Note: Descriptions are shown in the official language in which they were submitted.
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
1
METHOD AND APPARATUS SUPPORTING TDD/TTY
MODULATION OVER VOCODED CHANNELS
BACKGROUND OF THE INVENTION
I. Field of the Invention
Generally, the present invention relates to the field of telecommunication
devices for the deaf (TDDs) or text telephone yokes (TTYs). More particularly,
the invention relates to modification of standard vocoder operation to enable
reliable transport of TDD/TTY signals within a telecommunication system. The
system may include wireless links.
II. Description of the Related Art
Many deaf or hearing-impaired people use communication terminals
specifically constructed and designed to enable them to communicate over
standard telephone lines. Such devices, referred to as telecommunication
devices for the deaf (TDDs) or Text Telephone Yokes (TTYs), are collectively
referred to as TDDs in this application. Typically, TDDs include a keyboard
and a display connected to a telephone via a modem
(modulator/demodulator). The modem is built into the TDD and is either
directly connected to a telephone line or coupled by an acoustic coupler to a
normal telephone handset. TDDs are capable of transmitting information over
telephone lines by means of coded tones to other TDDs connected at opposite
ends of the telephone line through another modem. These tones are referred to
as low activity communications because the frequency and amplitude
envelopes remain relatively constant.
The code and protocol that is in widespread conventional use for TDD
communications is an idiosyncratic one. The code set, known as Baudot, and
the communication protocol (TDD protocol) evolved historically at a time when
many telecommunication devices for the deaf were based on mechanical or
electromechanical devices rather than electronic devices. Accordingly, the TDD
protocol was constructed for a set of constraints that no longer are relevant
to
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
2
present day devices. Those constraints work to create a code protocol and a
telecommunication network of users and devices operating under that protocol,
that is somewhat antiquated.
Traditionally, TDD communications are conducted at 50 Baud (45.5 Baud
in some countries), representing a transfer of 6 characters per sec. Other
protocols now available for TDD communications incorporate higher Baud
rates, such as the ASCII (American Standard Code Information Interchange)
and enhanced Baudot protocols. Regardless, a normal TDD communication
character set consists of characters that are 5 bits long. These characters
are
analogous to a letter in an alphabet where each letter represents a word or
idea.
A character is grouped with overhead information bits prior to transfer, where
each group of bits to be transferred has a duration or unit interval equal to
22
milliseconds. For example, under conventional TDD protocol, a group of bits to
be transferred comprises 8 bits: a start bit (one source or zero bit), five
bits
representing the character, and at least one and 1/2 bits marking the stop
point
of the transfer group.
Compared to modern telecommunication systems, TDD transmissions
occur at a snail's pace. A bigger problem is that TDD signals are
substantially
constant. These slow paced, monotone signals can create havoc in digital
telecommunication systems that transmit higher activity signals at very high
rates, and especially in telecommunication systems that include wireless
links.
One example of such a telecommunication system is a code division multiple
access (CDMA) system having a large number of wireless subscriber units.
Each subscriber unit has a transceiver and communicates within the system
through satellite repeaters or terrestrial stations referred to as cells. Each
cell
includes a physical plant called a base station. A cell covers a limited
geographic area and routes calls carried over subscriber units to and from the
telecommunication network via a mobile switching center. When a subscriber
moves into the geographic area of a new cell, the routing of that subscriber's
call may be eventually made through the new cell by a process called a
"handoff."
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
3
A subscriber unit, generically referred to as a cell phone, transmits a
signal that is received by a base station. The signal is relayed to a mobile
switching center that routes the signal to a public switched telephone network
(PSTN) including telephone lines or other subscriber units. Similarly, a
signal
may be transmitted from the PSTN to a subscriber unit via a base station and a
mobile switching center
The interface between the subscriber unit and the base station is referred
to as the air interface. The telecommunications industry association (TIA) has
provided a standard for CDMA call processing on the air interface entitled "IS-
95 Mobile Station - Base Station Compatibility Standard for Dual Mode
Wideband Spread Spectrum Cellular System." Addendum to IS-95 are
provided as Telecommunications Service Bulletins (TSB). The standard IS-95 +
TSB74 includes provisions for service negotiation on the air interface and is
incorporated herein by reference.
