module (FEPC)
224.
[0038] The downhole telemetry cartridge 226 can include a downhole
master node controller 227 that may examine packets sent by each
respective tool 230, 230', etc. Data communicated in either
direction may be copied and buffered at the master node controller
227, and sent to the recipient.
[0039] A surface computer 234 can store and execute a surface data
dispatcher module 236 (which may be, in an embodiment, a software
data routing module, such as SCHLUMBERGER's.TM. MAXWELL.TM.
framework). The surface computer 234 can also store and execute a
plurality of surface tool-specific applications 238, 238', 238'',
238''', etc. that analyze and use data obtained, respectively, by
tools 230, 230', etc.
[0040] The wireline cable 114 may be a monocable, coaxial cable, or
a multi-conductor cable, such as a heptacable. Embodiments may
extend to wired drill pipe or in a system that uses a single
insulated conductor (wire) and borehole casing as a system. In
construction, the monocable may have a single conductor with a
return on the armor of the wires, while the coaxial cable can have
a single conductor with a return on the serve around it isolated
from the armor.
[0041] Heptacables may contain several electrical conductors, for
example, seven wires. The outer armor, for example, may have a
center conductor with six conductors wound helically around the
center conductor. Data may be simultaneously transmitted on these
several conductors. The distinct combinations of conductors used in
heptacable or other multi-conductor cables are referred to herein
as "propagation modes."
[0042] Cross-talk between several propagation modes, when in use
simultaneously, can be a source of noise in data transmission.
"Far-end cross-talk" is the interference between data transmitted
in a propagation mode and data transmitted in another propagation
mode. Far-end cross-talk can be caused, for example, by
imperfections in the symmetry or insulation of the wireline cable,
as well as circuitry that is used for interfacing to the cable
downhole and at the surface. Far-end cross-talk can impact both
data rate and robustness of the data transmission. Cross-talk may
limit the available data rate and reliability. For example,
cross-talk can lead to transmission failures during the progress of
a logging job.
[0043] A heptacable can provide for various signal propagation
modes, each of which transmits signals on a specific combination of
the seven conductors and armor. FIGS. 3-1 through 3-3 depict end
views of example wireline heptacable with conductors 351a-g in a
cable 350. These views highlight which conductors are activated for
T5, T6 and T7 propagation modes, respectively. These views depict
which conductors are activated per propagation mode for the T5, T6,
and T7 propagation modes. In the T5 mode of FIG. 3-1, a signal is
propagated on conductors 351c and 351f, and the return is provided
on conductors 351a and 351d. In the T6 mode of FIG. 3-2, the signal
is propagated on conductors 351b, 351d, and 351f, and the return is
on conductors 351a, 351c, and 351e. In the T7 mode of FIG. 3-3, the
signal is propagated on conductor 351g and the return is on
conductors 351a-351f and on the surrounding armor 350.
[0044] Referring now to FIG. 4, uplink and downlink signals are
transmitted in a multi-carrier frequency, full duplex architecture.
The graph shows signal magnitude (y-axis) vs. frequency (x-axis)
for a transmitted uplink and downlink spectrum allocation. The full
duplex architecture transmits an uplink message 460 and a downlink
message 462 simultaneously or overlapping in time. The transmission
is with a first propagation mode being devoted to uplink messages
and a second propagation mode being devoted to downlink messages
across a common frequency bandwidth.
[0045] Uplink message 460 propagates in the first propagation mode
during time period 461. Downlink message 462 propagates in the
second propagation mode during time period 463. The time period 463
overlaps in time with a portion of time period 461, but in
differing frequency bands. In the example illustrated in FIG. 4, T7
propagation mode may be used for downlink messages, while T5
propagation mode may be used for uplink messages. Given that uplink
data volume is greater than downlink volume data, downlink messages
overlap in time with a fraction of uplink messages.
[0046] In full duplex architectures simultaneously using two modes
(e.g., both T5 and T7) for data transmission in either uplink or
downlink transmission, several effects can be noted. Using two
modes for uplink results in both modes using overlapping frequency
bands. Due to the properties of wireline cables, frequency band
overlap results in some far-end cross-talk that causes the uplink
signal transmitted on one mode to be seen in the other mode's
uplink receiver.
[0047] For example, the T5 downhole transmitter's signal (intended
for the surface T5 receiver) may also be detectable in the surface
T7 receiver due to far end cross-talk (FEXT). The magnitude of the
FEXT depends mostly on the construction and maintenance of
heptacable symmetry. As there may not be a dedicated mode for
downlink, it is possible there may be another cross-talk called
"ECHO" in that a receiver may see a portion of the locally
transmitted signal on the same mode. As the cross-talk sources are
additive in nature, the maximum achievable data rate may be lowered
as the cross-talk sources increase in magnitude.
[0048] ECHO, in turn, may result in two effects on long cables
using full duplex architecture. On long cables (with high
attenuation), the receiver may use a high gain in order to minimize
analog to digital converter (ADC) quantization noise. However, the
gain may also amplify the ECHO, and the receiver can be saturated.
