[0001] The present invention relates to method and apparatus for effecting separation and
substantial cancellation of interference at a receiver between a first and a second
received digital signal which use the same frequency spectrum and are received from
the same general direction.
[0002] In a domestic satellite communication system the coexistence of spot and area coverage
beams can be desirable. For example, a separate spot coverage beam can be used for
communication between the satellite and each high traffic ground station while an
area coverage beam can be used for communication between the satellite and a plurality
of low traffic ground stations under conditions where it might not be desirable to
interconnect the individual low traffic ground stations to a nearest high traffic
ground station for access to the satellite system. To avoid signal degradation and
permit separation of the overlapping spot coverage and area coverage beams especially
at each spot coverage receiving station, a typical prior art technique would be to
use separate bandwidths or polarizations, if possible, for the spot coverage beams
and the area coverage beam. Using separate bandwidths, however, results in inefficient
use of the frequency spectrum and different polarizations may not be available where
dual polarization beams are already used.
[0003] Various techniques have been devised to,suppress interference between two beams arriving
at a receiver from separate directions. In this regard see, for instance, U. S. Patents
2,520,184; 3,094,695; 3,369,235; and 3,987,444. Since the area and spot coverage beams
transmitted from a satellite arrive at each spot beam ground station from the same
direction, techniques for separating signals from different directions are not usable.
[0004] An alternative technique to enable reception of only one signal of a plurality of
signals concurrently received from a plurality of transmitters at an FM receiver would
be to modulate the carrier of each transmitter with a separate frequency to provide
a unique adress that is assigned to an associated receiver as disclosed, for example,
in U.S. Reissue Patent Re. 27,478. Such arrangement may be applicable to FM communication
systems but does not appear applicable to a digital communication system.
[0005] The problem remaining in the prior art is to provide a technique which permits two
digital signals using the same frequency spectrum and general transmission direction
to be simultaneously transmitted on one radio channel or overlapping spot and area
coverage beams with the ability for the signals to be separated at a receiving station
intercepting both signals.
[0006] The foregoing problem is solved, according to the invention, by the method characterized
by performing a detection process on an uncoded first information signal in a predetermined
frequency band in a first direction and on a coded second information signal having
a different informational content and a lower capacity than the first information
signal in the predetermined frequency band in the first direction for generating the
most likely digits representative of the received first information signals and decoded
second information signal, said code of the second information signal comprising a
forward error correcting code.
[0007] A receiver for practizing the method is characterized in that the receiver is disposed
in the path of a first uncoded digital information signal modulated to a predetermined
frequency spectrum and a second digital information signal having a different informational
content and a lower capacity than said first digital information, said second
signal being encoded with a forward error correcting code and modulated to the predetermined
frequency spectrum, said receiver comprising a detector capable of generating from
the combined received signal of the interferring uncoded first information signal
and coded second information signal the most likely digits representative of any combination
of the separated first information signal and decoded second information signal.
[0008] The present invention has been and will be described primarily in relationship to
a satellite communication system to enable the concurrent use of an area coverage
satellite radiated beam and a plurality of spot coverage satellite radiated beams
where all of the beams use the same frequency spectrum and the spot coverage beams
are received within the area encompassed by the area coverage beam. However, it will
be understood that such description is exemplary only and is for the purpose of exposition
and not for purposes of limitation. It will be readily appreciated that the inventive
concept described can be equally applicable to other radiated wave transmission systems
which comprise two or more beams which have different destinations but interfere with
each other at one or more of the destinations. Alternatively, the present invention
can be used to increase the capacity of a radio channel by 50 percent by simultaneously
transmitting an uncoded first digital signal and a coded second digital signal with
reduced capacity on each radio channel according to the concept to be described hereinafter
for the individual area and spot coverage beams.
[0009] In the drawings, in which like numerals represent like parts in the several views:
FIG. 1 diagrammatically illustrates a satellite communication system for providing
both an area coverage beam and a plurality of spot coverage beams between the satellite
and associated ground receiver stations;
FIG. 2 illustrates an arrangement according to the present invention to effect interference
cancellation between the area coverage beam and each of the spot coverage beams at
each of the receiver stations;
FIG. 3 depicts a typical prior art encoder for generating a constraint length = 3,rate=
1/2 binary convolutional code from an input stream of data;
FIG. 4 is a decoding diagram illustrating the allowable'State transitions and channel
symbols for both the convolutionally encoded area coverage beam signal and one interfering
uncoded spot coverage beam signal of FIG. 2;
FIG. 5 is a block diagram of an arrangement for implementing the joint maximum-likelihood
detector at a receiving station in accordance with the present invention.