Service negotiation is critical to successfully transmit any
communication, especially a low activity TDD communication, over a digital
telecommunication system. One problem with modern systems, including the
one described above, is that a vocoder - a device used in the system to encode
a
voice or TDD analog signal into a digital signal, and to decode a digital
signal
into a voice or TDD analog signal- has difficulty in handling the
substantially
monotone signal and slow speed dictated by the TDD protocol. In current
systems, a low activity communication signal such as a TDD communication
would probably be treated by the vocoder as background noise or signal
interference and be disregarded.
What is needed is an invention that can easily be integrated into existing
communication systems and can be capable of reducing frame error rates by
invoking a protocol to be used by the vocoders during transmission of the low
activity communication signal.
The invention should be compatible with wireless telecommunication
modulation systems, such as CDMA systems, servicing large numbers of
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
4
system users. A more robust discussion of CDMA systems and techniques used
in multiple access communication systems may be found in U.S. Pat. No.
4,901,307, entitled "SPREAD SPECTRUM MULTIPLE ACCESS
COMMUNICATION SYSTEM USING SATELLITE OR TERRESTRIAL
REPEATERS," assigned to the assignee of the present invention and
incorporated by reference herein. Further, the invention should also be
compatible with other modulation systems and techniques used in other types
of communication systems, such as time division multiple access (TDMA),
frequency division multiple access (FDMA), and amplitude modulation (AMPS)
schemes.
SUMMARY OF THE INVENTION
Broadly, the present invention involves the modulation of a low activity
communication by a telecommunication system using encoded signals and
increased transmission power levels. More particularly, the invention concerns
a method that uses specialized encoding, decoding, or both, on a low activity
communication signal to minimize a transmitted signal's frame error rate. The
invention also provides for decoding a low activity signal by looking at "soft
bits" contained in erred frames, or in frames adjacent to an erred frame, in
an
attempt to determine the content of the original frame.
Certain disclosed embodiments of the invention provide unique
decoding methods for a TDD signal that was encoded using standard encoding
protocol. In one embodiment, the decoder may compare a frame containing
transmission errors (erred frame) with a vocoded frame from a known TDD
signal and determine the most likely vocoded frame that was transmitted. In
another embodiment, the decoder may review adjacent frames to determine the
most likely vocoded frame that was transmitted but received in error. In yet
another embodiment, the decoder can be modified to include a signal enhancer
or repeater that "cleans up" corrupted bits in the transmitted frame before
the
decoding methods are applied upon the transmitted frames. And although a
TDD communication is discussed throughout this application, it should be
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
understood that any slow or low activity communication may be transmitted
using this invention.
Another embodiment of the invention provides for decoding as
discussed above but invokes vocoder parameters that are different from
5 standard vocoder parameters. When a TDD signal is received, the encoder
switches to "Baudot encoding mode," notices the decoder of the protocol
change, and uses channel coding redundancy to further improve the decoder's
chances of determining the correct TDD signal sent even if it is contained in
a
bad frame. This version of the invention replaces standard vocoder parameters
with vocoder "signatures" that are better spaced apart, thus making it easier
to
distinguish between tones.
Another version of the invention provides for encoding a TDD signal in
vocoder frames using redundancy, but doing the encoding across numerous
vocoder frames. The information is interleaved across "N" frames so that if a
frame is lost, the decoder can extract necessary information from adjacent
frames to determine the content of the lost frame.
Yet another version of the invention provides a cost-efficient system for
combining the aforementioned encoding and decoding methods of the
invention with methods for controlling transmission power levels. Modifying
the design of a standard vocoder chipset within a mobile station is expensive,
which could bar the implementation of the aforementioned encoding and
decoding methods. However, a communication system can be implemented
wherein the aforementioned methods are utilized in the system's base stations,
but not in the mobile stations. Frame error rates for low activity
communication signals transmitted from the base station to the unmodified
vocoder in a mobile station can be minimized using methods for controlling
transmission power levels.
The invention provides its users with numerous advantages. One
advantage is that a TDD message can be transmitted using a digital
transmission medium having wireless links. Yet another advantage is that a
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
6
TDD device can be connected to a mobile device or subscriber's unit, such as a
digital cellular telephone, connected to the telecommunications system by a
wireless link. The invention also provides a number of other advantages and
benefits that should become even more apparent after reviewing the following
detailed descriptions of the invention.