As saturation is a non-linear distortion, the signal-to-noise ratio
(SNR) may be affected across frequencies, not just those used by
the downlink. Thus, for a full duplex architecture, receiver gain
may be purposefully limited to reduce the chance of saturation.
However, such a measure can increase the impact of ADC and receiver
noise and, consequently, degrade SNR, resulting in lower data
rates.
[0049] A second result with ECHO is that frequency-domain
separation between uplink and downlink is possible when the
shortened impulse response of the main channel and the ECHO channel
can be simultaneously contained within a discrete multi-tone (DMT)
cyclic prefix. The impulse-response of the ECHO may be affected by
the impedance mismatch (both in magnitude and phase) between the
transceiver and the cable's impedance. This may make the ECHO
impulse response long and difficult to shorten.
[0050] In some cases, the ECHO may be sufficient to disrupt the
integrity of the telemetry link in full duplex architecture.
However, the ECHO may have an impact in the full duplex
architecture where the uplink and downlink happen simultaneously
and the separation is frequency based. The architecture disclosed
herein, referred to as "half duplex" architecture, can avoid the
effects of ECHO by separating uplink and downlink in time. The half
duplex architecture does not suffer from ECHO as there may not be
simultaneous transmission of uplink and downlink.
[0051] Referring now to FIG. 5, uplink signals and downlink signals
transmitted in a multi-carrier frequency, half duplex architecture
graphically plotted as signal magnitude versus frequency are shown.
The graph illustrates signal magnitude (y-axis) vs. frequency
(x-axis) showing transmitted uplink and downlink spectrum
allocation. The half duplex architecture described herein retains
the discrete multi-tone modulation technique but transmits uplink
and downlink during separate time intervals. The half duplex
architecture described herein also inserts a block of "guard"
samples where both the uplink and downlink are quiet and do not
transmit any signal. The guard period gives time for the ECHO
impulse response caused by the last transmitted frame to dissipate
before the other direction's receiver begins receiving data.
[0052] Uplink signals 570 are transmitted in both T7 mode 571 and
T5 mode 573 during time periods for uplink 576. Downlink messages
572 are transmitted in T5 mode 573 during time periods for downlink
575. Time periods for downlink 575 and time periods for uplink 576
do not overlap, thus the signals are transmitted sequentially, not
simultaneously.
[0053] A guard period 574 is interjected between uplink signals 570
and downlink signals 572. The guard period, during which signals
are not transmitted in either direction, allows sufficient time for
the residual energy caused by the most recently transmitted frame
to dissipate before the other direction's telemetry unit begins
receiving data. A long guard interval may affect the duty cycle
(and data rate) due to the increased time overhead.
[0054] In embodiments employing shorter cables, the effect of ECHO
may be less severe, and the quiet period 574 can be shortened. In
an embodiment, the guard period 574 may be configurable at runtime
to optimize the uplink duty cycle. In an embodiment, the guard
period 574 may be a parameter that is configured by the surface
modem and transmitted to the downhole modem during a training phase
as disclosed in related and commonly assigned application entitled
"TOOLSTRING COMMUNICATION IN CABLE TELEMETRY" (Attorney Docket
Number IS12.2286) filed concurrently herewith. In an embodiment,
the length of the guard period 574 may be transmitted to the other
modem during training to allow for future modification/tuning, even
if initially fixed at a particular value.
[0055] At the surface modem, the guard period 574 provides time for
the downlink ECHO to decay away before uplink frames are received.
In a half duplex system, the downlink frames represent overhead
time that is not spent transmitting uplink data. Thus, to maximize
the uplink data rate, the number of downlink frames per
uplink/downlink transmission may be minimized. This overhead, the
guard period 574, may be minimized by transmitting as many uplink
frames as possible before switching to downlink.
[0056] An embodiment may have symmetrical use of the propagation
modes where uplink and downlink transmissions are present in the
propagation modes. Another embodiment may assign downlink or uplink
transmissions to one or more propagation modes depending on the
data communication rates of the system.
[0057] When the downlink portion of the telemetry link is
established, the downhole receiver may drop samples in order to
align itself with the cable delay. If the uplink frames are
transmitted at a fixed interval, then an insufficient guard period
may cause a portion of the received downlink frame to overlap the
transmitted uplink frame. The overlap may be seen as ECHO in the
downlink receiver and implies the previously mentioned effects
(clipping, inter-carrier interference, etc.) associated with ECHO
and full duplex systems. A half duplex system may have uplink and
downlink using the same frequencies. Thus, the guard period may be
selected as longer, at least, than the cable round-trip delay to
prevent the downlink and uplink frames from overlapping.
[0058] Referring now to FIG. 6, a flow chart for a method 600 of
multi-carrier frequency, half duplex cable telemetry. The method
600 begins with generating 680 downlink messages in a first
propagation mode in a surface telemetry unit. In an embodiment, the
cable includes a wireline heptacable. In an embodiment, the first
propagation mode includes, for example, T5 propagation mode. FIG. 7
depicts end views of seven different propagation modes T1 through
T7 that may be applied in wireline communication along a
heptacable.