[0010] in FIG. 1, a satellite communication system is illustrated wherein the present invention
is especially useful to permit the concurrent transmission from a satellite 10 of
both an area coverage beam 12 and a plurality of spot coverage beams of which, for
example, three beams 14a, 14b and 14c are shown with all beams being able to use the
same frequency spectrum. Spot coverage beams 14a, 14b and 14c are shown radiating
from antennae 15a, 15b, and 15c, respectively, and directed at respective ground areas
16a, 16b and 16c which include, for example, high traffic ground r stations 17a, 17b
and 17c, respectively. Area coverage beam 12 is shown radiating from an antenna 13
and directed at a ground area 18 which includes both the ground areas 16a, 16b and
16c and a plurality of low traffic ground stations of which, for example, four stations
19a-19d are shown. In the satellite communication system of FIG. 1, each of the high
traffic ground stations 17a-17c communicates with satellite 10 via
a separate spot beam 14a-14c, respectively, while the low traffic ground stations
19a-19d communicate with satellite 10 via area coverage beam 12 using any suitable
technique to assure that a particular message will be processed by only the appropriate
one of stations 19a-19d. Such arrangement permits low traffic ground
stations 19a-19d to communicate with satellite 10 under conditions where it is not
advantageous to connect a low traffic ground station 19 to a nearby one of high traffic
ground stations 17a-17c.
[0011] It can be seen from FIG. 1 that when area coverage beam 12 and spot coverage beams
14a-14c are transmitted concurrently and use the same frequency spectrum, each of
ground stations 17a-17c will receive both the associated one of spot coverage beams
14a-14c and area coverage beam 12 since these beams emmanate from approximately the
same point. Under such conditions the use of prior art arrangements such as, for example,
side lobe suppression arrangements to select a wave received from a particular direction
over waves received from other directions is not feasible.
[0012] The concurrent transmission of area coverage beam 12 and a plurality of spot coverage
beams 14a-14c which use the same frequency spectrum without interference can be effected
in accordance with the present invention by the typical arrangement shown in FIG.
2. There, a separate source of data 20a-20c generates the digital signals destined
to be transmitted via spot coverage beams 14a-14c, respectively. The digital data
signals generated by each of data sources 20a-20c are modulated to the desired frequency
spectrum for transmission in separate modulators 21a-21c, respectively. The outputs
from modulators 21a-21c are amplified in power amplifiers 22a-22c, respectively, prior
to being applied to the respective antennae 15a-15c for transmission via spot coverage
beams 14a-14c, respectively. The digital signals to be transmitted via area coverage
beam 12 are similarly generated by a data source 20d but at a reduced data rate which
is, for example, approximately one-half the rate of sources 20a-20c. These latter
signals are, however, first encoded in channel encoder 23 using a forward error correcting
code such as, for example, a block or convolutional code prior to being sequentially
modulated in modulator 21d, amplified by power amplifier 22d and transmitted by antenna
13 in area coverage beam 12. In this manner an area coverage beam of lower capacity
is provided, and redundancy is added using coding at the transmitter such that the
transmitted area coverage beam 12 occupies the entire spectral band used by the spot
coverage beams 14a-14c. Coding provides the advantages of
(1) reducing the area coverage beam transmitter power required to achieve some given
bit error rate which is highly desirable for satellite communication, and
(2) reduces the mutual interference between the area and spot beam signals.
[0013] It is to be understood that data sources 20a-20d, modulators 21a-21d and power amplifiers
22a-22d can comprise any suitable means capable of providing the function described
hereinabove. Similarly channel encoder 23 can comprise any suitable means for encoding
the digital data signals supplied by data source 20d into a forward error correcting
code. For example, where the code used is a convolutional code, an encoder of any
desired constraint length and code rate may be used.