BRIEF DESCRIPTION OF THE DRAWINGS
The nature, objects, and advantages of the invention will become more
apparent to those skilled in the art after considering the following detailed
description in connection with the accompanying drawings, in which like
reference numerals designate like parts throughout, and wherein:
FIGURE 1A is a block diagram of hardware components and
interconnections of a telecommunications system incorporating wireless links
in
accordance with one embodiment of the invention;
FIGURE 1B is a block diagram of a vocoder capable of implementing the
present inventions encoding and decoding methods coupled to a prior art
noticing apparatus in accordance with one embodiment of the invention;
FIGURE 2 illustrates a typical prior art TDD communication device used
in accordance with one embodiment of the invention;
FIGURE 3 shows a traffic channel frame format for a rate set 1 used by a
variable rate vocoder;
FIGURE 4 is a flow diagram of a method aspect in accordance with one
embodiment of the invention;
FIGURE 5 illustrates a block diagram of a wireless telecommunication
system configured according to an embodiment of the invention;
FIGURE 6 illustrates a block diagram of a wireless telecommunication
system configured according to an embodiment of the invention; and
FIGURE 7 is a flow diagram of a method for controlling transmission
power levels between a base station and a mobile station.
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
7
DETAILED DESCRIPTION OF SELECTED EMBODIMENTS
FIGURES 1 through 7 illustrate examples of various method and
apparatus aspects of the present invention. For ease of explanation, but
without
any limitation intended, these examples are described in the context of a TDD
communication device attached to a digital telecommunication system
incorporating wireless links, one example of which is described below.
HARDWARE COMPONENTS AND INTERCONNECTIONS
FIG. 1 illustrates one type of telecommunications system 100 including
wireless links and a TDD communication device (TDD) 200 as used in the
present invention. As shown in detail in FIG. 2, TDDs usually include a
keyboard and a display that are connected to a telephone via a modem
(modulator/demodulator). The modem is built into the TDD and is either
directly connected to a telephone line or coupled by an acoustic coupler to a
normal telephone handset. TDDs are capable of transmitting information over
telephone lines by means of coded tones to other TDDs, such as TDD 102
shown in FIG.1, connected at opposite ends of a telephone line through another
modem.
In digital telecommunications systems using wireless links, the TDD 200
may be coupled to a subscriber unit 104 that is used in the telecommunications
system 100 to transmit received signals. Exemplary embodiments of a
subscriber unit 104 are digital signal telephones, such as the Q-800
manufactured by Qualcomm Incorporated, and commonly referred to as cell
phones. The subscribers unit 104 as shown in FIG. 1 includes a noticing
apparatus 106 communicatively coupled to circuitry of the subscribers unit
104.
A hardwire 108 may be used to connect the TDD 200 to the subscribers unit 104
via the noticing apparatus 106, or a device port may be used. Examples of such
a noticing apparatus and device ports are disclosed in the U.S. Patent
application entitled "METHOD AND APPARATUS FOR ESTABLISHING
TDD/TTY SERVICE OVER VOCODED CHANNELS, serial number
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
8
09 / 114,344, filed July 13, 1998, assigned to the assignee of the present
invention
and incorporated by reference herein.
The device port may be configured to receive a low activity
communication device attachment such as a plug, connector, or receiver. These
items are commonly used today for connecting telephone and computer
equipment, and are readily available from electronics suppliers. The device
port interfaces with the attachment to communicatively connect a low activity
communication device (not shown) such as the TDD 200 to the subscriber unit
104 of the telecommunications system100. The device port allows information
to be exchanged between a low activity communication device and the
subscriber unit 104. Regardless of whether a device port or a hardwire is
used,
the noticing apparatus 106 allows for the system 100 to be noticed that a TDD
signal needs to be transmitted.
Returning to FIG. 1, after the noticing apparatus 106 receives the low
activity communication signal, the signal is processed by the subscriber unit
104. Very basically, a signal for transmission is created that includes the
information contained in the low activity signal. Because the
telecommunications system 100 has been noticed that a low activity signal is
being transmitted, the system adapts to assure a decipherable transmission
occurs. For example, an analog signal received from the analog circuitry 228
shown in FIG. 2 normally would undergo signal or "voice" processing
including digitizing the signal, setting a transmit power level to protect
against
signal fading during transmission, compressing the signal, and filtering.
