Pulse or digital communications – Spread spectrum – Direct sequence
Reexamination Certificate
2000-05-31
2004-06-08
Chin, Stephen (Department: 2634)
Pulse or digital communications
Spread spectrum
Direct sequence
C375S143000, C375S150000, C375S152000, C375S317000, C327S060000, C327S068000, C455S214000, C455S231000
Reexamination Certificate
active
06748007
ABSTRACT:
TECHNICAL FIELD
The subject of this invention is a method of processing a pulse response with an adaptive threshold and a corresponding receiver. It finds general application in the processing of signals, every time a signal arrives in the form of pulses accompanied by overwritten replicas with noise. This may be the case in seismology, in radar or sonar detection or in the case of digital radio-communications with Direct Sequence Spread Spectrum (DSSS).
STATE OF THE PRIOR ART
The invention will be more specifically described in the context of spread spectrum digital radio-communications, although its scope is wider than this.
The technique of direct sequence spread spectrum consists schematically of multiplying a data symbol (for example one or more bits, by a pseudo-random code made up of a sequence of elements called chips). This operation has the effect of spreading the spectrum of the signal. On reception, the received signal is processed by correlation with a pseudo-random code identical to that used for transmission, which has the effect of reducing (“unspreading”) the spectrum. The “unspread” signal is then processed in order to recover the data symbol.
The technique of modulation by direct sequence spread spectrum is widely described in the specialist literature. The following books can be mentioned:
“CDMA Principles of Spread Spectrum Communication” by Andrew J. VITERBI, Addison Wesley Wireless Communications Series,
“Spread Spectrum Systems” by Robert C. DIXON, John WILEY and Sons,
“Spread Spectrum Communications” by Marvin K. SIMON, Jim K. OMURA, Robert A. SCHOLTZ and Barry K. LEVITT, Computer Science Press, 1983, vol. I.
FIG. 1
appended gives the simplified block diagram for a spread spectrum receiver in the case where a differential type modulation has been used on transmission.
In this Figure, one can see a receiver comprising an aerial
10
, a conversion oscillator
12
, a multiplier
14
, an amplifier
16
, a matched filter
18
, a delay line
20
, a multiplier
22
, an integrator
24
and a decision circuit
26
.
The operating principle of this receiver is as follows. The matched filter
18
carries out the correlation operation between the received signal and the spread sequence used. The principle of the phase differential modulation, which is sometimes chosen on transmission, leads to data being carried by the phase difference between the signals at the output from the matched filter
18
and at the output from the delay line
20
. This data is reconstructed by the multiplier
22
.
A correlation peak at the output from the multiplier
22
corresponds to each propagation path. The role of the integrator
24
is to take into account the data carried by each of the propagation paths. The propagation paths being statistically independent in a multiple path environment, with this particular reception technique, diversity processing is carried out, the order of which can be raised when the pulse response is complex. The decision circuit
26
enables one to reconstruct the transmitted data and, in addition, regenerate a clock signal used to command the various circuits.
Document FR-A-2 742 014 describes a digital embodiment of this receiver, which is illustrated in FIG.
2
. This receiver comprises two similar channels, one to process part I of the signal in phase with the carrier and the other to process part Q in quadrature with this same carrier.
Channel I comprises filtering means
50
(I) matched to the pseudo-random sequence used on transmission; these first means supplying samples I
k
. Channel I further comprises delay means
60
(I), the delay period being equal to the period Ts of the symbols; these means supplying samples I
k−1
.
Channel Q also comprises filtering means
50
(Q) matched to the pseudo-random sequence and supplying samples Q
k
. Channel Q additionally comprises delay means
60
(Q), the delay being Ts and supplying samples Q
k−1
.
The multiplier
70
calculates combinations of products of these samples and notably a signal designated below Dot(k) which is equal to I
k
I
k−1
−Q
k
Q
k−1
and a signal designated Cross(k) equal to Q
k
I
k−1
−I
k
Q
k−1
. The signals Dot(k) and Cross(k) allow one to calculate the product of a sample S
k
obtained at the instant k by the S
k−1
conjugated sample obtained at the instant t−Ts, where Ts is the duration of the symbols. This calculation is specific to the differential modulation.
The circuit in
FIG. 2
further comprises a programming means
72
. A decision circuit
90
finally supplies a clock signal H and the reconstructed symbol D (over one or more bits).
By way of an explanatory example,
FIG. 3
appended, shows the throughput of a Dot signal obtained by simulation, in the case where only a single propagation path exists between the transmitter and the receiver. Some of the peaks shown are positive, some negative, depending on the value of the binary data transmitted. The interval between two consecutive peaks corresponds to the duration Ts of one symbol.
In the case of Four Phase Shift Keying or DQPSK (Q for quaternary), the two signals Dot and Cross must be examined simultaneously in order to recover the transmitted data.
FIGS. 4 and 5
respectively show the throughput of the signals Dot and Cross still provided by simulation in the case of a single path.
In the case of several paths, the peaks illustrated in
FIGS. 3
to
5
would be doubles, triples, quadruples etc., for each symbol, the number of peaks detected being equal to the number of paths assumed by the radio wave between the transmitter and the receiver.
A simple integrator, like the integrator
24
in
FIG. 1
, integrated into the circuit
90
in
FIG. 2
, integrates all of the signals present, that is to say both the peaks (corresponding to true data) and the noise (that doesn't correspond to any data). The signal to noise ratio is therefore low.
Document FR-A-2 757 330 proposes a solution to improve this signal to noise ratio. It consists firstly of calculating the sum of the squares of the Dot(k) and Cross(k) signals, and then extracting the square root of this sum, a quantity that reflects the energy distribution of the various propagation paths, each peak having as its amplitude, the energy carried by the corresponding path. Hence the quantity E(k) is measured defined as:
E
(
k
)=[Dot(
k
)
2
+Cross(
k
)
2
]
½
Next an operation is carried out finding the mean of the energy E(k) over a few symbols, that is to say over a few values from row k. The number of symbols used for this estimation of the mean must correspond to a duration less than the coherence time of the channel, that is to say less than the time beyond which two separate waves from the same origin no longer interfere.
Using these means, designated E
moy
, the instantaneous signals Dot(k) and Cross(k) are weighted, for example, by multiplication of Dot(k) and Cross(k) by E
moy
. In this way, two new weighted signals are obtained, namely Dot(k)
moy
and Cross(k)
moy
. It is on these weighted signals reflecting the average of the energy of several symbols, that one then carries out the integration over a period Ts for the symbol and then the regeneration of the clock and the recovery of the data.
While this known technique gives satisfaction in certain regards, it nevertheless has the disadvantage of taking the noise into account even though it may be slight. Furthermore, the integration carried out on the pulses prevents the determination of the number of paths present.
SUMMARY OF THE INVENTION
The purpose of this invention is to remedy these disadvantages by allowing one to discard the noise and keep the identity of the paths.
To this end, the invention proposes that a threshold is determined starting from which the signal will be used in subsequent processing. Those skilled in the art would be inclined to choose a fixed threshold, determined once and for all. However, the value of the threshold would often be inappropriate. In effect, depending on the transmission conditions,
Bouvier des Noes Mathieu
Lattard Didier
Lequepeys Jean-Rene
Varreau Didier
Chin Stephen
Commissariat a l'Energie Atomique
Ha Dac V.
LandOfFree
Method of processing a pulse response with an adaptive... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of processing a pulse response with an adaptive..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of processing a pulse response with an adaptive... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3322910