[0059] Referring still to FIG. 6, the method continues with
generating 682 uplink messages in the first propagation mode and a
second propagation mode in a downhole telemetry unit. In an
embodiment, the second propagation mode includes, for example, T7
propagation mode.
[0060] The method continues with transmitting 684 the downlink
messages and the uplink messages over a cable operatively coupling
the surface telemetry unit and the downhole telemetry unit
sequentially in a plurality of time periods across a single
frequency bandwidth.
[0061] The method continues with separating 686 each message from
the subsequent message by a quiet time sample during which no
signal is transmitted. In some embodiments, the length of the quiet
time sample is determined based on the length of the cable. In some
embodiments, the length of the quiet time sample is determined
based on at least one of a time interval sufficient for a
cross-talk ECHO to dissipate, and a travel latency time sufficient
to avoid overlap of the downlink messages and the uplink messages.
In an embodiment, the length of the quiet time sample includes a
variable user input.
Multi-Carrier Frequency, Half Duplex Cable Telemetry
Explanation
[0062] Wireline operations may be performed on a number of
different cable types based on the number of conductors in the
cable. The monocable may have a single conductor with a return on
the armor wires, while the coaxial cable can have a single
conductor with a return on the serve around it isolated from the
armor. The heptacable may have, for example, seven conductors and
armor.
[0063] Enhanced Digital Telemetry System (EDTS) version 2.0
(commercially available from SCHLUMBERGER TECHNOLOGY
CORPORATION.TM. see: www.slb.com) cable telemetry is intended to
work with the heptacable-construction cable. The heptacable may
have seven natural modes. EDTS 2.0 cable telemetry uses two of
these modes: T5 and T7. T5 mode is a differential mode that uses
wires 2, 3, 5 and 6, and has immunity to common mode noise. T7 uses
wire 7, a return on the remaining wires and the armor, and, hence,
behaves as coaxial cable. These two modes may possess the select
telemetry channel characteristics in earlier telemetry systems, and
may be used in wireline telemetry systems. FIG. 7 shows an end view
of seven different propagation modes T1 through T7 applied in
wireline communication along a heptacable.
[0064] A purpose of cable link telemetry (as shown in FIGS. 8 and
9) may be to enable data reception from the tools of the toolstring
and to enable sending commands to the tools via the wireline
heptacable. FIG. 8 depicts data transfer between a MAXWELL.TM. tool
application, an uphole modem, a downhole modem and a tool. The
communication between the MAXWELL.TM. tool application and an
uphole modem and the communication between the downhole modem and
the tool involve uplink data and commands. The communication
between the uphole modem and the downhole modem involves modem
signals (cable telemetry). The cable telemetry may provide delivery
from uphole to downhole modem only. The EDTS 2.0 components may be
used to provide delivery from MAXWELL.TM. tool application to
tool.
[0065] FIG. 9 shows data transfer between a surface modem (e.g.,
with MAXWELL.TM. tool applications) and a downhole toolstring. The
transfer involves downlink via logging cable from a cable telemetry
to a toolstring with a downhole modem m, tools 1 and 2, and uplink
back thereto.
[0066] Cable telemetry is configured to avoid altering data or
commands. EDTS 2.0 cable telemetry is built in the uphole and
downhole modems. This uses a multi-carrier modulation technology
called DMT. This technology may also be used in other telemetry
systems, such as EDTS (a predecessor of EDTS 2.0 commercially
available from SCHLUMBERGER TECHNOLOGY CORPORATION.TM. see:
www.slb.com). While EDTS cable telemetry uses a single uplink
channel on T5 and a single downlink channel on T7, EDTS 2.0 may use
both T5 and T7. Using both T5 and T7 may provide uplink data
transmission at nearly double the data rate, and may use T5 for
downlink transmission.
[0067] Due to the properties of wireline cables, there may be some
FEXT that causes the uplink signal transmitted on one mode to be
seen in the other mode's uplink receiver. For example, the T5
downhole transmitter's signal (intended for the surface T5
receiver) may also be detectable in the surface T7 receiver due to
FEXT. The magnitude of the FEXT may depend on the construction and
maintenance of heptacable symmetry. As there may be no dedicated
mode for downlink, there may be another FEXT called "ECHO" in that
a receiver may see a portion of the locally transmitted signal on
the same mode. As the FEXT source may be additive in nature, the
maximum achievable data rate may be lowered as the cross-talk
sources increase in magnitude.
[0068] As shown in FIG. 10 a cumulative signal (i.e. Actual signal
seen at receiver) may be generated from a combination of a receiver
signal (RX), cross-talk (ECHO), near end cross-talk (NEXT), and
noise effects added.
[0069] FIG. 11 shows a schematic of cumulative cross-talk (ECHO)
sources on a wireline heptacable (i.e. transmission cable). The
heptacable may include NEXT, FEXT and other components. The ECHOS
extend from TX A and B to hybrids and back to RX C and D on one
side, and ECHOS extend from TX C and D to hybrids and back to RX A
and RX B on another side of the heptacable.