These
functions may be performed by the circuitry (not shown) of the subscriber unit
104 that includes a vocoder. Depending upon the signal received, a variable
rate vocoder - generically referred to in this application as a vocoder - may
dynamically determine and negotiate service within the telecommunications
system 100 to provide successful transmission and decoding of the signal. This
negotiation involves establishing the values for multiple parameters, such as
the
rate the vocoder should use, the transmission power, and compression
technique. A fuller discussion concerning the processing of signals for
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
9
transmission in telecommunication system may be found Electronic Industry
Association standard TIA/EIA/IS-95-A entitled "Mobile Station-Based Station
Compatibility Standard for Dual-Mode Wideband Spread Spectrum Cellular
Systems, referred to as "IS-95" and incorporated by reference herein, and
other
transmission standards, including standard vocoder protocol, are well known
in the art.
However, when a low activity signal is received, a vocoder may identify
the signal as either noise, a pause, or a signal not intended to be
transmitted.
Simply, a vocoder doesn't know what service to use because it cannot identify
the low activity signal received. By noticing the system 100 that a low
activity
signal is being sent, the vocoder will establish the service needed to assure
the
best possible transmission and decoding of the signal.
After the low activity communication signal has been processed and the
service determined, a signal may be transmitted using an antenna 112 over a
wireless link 114. The digitized signal is received by another antenna 116 at
a
remote location, such as a base station 118, and processed by base station
circuitry (not shown) including a vocoder 120. Various based station circuitry
arrangements for telecommunications systems are well known in the art, and a
further understanding may be found in TIA/EIA/IS-95-A referenced above. By
processing the signal after receipt, a low activity signal reflecting the
information contained in the transmitted low activity signal may be delivered
to the low activity device 102 via communication link 120. A second noticing
apparatus 106 is shown coupled to the base station 106. This provides for a
low
activity signal to be sent from the low activity communication device 102 back
to the TDD communication device 200.
Communication link 120 appears bifurcated to emphasize that the base
station 118 may not be connected directly to the low activity device 102. The
base station 118 is usually connected to a standard PSTN switching station
commonly used by telephone companies for coordination of telephone calls and
the low activity device 102 is connected to the PSTN. In another embodiment, a
second mobile station (not shown) connected to the low activity communication
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
device 102 may be linked to the base station 118. Further, the
telecommunication system may include mobile switching stations as mentioned
above.
Shown in FIG. 2 is a schematic block diagram of the circuitry of a typical
5 TDD device 200, either a standard or enhanced TDD, operating in accordance
with the present invention. In the TDD device 200 of FIG. 2, a keyboard 202 is
provided into which the user may input data characters. The output of the
keyboard 202 is connected to a processor 204 that serves to control the
circuit
elements contained in FIG. 2. Characters that are received or transmitted by
the
10 processor 204 are also displayed on a display 206. Optionally, the same
characters received or transmitted may be reproduced on a device such as
printer 208. Some TDD devices may not have a printer, although it standard for
TDDs to have a visual display of some kind so that a user can see the
characters
being typed and received. The keyboard 202 thus functions as an input source
of data characters to the processor 204 while either or both the display 206
and
the printer 208 serve as local destinations for the data stream characters.
The processor 204 may be connected by a suitable data and address bus
that would typically be used for this type of application by one schooled in
the art. In FIG. 2, the bus 210 connects a read only memory (ROM) 212 to a non-
volatile random access memory (NVRAM) 214. Appropriate control lines 216
and 218 are connected from the processor 204 to the ROM 212 and the NVRAM
214 providing interactive control of these units. The ROM 212 is intended to
permanently store the program that dictates the operation of the processor 204
as well as certain data used by the program. For example, special character
strings for machine-to-machine communication and for synchronizing two
TDDs in an enhanced operating mode may be stored. The NVRAM 214 is used
as a buffer, a floating storage place for data coming into or out of the TDD
device 200, and for storage of standard messages as entered by the user
through
the keyboard 202 and intended for rapid. Other circuitry configurations may be
used, such as combining the microprocessor 202 with the ROM 212 and the
NVRAM 214 in a single integrated circuit.