[0014] A typical prior art arrangement for encoder 23 to produce a convolutionally encoded
area coverage signal having a simple constraint length K = 3, rate = 1./2 binary convolutional
code is illustrated in FIG. 3. It is to be understood that such description is exemplary
only and is for purposes of exposition and not for purposes of limitation. It will
be readily appreciated that other constraint lengths and rates or codes and their
implementations are equally applicable to permit channel separation at the receivers.
In FIG. 3, digital data signals for the area coverage beam are generated in data source
20d at the rate of one bit every T seconds for transmission over line 24 to encoder
23. At encoder 23 the received data signals are shifted into a three-bit shift register
25, or any other suitable means, at the rate of one bit every T seconds. A first modulo-2
adder 26 operates on the information stored in all three bits in register 25 to produce
a resultant first binary bit on transmission line 27 while, simultaneously, a second
modulo-2 adder 28 operates on the first and third bits in register 25 to produce a
resultant second binary bit on transmission line 29. During each T second, a commutator
30 first selects the signal on first transmission line 27 and then the signal on second
transmission line 29 and thereby transmits two binary digits over line 31 to modulator
21d for each data bit shifted into register 25. Therefore, for each data bit from
data source 20d the encoder 23 generates two data bits at its output and the encoded
data is then modulated and amplified prior to transmission via area coverage beam
12 to the ground stations disposed within ground area 18.
[0015] In the present arrangement, data sources 20a-20c generate two bits of data every
T seconds for transmission via spot coverage beams 14a-14c, respectively, while data
source 20d, generating one data bit every T seconds, in combination with encoder 23,
which provides redundancy and generates two data bits for every data bit from source
20d, also generates two bits of data every T seconds for transmission via area coverage
beam 12. Therefore, the data rate of all antenna radiated beams is the same with area
coverage beam 12 having a lower capacity than each of spot coverage beams 14a-14c.
[0016] Separation of interfering uncoded spot beam and coded area beam signals is achieved
in accordance with the present invention by providing a suitable detector at each
of spot beam ground stations 17a-17c and area beam ground
* stations 19a-19d. At each spot beam ground station 17a-17c, the receiver performs
a suitable detection of the spot beam signal received by that ground station plus
the received area beam signal, as will be described hereinafter. After the signals
are separated the information content of the interfering area beam is discarded. At
each area beam ground station 19a-19d which experiences interference from a spot beam
signal, a suitable detection of the desired area beam signal plus the unwanted interfering
spot beam signal is again performed and the information content of the interfering
spot beam signal is discarded after separation.
[0017] It is to be understood that a suitable detection process can comprise any process
which will enable the separation of the two digital signals and the decoding of the
forward error correcting coded signal. For example, where the two digital signals
have different signal strengths at the receiver, separation may be achieved using
a threshold detecting process. Alternatively, where the two received digital signals
have approximately the same signal strength, a maximum-likelihood detection process
may be performed. The type of detection process employed, however, will depend primarily
on the amount of signal degradation which can be tolerated since each of the known
detection processes would produce a certain amount of degradation in separating and
decoding the two signals described hereinbefore.
[0018] Although the above detection methods can be used the preferred method of separating
and decoding a simultaneously received uncoded first digital signal and a forward
error correcting encoded second digital signal with minimal degradation is accomplished
using the novel technique of joint maximum-likelihood detection as will be described
hereinafter.
[0019] Convolutional decoders and maximum-likelihood detection systems are well known in
the art. In this regard see, for example, U. S. Patents 3,789,360 and 3,815,028; and
U. S. Patent Application Serial No. 788,887. In the convolutional decoding process
described in the above-mentioned references, data is not decoded as soon as it is
received from the channel. Instead, a sequence of data, having a predetermined decoding
depth, following the digit to be decoded is first collected. Then, by computing what
are known as path metrics, a limited number of possible messages are selected, each
extending throughout the decoding depth far beyond the digit presently to be decoded,
with one such survivor sequence ending in each of the data states. A correlation between
each survivor sequence and the data actually received is computed for the entire decoding
depth under consideration. The highest correlated of the survivor sequences is then
selected to be the sole survivor sequence. The earliest received digit or digits within
the decoding depth is then permanently decoded under the temporary assumption that
the sole survivor sequence is the correct sequence.