[0070] The ECHOs may result in two effects on long cables for the
full duplex architecture. On long cables (e.g., with high
attenuation), the receiver may use a high gain in order to minimize
ADC quantization noise. However, the gain may also amplify the
ECHO, and the receiver can be saturated. As saturation is a
non-linear distortion, it may impact the SNR on various
frequencies, not just those used by the downlink. So with a full
duplex architecture, receiver gain may have to be limited to reduce
the chance of saturation. However, such a measure may increase the
impact of ADC and receiver noise and consequently degrade SNR,
which may result in lower data rates.
[0071] A second effect with ECHO is that frequency-domain
separation between uplink and downlink may be possible when the
shortened impulse response of the main channel and the ECHO channel
can be simultaneously contained within a DMT cyclic prefix.
Additionally, the impulse response of the ECHO can be affected by
the impedance mismatch (both in magnitude and phase) between the
transceiver and the cable's impedance. This may make the ECHO
impulse response long and difficult to shorten.
[0072] The effect of ECHO can be severe in full duplex
architectures. The level of ECHO may be as much as 50 dB more than
NEXT on a telemetry interface module (e.g., eWAFE.TM.), and can
have a negative impact on the uplink data rate for full duplex
architecture. In some cases, the ECHO could be sufficient to
disrupt the integrity of the telemetry link.
[0073] ECHO is a property of the physical interfaces, such as the
hybrid and its impedance matching to the cable, as well as the
relative size of the signal to the ECHO which may depend on the
cable characteristics. The ECHO may have an impact in the full
duplex architecture where the uplink and downlink happen
simultaneously and the separation is frequency based. Thus, the
half duplex architecture may avoid the effects of ECHO by
separating uplink and downlink in time. The half duplex
architecture may not suffer from ECHO as there is no simultaneous
transmission of uplink and downlink.
[0074] To overcome the effects detailed above, the following
physical layer architecture may support, for example, about 150
KHz, 200 KHz, and 250 KHz bandwidths, have an uplink/downlink
structure, about 36 uplink frames per 1 downlink frame at 150 KHz,
200 KHz, 250 KHz bandwidths, and about 100 fixed samples of quiet
time between uplink/downlink at 150 KHz, 200 KHz, 250 KHz
bandwidths. The quiet time value may be transmitted to other modes
during training to allow for future modification. The architecture
may also allow for single-mode operation on T5 or dual uplink mode
(T5+T7) operation. The architecture may also allow variable
downlink bit allocation (150 KHz: .about.900 bits total, .about.600
data bits, .about.300 maintenance bits), thus allowing for data
rates to be met at 150 KHz bandwidth and above.
[0075] FIG. 12 shows a time-domain diagram of uplink and downlink
frames in a half duplex architecture. This time-domain diagram
depicts T5 and T7 schedules. The T5 schedule includes 1 T5 downlink
(DL) frame, 50 T5 quite samples, 36 T5 uplink (UL) frames and 50 T5
quiet samples. The T7 schedule includes a quite frame, 50 T5 quite
samples, 36 T7 uplink frames and 50 T7 quiet samples. FIGS. 13-1
and 13-2 show oscilloscope captures for a time domain diagram.
These figures showing uplink frames and downlink frame
therebetween. FIG. 13-2 is a zoomed-in version of FIG. 13-1.
[0076] The half duplex architecture retains the discrete multi-tone
modulation technique, but transmits uplink and downlink during
separate time intervals. It also inserts a block of "guard" or
"quiet" samples where both the uplink and downlink do not transmit
any signal. This period gives time for the (shortened) ECHO impulse
response caused by the last transmitted frame to dissipate before
the other direction's receiver begins receiving data.
[0077] Sampling timing synchronization may be performed for the
systems herein. Example sampling timing synchronizations that may
be used are described in U.S. Pat. No. 6,778,622. In the half
duplex system of the present disclosure, there are two ways
sampling timing can be maintained. The surface FPGA sample
frequency offset may change value slowly (for example, in an
embodiment, each few minutes). This mechanism may track without
updates for a few frames using the previously known offset. Since
per-frame clock shifts by design occur with probability less than
one half (50/100) and the downlink+quiet interval may be short (in
an example embodiment, less than 2 frames), the per-frame clock
shifts can account for the lack of adjustment during this interval.
The first few uplink frames may experience a slightly larger shift,
but the effect of this shift can be remedied through digital
rotations.
[0078] Linear regression estimated during the uplink interval could
be used to extrapolate the pilot angle through the quiet &
downlink periods. This may permit the clock cycle shift to be
applied during the downlink & quiet intervals using the
predicted pilot tone phase. As EDTS may perform linear regression
over hundreds of frames (e.g., extrapolation over one or two frames
can be performed). By the next uplink frame, the extrapolation
would predict the true pilot tone phase taking into account any
clock cycle adjustments.
[0079] The system may also be used with variable quiet time. At
first glance, the quiet period could be an entire frame's worth of
samples, as the ECHO impulse response may die off exponentially.
However, such a long guard interval may lower the duty cycle (and
data rate) due to the increased time overhead. Furthermore, the
quiet period may be long enough for the shortened (post-TEQ) ECHO
impulse response to dissipate. Although the time domain equivalent
(TEQ) may or may not explicitly try to shorten the ECHO impulse
response.