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
11
Also connected to the processor 202 in FIG. 2 is a telephone keypad 220
that permits the entry of telephone numbers for dialing by the processor 202
through telecommunications system 100. A standard telephone handset 224
rests on a cradle 226 that incorporates a switch (not shown) indicating
whether
the handset 224 is in use and thus removed from the cradle 226.
The processor 204 is communicatively connected through analog
circuitry 228 to the telecommunications system 100. This connection is shown
as a hardwire connection 230, but may be any type of connection that can
communicatively link the analog circuitry 228 with the telecommunications
system 100. The analog circuitry 228 provides a connection between the
handset and the processor 202 allowing both Baudot tones and dialing tones to
be received by the telecommunications system 100. The analog circuitry 228
provides an interface of voice information to and from the handset 224. The
analog circuitry 228 of the TDD device 200 is connected to the
telecommunication system 100 using a connector such as the device discussed
above.
Despite the specific foregoing descriptions, ordinarily skilled artisans
having the benefit of this disclosure will recognize that the apparatus
discussed
above may be implemented in a telecommunications system of different
construction without departing from the scope of the present invention. As a
specific example, multiple subscriber unit 104 may be linked to the base
station
118, or the low activity communication device 200 may be integrated with the
subscriber unit 104.
OPERATION
After a TDD signal is received, vocoders used by the system 100 during
processing of the signal are noticed or detect that a low activity signal has
been
received for transmission and may use an eighth rate traffic channel frame
format to transmit the signal. However, adaptation of the following methods
for quarter-to-full rate traffic channel transmissions may be accomplished, as
discussed below.
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
12
FIG. 3 shows a typical variable rate vocoder frame format for a traffic
channel using a rate set 1. The variable rate vocoder produces a frame every
20
milliseconds using Code Excited Linear Prediction (CELP) techniques that are
well known in the art. The frames may be formatted in full, half , quartex or
eighth rate formats depending upon voice activity. If a Baudot tone is
received,
the variable rate vocoder will usually detect low activity and use the eighth
rate
format, assuming the standard vocoder currently in use can detect that a
signal
is being sent. Commonly, a Baudot signal will be treated as noise and
generally
ignored.
Full rate refers to the fact that each bit contained in each frame is not
repeated. Half-rate refers to sending the same number of bits per frame, but
each bit is repeated once in the frame; that is, each unique bit will appear
twice
in the frame. Quarter-rate refers to each unique bit appearing four times per
frame, and so on. The more repetitively a bit of information is sent, the less
total information is sent per frame. At full rate the signal is sent at a
higher
power because a given bit is sent only once. This full rate power level is
referred to as the reference power for purposes of this application. Because
bits
are repeated at lower rates, a reduced power level is used because the power
for
each repeated bit is accumulated over the frame. Assuming a fixed minimum
power is used for the transmission, a full rate transmission will contain more
frame errors than would a half rate transmission of the same information.
Typically, the power level is set based upon a selected frame error rate
(FER) for the transmitted signal as received at a remote location, also
referred to
as the target of the transmitted signal, such as the subscriber unit. A
desired
FER is selected because when a low activity signal is being sent, the actual
FER
increases using current methods. This selected FER range is between a 0.1%
and a 1.0% error rate, but may be greater or lesser if necessary for
preservation
of the quality of the transmitted signal. Preferably, an FER of 0.2% is
desirable
for low activity signals.
In the present invention, implementing specialized encoding and
decoding techniques controls the frame error rate. In the circumstance that
the
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
13
disclosed techniques fall short of a desired FER - in this case FER being
defined
as the total number of erred frames even after reconstruction of vocoder frame
information - methods to adjust transmission power levels can also be used
along with the specialized encoding and decoding techniques. Typically, the
vocoders will be locked at full rate and the transmission power will be
increased for transmitting low activity signals. It should be realized that
any
increase required would still be less than the increase required if the
present
encoding/decoding techniques were not implemented.