[0020] More particularly, decoding is accomplished by forming the log-likelihood function
which hereinafter will be referred to as the path metric. Two samples are taken every
T seconds and the path metric is formed for each possible source sequence, and that
sequence for which the metric is largest is selected as the best estimate to the true
transmitted sequence.
[0021] As was shown in the prior art references cited hereinbefore, metric calculations
are an application of dynamic programming techniques and that maximum-likelihood decoding
can be performed without actually finding the path metric for each sequence. The procedure
for decoding the convolutionally encoded area beam signal is illustrated by the State
diagram of FIG. 4. The State is defined as the contents of the first two stages of
shift register 25, which changes at a T-second rate. When considering State 00, at
time to + T, this State can be reached from either of States 00 or 01, both transitions
corresponding to a data bit 0 having entered the coder. If the partial metrics corresponding
to each of the two merging paths are known up to and including time t
0 = T, then since the two paths have merged, the most likely path.leaving State 00
for t > t
0 + T must contain, as a subset, the path with the greatest metric up to that point
since future samples cannot affect past metrics.
[0022] At each of the spot beam ground stations 17a-17c and each of area beam ground stations
19a-19d having interference from a spot beam signal, the input signal to the maximum-likelihood
detector comprises two coded area beam channel symbols and two uncoded spot beam channel
symbols every T seconds which interfere with each other. Therefore, in the State diagram
of FIG. 4, four most-likely paths actually exist for each of the single paths shown
for the transitions between States. More particularly, as shown for the transition
from State 00 to State 00, the first two symbols for each of the four paths denote
the source coding for this particular transition, which is common to each of the possible
paths, while the last two symbols denote the four possible data symbols which may
exist for the first and second spot beam symbols received during each T seconds. It
is to be understood that each of the other transitions between States similarly comprises
four possible paths with corresponding symbols to denote the possible received symbols.
[0023] FIG. 5 is a block diagram of a typical arrangement for a novel joint maximum-likelihood
detector capable of decoding a constraint length = 3, rate = 1/2 binary convolutional
coded area beam 12 and uncoded spot beam 14 and providing a binary output every T
seconds indicating both the two most likely information digits received via the interfering
spot beam 14 and generated by the associated data source 20, and the decoded most
likely information digit received in coded form via interfering area beam 12 and generated
by data source 20d. It is to be understood that the arrangement of FIG. 5 is exemplary
only and is for purposes of exposition and not for purposes of limitation. It will
be readily appreciated that the inventive concepts described are equally applicable
to decode an interfering uncoded first beam and convolutionally coded second beam
having different constraint lengths and rates or nonbinary or multilevel alphabets,
transmissions, and the like, after the appropriate modification is made as will be
easily determined by one skilled in the art once the constraint length and rate is
known.
[0024] As shown in FIG. 5, the present joint maximum-likelihood detector is segmented into
four States, each State corresponding to a different one of the possible combinations
of one's and zero's in the first two stages of register 25 in encoder 23. A separate
sample of the received waveform at each ground station is taken every T/2 seconds,
and every T seconds the two samples which may be in digital or analog form, are made
available at input 40 of the present detector, each sample comprising elements of
the interfering area beam and spot beam signals. During each clock cycle T, the detector
recursively computes in processors 41a-41d the path metric of the most likely path,
of the eight paths, leading to each State. This computation is in the form:
where
An = the new path metric
An-1 =.the most likely path metric leading into the state from which a transition is made
r2n-2' r2n-1 = the two samples of the received process received during nth clock cycle
E1 = the signal energy/channel bit of the spot beam signal
E2 = the signal energy/channel bit of the area beam signal
ξ2n-2' ξ2n-1 = the assumed spot beam channel digits corresponding to a transition
Y2n-2' Y2n-1 = the assumed area beam channel digits corresponding to a transition.
[0025] As was stated hereinbefore, there are a total of eight paths leading to each of the
four States. For each State the detector computes the eight path metrics, finds the
largest one of the eight path metrics, saves the largest path metric, and stores the
path corresponding to the largest metric. This process will now be described in greater
detail for processing the path metrics for State 00, and it is understood that a corresponding
process is concurrently performed for processing the path metrics for the other States
01, 10 and 11.