[0080] FIGS. 14-1 and 14-2 are graphs showing a shortened ECHO
impulse response after TEQ, and a corresponding un-shortened ECHO
impulse response before TEQ, respectively. Each graph depicts value
(y-axis) versus sample number (x-axis). These figures show the ECHO
impulse response on 36 k feet (ft) (10.97 km) 7-48 A heptacable for
200 KHz bandwidth. Although clipping may occur in the un-shortened
impulse response, the duration in which the un-shortened impulse
response has large enough amplitude to cause clipping can be
neglected.
[0081] Using the above observations, in an example embodiment, 50
samples of quiet time may be sufficient at 150 KHz bandwidth on 36
k ft (10.97 km) of 7-48 A heptacable. This amount of quiet time may
not be sufficient for 200 KHz and 250 KHz. On shorter cables the
effect of ECHO may be less severe and the quiet interval may be
shortened. Thus it may be useful to have the quiet portion
configurable at runtime to optimize the uplink duty cycle. It could
be a parameter that is configured by the surface modem and
transmitted to the downhole modem during training. The quiet
interval length may be transmitted to the other modem during
training to allow for future modification/tuning (even if it is
initially fixed at a particular value, such as 100 samples).
[0082] The length of the guard interval is a design parameter that
may be used for half duplex architecture. The guard interval
provides time for the downlink ECHO to decay away before the uplink
frames are received. Thus, the guard interval may be longer than
the (shortened) ECHO impulse response (at the surface).
[0083] The guard interval may provide a "cushion" at the downhole
modem, between the downlink receiver receiving downlink frames and
the uplink transmitter transmitting uplink frames. In EDTS, the
downlink frame boundaries may be aligned to the uplink frame
boundaries on the surface modem. This alignment may result from the
downhole (uplink) transmitter initiating the telemetry link at
power up. The same alignment may carry over into EDTS 2.0. When the
downlink portion of the telemetry link is established, the downhole
receiver may have to drop samples in order to align itself with the
cable delay. If the uplink frames are transmitted at a fixed
interval, then an insufficient guard period may cause a portion of
the received downlink frame to overlap the transmitted uplink
frame.
[0084] The overlap may be seen as ECHO in the downlink receiver and
implies the previously mentioned issues (e.g., inter-carrier
interference, etc.) associated with ECHO and full duplex systems.
Moreover a half duplex system may have uplink and downlink using
the same frequencies. Thus, the guard period may be selected as
longer than the cable round-trip delay to prevent the downlink and
uplink frames from overlapping.
[0085] FIG. 15 graphically shows an effect of delay on guard
interval length of the frames. This figure includes three graphs of
T5 downhole (conceptual), T5 surface, and T5 downhole (actual)
(y-axis) versus time (x-axis), respectively. Each figure also
includes an uplink and a downlink with two guards and downlink
therebetween. The downlink is positioned between the two guards.
Cable delays extend between the uplinks of the top two graphs.
Another cable delay extends from the downlink of the middle graph
to the downlink of the bottom graph.
[0086] Table 1 provides example (one-way) cable latencies measured
for 150 KHz bandwidth:
TABLE-US-00001 TABLE 1 T5 Delay T7 Delay Cable (Samples) (Samples)
7 k ft P 4 4 21 k ft A 13 11 31 k ft NT 23 21 31 NT + 7p 27 25
[0087] In an example embodiment, a 50 sample guard interval may
work for up to about 36 k ft (10.97 km) 7-48 A cable, the
round-trip delay of such a cable (at about 150 KHz) is
approximately 45 samples (36 k ft (10.97 km)/21 k ft (6.4
km).times.13 samples.times.2). At 200 KHz the round-trip delay is
about 60 samples. If a 3 tap TEQ is used, the cyclic prefix of the
first uplink frame provides an amount of additional guard time
(e.g., about 14 samples). In short, the guard interval may be set
to a value greater than 50 samples for 200 KHz (or about 250 KHz)
bandwidth operation, for example on longer cables. At least 100
samples may allow for half duplex operation on cables up to about
80 k ft (24.38 km) at about 150 KHz and 48 k ft (14.63 km) at 250
KHz. Furthermore, a 100 sample guard interval may not decrease the
duty cycle.
[0088] In terms of the physical layer, the following tasks may be
performed at each uplink frame: update the frequency-domain
equalizer (FEQ), update the mean-squared noise (MSN) measurement
(for SNR), generate bit swapping tone list for the uplink channel
(each .about.2 seconds), and optionally bit swapping tone list for
the downlink channel.
[0089] In the half duplex architecture of the present disclosure,
these operations could be split across data frames and/or performed
during the quiet and downlink intervals. The SYNC frame may not be
called for in EDTS 2.0 and may be reduced in the half duplex
architecture.
[0090] For the downlink, half duplex EDTS 2.0 may not call for
bit-swapping. In an embodiment, the minimum downlink margin may be
about more than 20 dB on about 36 k ft (10.97 km) cable across
various tones on a DMT testbed. Optionally, downlink bit-swapping
could be implemented using the first or last uplink frame as a
reference.