A. Decoder Using Soft Bits
In one embodiment, when a TDD call is received, the system 100 is either
noticed or detects the call type. The system 100 processes the call from TDD
unit 200 for transmission using standard processing techniques known in the
art. When the frame is received at a remote point, for example base station
118,
the call is decoded using the present invention. If a frame error has occurred
in
the physical layer, that is, if the frame does not pass the checksum as
described
within IS-95, the frame is still delivered to the vocoder 120 for decoding.
Delivering the erred frame to the vocoder is currently not done in standard IS-
95 implementations. Bits contained in an erred frame are referred to as "soft
bits" because they may not all be in error and information may be gleaned from
them individually to reconstruct information contained in erred frames.
However, detecting or being noticed that a TDD call has been received,
the vocoder decoder in the present invention processes erred frames by looking
at the vocoder parameters received and comparing these parameters against
"signatures" of TDD modulation signals or tones as seen in the vocoder
parameter space. This compares the vocoder parameters of stored vocoded
TDD tones with those received. This comparison results in a determination
being made as to which TDD signal was most likely received.
For example, suppose a vocoder representation of a Baudot tone of "0"
is represented as sixteen "0"s in sequence, and that the representation of a
Baudot tone of "1" is represented as sixteen "1"s. The present method
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
14
considers these to be voice-parameter-space signatures. For the following
examples, three layers are identified as:
vocoder frame boundaries: I -- voc frame 'N' -- I
baudot tone boundaries: I -- baudot 'X' -- I , and
received vocoder parameter: 000000000000000 or
11111111111111111.
Assume that the vocoder decoder receives the following parameters:
{erred frame}
I --voc frame 1-- I --voc frame 2-- I --voc frame 3-- I --voc frame 4 -- I --
voc
t ' 1' -- I -- baudot '0' -- I -- baudot '0' -- I -- baudot '1' -- I -- baudot
'0' -- I
1111111100000000000000000000000000000110000111111111110000000000000000
nnnn~
[frame errors]
The decoder recognizes the baudot tone boundaries and recognizes that the
received parameters for the second baudot '0' are closer to '0' than '1.' The
decoder decides on baudot tone '0' and modifies the suspected error bits
before
decoding. For the next baudot tone '1,' the decoder recognizes that the
vocoder
parameters are closer to '1' than '0' and modifies the bits accordingly. The
decoder now uses the following sequence to produce a corrected TDD signal:
{corrected frame}
I --voc frame 1-- I --voc frame 2-- I --voc frame 3-- I --voc frame 4 -- I --
voc
t '1' -- I -- baudot '0' -- I -- baudot '0' -- I -- baudot '1' -- I -- baudot
'0' -- I
1111111100000000000000000000000000000011111111111111110000000000000000
This example shows the error transitions to occur at a frame boundary, which
isn't always the case. If these transitions commonly fall within a frame,
another
version of the invention can be used as follows.
If the vocoder decoder receives an erred frame where the erred bits are
contained within the frame, the vocoder may look to adjacent non-erred or
"good" frames to reconstruct the erred frame. The adjacent frame will contain
a
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
portion of the baudot tone that was lost in the erred frame. For example,
suppose the following signal is received:
{erred frame}
I --voc frame 1-- I --voc frame 2-- I --voc frame 3-- I --voc frame 4 -- I --
voc
5 t '1' -- I -- baudot '0' -- I -- baudot '0' -- I -- baudot '1' -- I --
baudot '0' -- I
1111111100000000001111111111110000000001111111111111110000000000000000
nnnnnnnnnn
[frame errors]
The vocoder parameters for the second baudot tone '0' are too ambiguous to
10 make an accurate decision on the tone because the number of '0's is almost
the
same as the number of '1's in the vocoder frame parameters. To make a better
determination, the vocoder looks at the next adjacent frame (voc frame 3) and
determines that the tone appears to continue as a '0' into this frame. The
decoder therefore decides that this is meant to be a baudot '0' tone in the
latter
15 half of vocoder frame 2.
As shown in the flow chart of FIG. 4, after it is determined if a low
activity signal is being received in tasks 402 and 404, the decoder
continuously
monitors and updates the received baudot tone boundaries in task 408.
Otherwise, any non-low activity signal is processed using traditional methods.
If an erred frame is received as detected in the physical layer, the frame is
assigned an indicator N and the vocoder examines the erred frame in task 410.