[0026] In FIG. 5, the largest path metric for States 00, 01, 10 and 11 computed in the previous
T second cycle is stored in storage devices 42a-42d, respectively, and have the respective
designations M
1 - M
4 . The outputs from storage devices 42a-42d are normalized in normalization means
43 by, for example, arbitrarily setting one of the four old metrics, M
1 - M
4 , equal to zero after first having effectively subtracted its value from the remaining
three metrics. These normalized old path metrics are designated M
nl - M
n4. This step prevents the successive path metrics from growing linearly with time.
[0027] From FIG. 4, it can be seen that the transition into State 00, at time to + T, can
only be effected from prior States 00 and 01. Therefore, in processor 41a of FIG.
5, the old path metrics M
nl associated with State 00 is used together with the appropriate value of the two samples
available at input 40 in correlator 44a to compute the path metrics for each of the
four possible paths between State 00 at time to and State 00 at time to + T in FIG.
4. These four path metrics are indicated by the symbols M'
11 - M1
4 at the output of correlator 44a. Similarly the old path metric M
n2 associated with State 01 is used together with the appropriate value of the two samples
available at input 40 in correlator 44b to compute the path metrics for each of the
four possible paths between State 01 at time to and State 00 at time to + T. These
four path metrics are indicated by the symbols
-
at the output of correlator 44b. The eight path metrics computed in correlators 44a
and 44b are compared in comparator 45 and the largest one of the eight metrics is
determined. The comparator 45 is strobed by a system clock 46 via a signal on lead
47 to provide the result of comparison at the appropriate sampling instance once every
T seconds.
[0028] The value of the largest path metric for State 00 is transmitted from comparator-45
to storage means '42a via lead 48 where it is stored for use during the next processing
cycle T. The one of eight paths leading into a State having the largest value also
indicates the most likely digital value for both the decoded area beam signal and
the two sequential uncoded spot beam signals generated during a prescribed T second
period by the associated data sources 20 at the satellite 10. For example, if comparator
45 determined that the largest path metric corresponded to the uppermost path between
State 00 and State 00 in FIG. 4, then the most likely value for the bit generated
by data source 20d at satellite 10 during the corresponding T second interval would
be a zero while the most likely value for the first and second sequential bits received
via the interfering spot beam 14 during that same T second interval would be a 0,
0, respectively.
[0029] In FIG. 5, the decoded binary value of the most likely bit received via interfering
area beam 12 for State 00 is shown as being stored in a shift register 50a or other
suitable means, while the binary values for the most likely first and second sequential
bits for State 00 received via interfering spot beam 14 are stored in shift registers
51a and 52a, respectively, or any other suitable means. The outputs from comparators
45 in processors 41b-41d similarly load registers 50b-50d, 51b-51d and 52b-52d for
the most likely binary value for each decoded area beam bit and the first and second
sequentially received spot beam bits, respectively, for the respective States 01,
10 and 11. Each of shift registers 50a-50d, 51a-51d and 52a-52d have a path memory
length which preferably is about 4-5 equivalent constraint lengths, implying that,
with high probability, all surviving paths have a common prefix. Thus, the final state
of any one of registers 50a-50d may be selected as the decoded most likely information
digits for the received interfering area beam signal. Similarly the final stage of
any one of registers 51a-51d and 52a-52d may be selected as the most likely information
digits for the first and second sequential digits, respectively, received via the
interfering spot beam signal. Alternatively, the last stage of each group of registers
50a-50d, 51a-51d and registers 52a-52d can be used as a separate input to a separate
well-known majority logic gate associated with a particular group of registers which
functions to choose the output value indicated by the majority of the final stages
of the associated group, and in the event of a tie to output a 0 or a 1. A second
alternative would be to select the final stage of the register within each group of
registers indicating maximum likelihood.
[0030] Where the ground station performing the described joint maximum-likelihood detection
process is, for example, a spot beam ground receiving station, only the most likely
spot beam digits in registers 51a-51d and 52a-52d will be of interest for further
processing and, therefore, registers 50a-50d for storing the decoded most likely area
beam digits can be eliminated. Similarly at each area beam ground receiving station
only registers 50a-50d need be supplied. However, as was stated previously, when the
present invention is applied to increasing the capacity of a radio channel by transmitting
a first uncoded signal and a second coded signal of lower capacity on the same channel
with each signal using the same frequency spectrum, such combined signals will usually
be destined for the same receiver and, therefore, registers 50a-50d will be required
for generating the decoded most likely second signal digits along with registers 51a-51d
and 52a-52d for generating the most likely first signal digits at the receiver.