[0091] Downhole on-board-programming (OBP) mode may also be used.
Where the downlink data rate may be about 8-9 kbps, few (if any)
downlink commands may be sent after a toolstring is initialized. In
some circumstances, a large amount of downlink data may be sent.
For example, downhole EPROM programming and/or downhole firmware
upgrade operations call for a large amount of data to be
transferred downhole. In an embodiment, these operations may be
performed in a shop or lab, rather than during a field job and may
call for a minimal uplink data rate. For these operations, the
number of uplink frames between downlinks could be reduced in order
to support faster downlink data rates.
[0092] Implementation options, such as those in Table 2, may be
used to insure compatibility with previous systems. Uplink data
rates represent the total sum of T5 and T7 data rates.
TABLE-US-00002 TABLE 2 Uplink Data (As high as possible) Uplink
Maintenance >=4.48 kbps Downlink Data >=15 Hz, >=8.96 kbps
Downlink Maintenance >=8.235 Hz, >=4.48 kbps
[0093] In 150 KHz system, a DMT frame of 256 carriers and 16 sample
cyclic prefix may result in a frame rate of 568.18 Hz. In a half
duplex system, the quiet period is an overhead that decreases the
uplink and downlink throughput. This overhead may be minimized by
transmitting as many uplink frames as possible before switching to
downlink. The downlink latency requirement (e.g., greater than
about 15 Hz) constrains the maximum number of uplink frames between
downlinks. Since 568.18 Hz/15 Hz=37.87, there may be a downlink
frame, for example, at least each 37 frames. In other words, there
may be no more than 36 frames per 1 downlink frame (assuming that
an entire downlink packet could be transmitted in one downlink
frame, not including time for the quiet interval). Similarly, there
may be a downlink frame at least each 50 frames for 200 KHz
bandwidth, 63 frames for 250 KHz bandwidth.
[0094] In a half duplex system, the downlink frames represent time
that is not spent transmitting uplink data. The uplink data rate
may be maximized by minimizing the number of downlink frames per
uplink/downlink transmission. This may cause relative higher number
of bits to be allocated to the downlink frames.
[0095] About 1500 total bits may be called for the downlink to
support the data and maintenance channels at a 15 Hz transmission
rate. As there are fewer than 256 carriers per frame, using a fixed
bit allocation (for various cables) may not be a viable option. For
example, putting 4 bits on 200 tones may result in margins in the
lower tones (where SNR is good) and a lower margin on the higher
tones where the SNR is poor. In an embodiment, the SNR may change
by as much as 20 dB from the peak to the minimum. Such an
allocation may also be prone to narrowband interference generated
by the power systems, etc.
[0096] For half duplex, the downlink should dynamically allocate
bits based on the measured SNR. The per tone gains may not be
varied (e.g. "fine gains"), but the number of bits may be varied
according to SNR. Such an allocation may improve the minimum
downlink margin across various tones.
[0097] When considering half duplex systems, the "duty cycle"
(portion of time that the uplink is actively transmitting data) may
be computed. Duty cycle can then be used to compare the efficiency
of a half duplex system versus that of an EDTS (which transmits
data 68/69=98.6% of the time). The number of bits per frame to
sustain a particular data rate and to calculate the downlink
latency (due to quiet portions) may also be computed.
[0098] The half duplex duty cycle may be represented by Equations 1
and 2:
UplinkDutyCycle HalfDuplex = ( # UplinkFrames UplinkFrameTime ) T
UplinkDutyCycle HalfDuplex = ( # UplinkFrames UplinkFrameTime ) T
Equation ( 1 ) T = ( # UplinkFrames UplinkFrameDuration ) + ( #
DownlinkFrames DownlinkFrameDuration ) + ( # QuietSamples
QuietSampleDuration ) Equation ( 2 ) ##EQU00001##
where T is the total time.
[0099] If both uplink and downlink use the same FFT size (512
samples) and the same cyclic prefix (16 samples), then the frame
length is 528 samples and the duty cycle and consequent data rate
become the following in Equations 3 and 4:
UplinkDutyCycle HalfDuplex = ( # UplinkFrames ) ( # UplinkFrames )
+ ( # DownlinkFrames ) + 2 ( # QuietSamples / 528 ) UplinkDataRate
HalfDuplex = UplinkDutyCycle HalfDuplex * UplinkBitsPerFrame
Equation ( 3 ) DownlinkFrequency HalfDuplex = ( RawFrameRate ) ( 1
DownlinkFrame ) ( # UplinkFrames ) + ( 1 DownlinkFrame ) + ( 2 *
QuietSamples / 528 ) = ( RawFrameRate ) 1 ( # UplinkFrames ) + 1 +
( 2 * QuietSamples / 528 ) RawFrameRate = ( Bandwidth / 2 ) / ( 528
Samples ) Equation ( 4 ) ##EQU00002##
[0100] Table 3 below show various parameters for various
bandwidths:
TABLE-US-00003 TABLE 3 Frame Uplink Downlink Downlink Downlink
Downlink Bandwidth Rate Uplink Downlink Quiet Duty Duty Frequency
Data Maintenance (KHz) (Hz) Frames Frames Samples Cycle % Cycle %
(Hz) Bits/Frame Bits/Frame 150 568.18 36 1 0 97.3 2.7 15.36 583 292
150 568.18 36 1 50 96.8 2.7 15.28 586 293 150 568.18 36 1 100 96.3
2.7 15.20 589 295 150 568.18 36 1 250 94.9 2.6 14.97 598 299 150
568.18 36 1 528 92.3 2.6 14.57 615 308 200 757.58 36 1 0 97.3 2.7
20.48 438 219 200 757.58 36 1 50 96.8 2.7 20.37 440 220 200 757.58
36 1 100 96.3 2.7 20.27 442 221 200 757.58 36 1 250 94.9 2.6 19.96
449 224 200 757.58 36 1 528 92.3 2.6 19.43 461 231 250 946.97 36 1
0 97.3 2.7 25.59 350 175 250 946.97 36 1 50 96.8 2.7 25.46 352 176
250 946.97 36 1 100 96.3 2.7 25.33 354 177 250 946.97 36 1 250 94.9
2.6 24.96 359 180 250 946.97 36 1 528 92.3 2.6 24.28 369 185 Uplink
interval increased to maintain same downlink data rate as 150 KHz:
200 757.58 49 1 0 98.0 2.0 15.15 591 296 200 757.58 49 1 50 97.6
2.0 15.09 594 297 200 757.58 49 1 100 97.3 2.0 15.04 596 298
The downlink duty cycle can be calculated in a similar manner. It
is also related to the frequency at which downlink transmissions
are sent. For sending one downlink frame per uplink, the following
is found:
[0101] In an embodiment, downlink packets may arrive continuously
at a rate not less than about 15 Hz. If one downlink frame is
transmitted between uplinks at 150 KHz bandwidth, then no more than
about 36 uplink frames may be transmitted by the downlink frame as
shown in Table 3 above. The length of the quiet interval may have a
minor effect on the effective downlink frequency.
[0102] In an embodiment, if an entire downlink maintenance packet
(maximum of 160 bits) is transmitted at 15 Hz (in a single downlink
frame), then less memory may be used for the downhole uplink
superpacket buffers and the master-slave clock synchronization may
be improved over EDTS. As the 15 Hz rate is used for the downlink
data channel (and it is possible to allocate more than about 160
bits/frame for the downlink maintenance channel), the changes in
EDTS 2.0 may be obtained at low cost.
[0103] The calculations involving the number of uplink frames per
downlink frame consider the minimum number for 150 KHz bandwidth.
In an embodiment, the downlink data and maintenance channel may
increase in both rate and frequency due to the faster frame rate
achieved at 200 KHz bandwidth. For half duplex EDTS 2.0 at 200 KHz
bandwidth, the downlink data rate and frequency may not increase.
The downlink maintenance channel frequency also may not increase
since the uplink superpacket rate is not dependent on bandwidth. To
maintain 15 Hz downlink rate with a frame rate of 757.58 Hz (200
KHz bandwidth), there could be as many as 49 uplink frames per
downlink frame (757.58/15=50.503). With no quiet period, the
maximum 150 KHz duty cycle is 36/37=97.3%. The maximum 200 KHz duty
cycle would be 49/50=98%. The difference in duty cycles may affect
implementation. For example (not including quiet time), a system
that achieves 3 Mbps from a 97.3% duty cycle may achieve 3.022 Mbps
with a 98% duty cycle.
[0104] FIG. 16 shows a process for managing superpacket data flow.
This figure depicts uplink and downlink signals using a T5 and T7
cable. In an embodiment, the EFTB 2.0 is designed to be more
independent from the modem process compared to the EFTB in EDTS. At
the top of FIG. 16, four signals can be seen. From left to right,
an outgoing uplink signal out to, for example, T5 cable propagation
mode from an uplink digital to analog converter, an incoming
downlink signal from, for example, T5 cable propagation mode, into
a downlink analog to digital converter, an incoming downlink signal
from, for example, a T7 cable propagation mode to a downlink analog
to digital converter, and an outgoing uplink signal to, for
example, a T7 cable propagation mode, from an uplink digital to
analog converter.
[0105] In specific detail, the downlink signals will be discussed
first (the middle incoming signals at the top of FIG. 16
respectively). The T5 downlink signal (2.sup.nd from the left at
the top of FIG. 16) passes from the downlink analog to digital
converter to a downlink process (in internal SRAM) of the T5 uplink
MCM via input SP0RX (DMA0), to be buffered at internal SRAM in the
various buffers: Receiver Symbol Buffer, Receiver Superpacket
Buffer, and Receiver Superpacket Transfer Buffer, respectively. The
T5 downlink signal is then passed to the downlink path switch on
the EFTB+MCM via input SP1TX (DMA3). On the EFTB+MCM, the signal is
passed via input SP1RX to buffers downlink superpacket receiver
buffer, to downlink packet buffer, to EFTB downlink packet
transmitter buffer, to downlink process, and out to EFTB on input
SP0TX (DMA2). The data can also be buffered in external SRAM
3.times.4M at downlink superpacket buffer temporarily, and then
passed to downlink superpacket receiver buffer, downlink packet
buffer, and so on.