If a "reliable" decision concerning whether or not the frame is a baudot '0'
or '1'
can be made, such as when the frame parameters are quite distinct, then the
erred frame is modified to reflect the parameters of the decision. A reliable
decision is one that falls within a prescribed probability of obtaining the
original frame parameters. For purposes of this invention, the desired
probability would be in the range of 51% to certainty. If a modification is
made,
the method returns to task 402 and determines the next signal.
If a reliable decision cannot be made as shown in task 412, the vocoder
reviews the next adjacent frame N+1 or, alternatively, N-1. If this frame is
good
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
16
in task 416, the decision to modify the erred frame is made in task 418 based
upon the parameters contained within frame N+1, or alternatively, frame N-1.
If neither next adjacent frame is good, then a next best reliable decision is
made
based upon the parameters contained within next adjacent frame N+1 and
frame N's parameters are modified accordingly.
B. Encoder and Decoder Using Soft Bits
The decoder implementation in this embodiment of the invention is
similar to that disclosed above. However, to further reduce the error rate and
improve upon the accuracy and reliability of the signal decoded, the encoder
also takes advantage of the "soft bits."
When the vocoder encoder detects baudot tones are to be sent, the
encoder switches to a "baudot tone encoding mode." In this mode the encoder
decides whether the tone received for encoding is a '0' or a '1.' The encoder
then sends this decision to the decoder using a vocoder frame, but using
channel-coding redundancy to improve the decoder's chances of determining
the proper baudot tone. Even if the decoder receives a tone in an erred frame,
it
will have a greater likelihood of determining the correct tone sent because of
the forwarded decision.
In a simplified example, if the encoder detects a baudot '1' is to be
transmitted, it sends a series of is to the decoder. The series may be any
length,
but must be sufficient so that the decoder can operate as discussed above in
section A if necessary. This version of the invention replaces the standard
vocoder parameters with vocoder "signatures" that are better spaced apart
(i.e.,
easier to differentiate), thus making it easier to decide between two tones
even
when frames are in error.
C. Encoder and Decoder Not Using Soft Bits
This embodiment of the invention is another version of the methods
described in sections A and B, but the decoder is not given the soft bits from
any erred vocoder frames to process.
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
17
In this case, when the vocoder encoder detects a '1' or '0' baudot tone,
the vocoder also encodes the tone in a vocoder frame using redundancy, but the
encoding may be done across many vocoder frames. The '1's and '0's are
interleaved across a number of frames M so that if one frame is lost, the
decoder
can extract the necessary information from adjacent frames. The following
example shows interleaving taking place across four frames, but any number of
frames could be used. Assume the encoder detects the following baudot tones
for transmission:
11001
The encoder encodes the frames as follows for transmission to the decoder:
I --voc frame 1 -- I --voc frame 2-- I -- voc frame 3 - I --voc frame 4 -- I --
voc
frame 5
I -- baudot '1' -- I -- baudot '1' -- I -- baudot '0' -- I -- baudot '0' --- I
-- baudot
'1' -
xxxxxxxxxxxx1111xxxxxxxx11111111xxxx111111110000111111110000000011110
00000001111.
In this example, the vocoder frame parameters for each frame are segmented
where four bits represents the detected baudot tone in a particular vocoder
frame. The entire sixteen bits represents the detected baudot tones from the
last
four vocoder frames:
I baudot for frame N-3 I baudot for frame N-2 I baudot for frame N-1 I baudot
for frame N I .
To account for baudot tones not corresponding to vocoder frame boundaries,
the invention uses the following four-bit sequence where XXYY indicates that
the code in the current vocoder frame reflects a baudot code of 'X' followed
by a
baudot code of 'Y':
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
18
I --voc frame 1-- I --voc frame 2-- I --voc frame 3 -- I --voc frame 4 -- I -
voc
t '0' -- I -- baudot '1' -- I -- baudot '1' -- I -- baudot '0' -- I -- baudot
'0' -- I
xxxxxxxxxxxxx0011xxxxxxxx00111111xxxx00111111110000111111110000001111
D. Modified Vocoder with Increased Transmission Power
In another embodiment, the prohibitive cost of modifying the chipset in a
mobile station to accomplish the encoding and decoding processes discussed
above in sections A, B and C is advantageously minimized.