1. A method of separating and substantially reducing interference at a receiver between
a first and a second digital received signal which use the sane frequency spectrum
and are received from approximately the same direction
characterized by:
performing a detection process on an uncoded first information signal in a predetermined
frequency band in a first direction and on a coded second information signal having
a different informational content and a lower capacity than the first information
signal in the predetermined frequency band in the first direction for generating the
most likely digits representative of the received first information signals and decoded
second information signal, said code of the second information signal comprising a
forward error correcting code.
2. The method according to claim 1
CHARACTERIZED IN THAT
said first information signal is in a radiated beam covering a first receiving area
and said second information signal is in a second radiated beam covering a second
receiving area which is at least partially overlapped by said first receiving area.
3. The method according to claim 1
CHARACTERIZED IN THAT
said first information signal and said second information signal are in the same radio
channel received by the receiver.
4. The method according to claim 1
CHARACTERIZED IN THAT
.said forward error correcting code is a convolutional code and, in performing said
detection process,
(1) computing likelihood functions for all possible transitions into each state of
the convolutional code by using a first and a second sample of the received signal,
which includes elements of the interfering first and second information signals, and
the most recently computed most probable likelihood function for each possible state
of the convolutional code to generate signals representative of the computed likelihood
function,
(2) comparing the signals representative of the computed likelihood functions associated
with each of said states as generated in step (1) for determining the most probable
transition into each state of the convolutional code and generating a signal indicative
of said most probable transition into each state; and
(3) in response to the signal generated in step (2), concurrently storing (a) the
value of said signal for subsequent use in reiteration of the step of performing the
detection process and (b) the most likely data sequence into each state over a predetermined
length for any desired combination of the interfering first received digital signal
and ,the decoded second received digital signal for producing an output stream of
digital data corresponding to the most likely 'estimate of the desired digital signal.
5. A receiver for practizing the method of claim 1,
CHARACTERIZED in that
the receiver is disposed in the path of a first uncoded digital information signal
(14a) modulated to a predetermined frequency spectrum and a second digital information
signal (12) having a different informational content and a lower capacity than said
first digital information, said second signal being encoded with a forward error correcting
code and modulated to the predetermined frequency spectrum, said receiver comprising
a detector (Fig. 5) capable of generating from the combined received signal of the
interfering uncoded first information signal and coded second information signal the
most likely digits representative of any combination of the separated first information
signal and decoded second information signal.
6. A receiver according to claim 5
CHARACTERIZED IN THAT
said forward error correcting code is a convolutional code having a predetermined
plurality of possible states and an arbitrary code rate; and said detector is a joint
maximum-likelihood detector comprising
first means (44a, 44b) capable of computing likelihood functions for all possible
transitions into each of the plurality of possible states of the convolutional code
in combination with all possible values of the interfering uncoded first information
signal, said first means using both selective values of sequential first and second
samples (40) of the received waveform, which comprise elements of both the uncoded
first and coded second digital information signals, and signals (MN1-MN4) representative of the most recently computed most probable states of the convolutional
code and the associated most likely possible value for the received first digital
information signal for computing the likelihood functions;
second means (45) connected to the output of said first means for comparimg the signals
representative of the likelihood functions associated with each state of the convolutional
code and the associated possible values for the uncoded interfering first digital
information signal and generating an output signal indicating the most probable transition
into each state of said plurality of possible states and the most probable value for
the first digital information signal;
third means (42a-42d) coupled between said second and first means capable of temporarily
storing the most recently computed voltage signal generated by said first means representative
of the most probable likelihood function for each state and associated most probable
value of the first digital information signal for subsequent use by said first means
for computing the likelihood functions for each possible transition during the next
time'interval; and
fourth means (50a-50d; 5la-51d; 52a-52d) responsive to the output of said second means
for storing any combination of the most likely data sequence into each state of the
convolutional code (50a-50d) and the most likely data sequence for each state for
the first digital information signal over a predetermined length and for producing
at an output terminal thereof a stream of data corresponding to the data for the desired
uncoded first digital information signal (51a-51d) and decoded second digital information
signal (52a-52d).