[0106] The T7 downlink signal (3.sup.rd from the left at the top of
FIG. 16) is passed from the downlink analog to digital converter to
the T7 downlink MCM on the input SP0RX (DMA0) to downlink process
(in internal SRAM) of the T7 downlink MCM. The signal is buffered
in the various buffers of the internal SRAM of the T7 downlink MCM:
Receiver Symbol Buffer, Receiver Superpacket Buffer, and Receiver
Superpacket Transfer Buffer, before passing to the downlink path
switch on the EFTB+MCM via input SP1TX (DMA3). Once on the
EFTB+MCM, the T7 downlink signal can pass along the same data
pathway that the T5 downlink signal follows through the
EFTB+MCM.
[0107] An uplink signal enters the EFTB+MCM from the EFTB at the
bottom left input on FIG. 16, passing to a digital signal processor
DSP via input SP0RX (DMAO). The uplink signal then can pass through
various buffers: EIP Queue Message Buffer, Uplink Controller
Buffer, and Uplink Packet Buffer. The data signal is then passed to
the internal superpacket transmitter buffer (and optionally,
divided into headers and trailer portions of the signal, buffered
at uplink superpacket headers buffer and uplink superpacket
trailers buffer). The signal is then passed to the uplink process
of the DSP of the EFTB+MCM, and passed to the SP router of the
Field Programmable Gate Array (FPGA) of the EFTB+MCM that
determines whether the uplink is to be transferred via T5
propagation mode or T7 propagation mode. Uplink signals determined
to be transmitted using T5 propagation mode are passed to the T5
uplink MCM, and via input SP1RX (DMA1) passed to the internal
superpacket buffer in internal SRAM of the T5 uplink MCM. The
signal can be buffered in external SRAM at external memory
superpacket list nodes temporarily. The signal then passes to the
transmit buffer 209.times.32 bits, to the internal SRAM uplink
process, before passing to the output via input SP0TX (DMA2) to the
uplink digital to analog converter at top right of FIG. 16.
[0108] Uplink signals determined to be transmitted using T7
propagation mode are passed to the T5 uplink MCM, and via input
SP1RX (DMA1) passed to the internal superpacket buffer in internal
SRAM of the T7 uplink MCM. The signal can be buffered in external
SRAM at external memory superpacket list nodes temporarily. The
signal then passes to the transmit buffer, to the internal SRAM
uplink process, before passing to the output via input SP0TX (DMA2)
to the uplink digital to analog converter at top left of FIG.
16.
[0109] Data flow may be uplinked (UL SP Data Flow) using a dual
uplink EDTS 2.0. The uplink superpackets at about 4 Mbps may be
generated in the EFTB 2.0 DSP at a rate of 125 Hz (or about twice
as much as the EFTB rate (62.5 Hz)). These packets may be routed
(based on buffer availability) to the T5 and T7 DSPs resulting in
each FEDSP receiving the overall superpackets rate at approximately
62.5 Hz. Since each of the uplink DSPs operates with (roughly) the
same data rate as standard EDTS, the acknowledgement of previously
transmitted superpackets may be managed by supporting a downlink
with double the 8.23 Hz acknowledgement rate of EDTS.
[0110] Data flow may be uplinked (UL SP Data Flow) using single
mode EDTS 2.0. The UL superpackets at about 2 Mbps may be generated
in the EFTB 2.0 DSP at a rate of 62.5 Hz. These packets may be sent
to the T5 DSP. Since this DSP operates with the same data rate as
EDTS, the acknowledgement of previously transmitted SP may be
managed by supporting a DL at least the same 8.23 Hz
acknowledgement rate of EDTS.
[0111] Data flow may be downlinked (DL SP Data Flow). In half
duplex EDTS 2.0, a mode (T5) is used for downlink. A single-mode
downlink transmission path may be supported either on the T5 or T7
DSP where the T5 mode would be selected as the default due to
robustness against the common-mode noise.
[0112] Although a few example embodiments have been described in
detail above, those skilled in the art will readily appreciate that
many modifications are possible in the example embodiments without
materially departing from this disclosure. Accordingly, such
modifications are intended to be included within the scope of this
disclosure as defined in the following claims. In the claims,
means-plus-function clauses are intended to cover the structures
described herein as performing the recited function and not simply
structural equivalents, but also equivalent structures. Thus,
although a nail and a screw may not be structural equivalents in
that a nail employs a cylindrical surface to secure wooden parts
together, whereas a screw employs a helical surface, in the
environment of fastening wooden parts, a nail and a screw may be
equivalent structures. It is the express intention of the applicant
not to invoke 35 U.S.C. .sctn.112, paragraph 6 for any limitations
of any of the claims herein, except for those in which the claim
expressly uses the words "means for" together with an associated
function.
* * * * *
References