The encoding and decoding methods of sections A, B and C can be
advantageously accomplished through the use of a standard vocoder
connectively communicating with a signal enhancer, such as an estimator or a
repeater or any other device capable of performing a signal enhancing
function.
In addition, the methods of sections A, B, or C can be accomplished through
the
use of a modified vocoder. It would be apparent to one skilled in the art that
a
vocoder can be modified to further incorporate the functions of a signal
enhancer, i.e., to make an estimation of whether corrupted bits received by
the
communication unit were originally transmitted as '0's or '1's.
The system of FIG. 5 illustrates one embodiment of the invention. The
transmission from base station 550 to mobile station 510 is referred to as the
forward link and the transmission from the mobile station 510 to the base
station 550 is referred to as the reverse link. In the reverse link,
unmodified
vocoder 520 in mobile station 510 encodes the Baudot signal into standard
vocoder parameters and transmits the vocoder parameters to base station 550.
Modified vocoder 560 receives the encoded Baudot signal and enhances the
Baudot signal to restore corrupted bits. A clean version of the Baudot signal
is
then generated. In the forward link, unmodified vocoder 580 in base station
550
encodes the Baudot signal using standard vocoder parameters and transmits
the vocoder parameters to mobile station 510. Modified vocoder 530 receives
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
19
the encoded Baudot signal and enhances the Baudot signal to restore corrupted
bits. A clean version of the signal is then generated.
However, using signal enhancers or modified vocoders on both the
forward link and the reverse link to accomplish the decoding methods
discussed in sections A, B, and C can be prohibitively expensive to produce.
In
yet another embodiment of the invention, the frame error rate of the system
can
be reduced by using a signal enhancer in a base station on the reverse link,
and
by using power control techniques to adjust signal transmission power levels
from the base station to the mobile station on the forward link.
In the reverse link of the communication system of FIG. 6, base station
655 reconstructs the original Baudot signal contained within a vocoder encoded
frame according to the methods discussed in sections A, B and C. However, in
the forward link, base station 655 transmits the encoded Baudot signal using
power control techniques as discussed in U.S. Patent Application No.
09/114,344, entitled "METHOD AND APPARATUS FOR ESTABLISHING
TDD/TTY SERVICE OVER VOCODED CHANNELS," assigned to the assignee
of the present invention and incorporated by reference herein.
FIG. 7 is a flow diagram of a method for controlling transmission power
between the base station and the mobile station. The method begins in task
702,
where a Baudot signal is received by the mobile station 605. After the signal
is
received, vocoders used by mobile station 605 during processing of the signal
are locked into a full rate in task 704. In this embodiment, the transmission
power does not decrease from the transmission power used by the
telecommunications system for full rate transmissions in task 710. The power
level is typically set based upon a selected FER for the transmitted signal as
received at the mobile station 605. A desired FER is selected because when a
Baudot signal is being sent, the actual character error rate of the Baudot
signal is
about 9 to 10 times that of the FER. This selected FER range is between a 0.1%
and a 1.0% error rate, but may be less if necessary for preservation of the
quality
of the transmitted signal. Preferably, an FER of 0.2% is desirable for
transmitting Baudot signals. If the FER exceeds the selected range in task
712,
CA 02379973 2002-O1-16
WO 01/11840 PCT/US00/21441
mobile station 605 notifies base station 655 in conventional fashion during
task
706 that a system adjustment to reduce the FER is needed. Accordingly, an
adjustment is made in task 708. An adjustment typically includes increasing
the
transmission power for the full rate transmission, but may also include
5 adjusting other parameters known to reduce FER. If the FER is acceptable in
task 712, the signal transmission may continue in task 714 and dynamic
adjustments to the telecommunications system continue throughout the
transmission of the entire transmitted signal 710. Otherwise, when the
transmission of the Baudot signal ends, the vocoders are unlocked, and the
10 telecommunications system returns to normal operation.
OTHER EMBODIMENTS
While there have been shown what are presently considered to be
preferred embodiments of the invention, it will be apparent to those skilled
in
the art that various changes and modifications can be made without departing
15 from the scope of the invention as defined by the appended claims.
What is claimed is:
