|Publication number||USRE41931 E1|
|Application number||US 11/332,810|
|Publication date||Nov 16, 2010|
|Filing date||Jan 13, 2006|
|Priority date||Apr 2, 1999|
|Also published as||DE60008446D1, DE60008446T2, EP1041730A1, EP1041730B1, US6678338|
|Publication number||11332810, 332810, US RE41931 E1, US RE41931E1, US-E1-RE41931, USRE41931 E1, USRE41931E1|
|Inventors||Dominique Noguet, Jean-René Lequepeys, Didier Lattard, Norbert Daniele|
|Original Assignee||Commissariat A L'energie Atomique|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (10), Classifications (11), Legal Events (5)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The present invention relates to a receiver module and to a receiver formed from several cascaded modules.
The invention has a general application in digital communications and in particular in wireless local area networks (WLAN), wireless local subscriber loops (WLL), mobile telephony, electronic funds transfer, integrated home systems, communications in transportation vehicles, cable television and multimedia services on cabled networks, etc.
The invention relates to the spread spectrum technique. It is known that this technique consists of modulating a digital symbol by a pseudorandom sequence known to the user and the emission of said modulated symbol. Each sequence is formed from N elements known as chips, whose duration is the Nth of the duration of a symbol. This leads to a signal, whose spectrum is spread over a range N times wider than that of the original signal. On reception, a demodulation takes place by correlating the signal received with the sequence used on emission, which makes it possible to reconstitute the starting symbol.
This technique has numerous advantages:
However, this technique suffers from a disadvantage constituted by its limited spectral efficiency. This means the ratio between the binary data rate and the width of the occupied band. If each data symbol contains m bits, the binary data rate is equal to m times the symbol rate, i.e. mDs. With regards to the occupied band, it is equal to double the chip frequency, i.e. 2N times the symbol rate, i.e. 2NDs. Thus, finally, there is a spectral efficiency equal to the ratio
Consideration could be given to an increase in the spectral efficiency by decreasing N, but this would be in opposition to the qualities inherent in the spread and would in particular prejudice the immunity of transmissions to interference. Consideration could also be given to increasing the symbol rate, but the interference phenomenon between symbols would be aggravated.
Another solution consists of increasing m, the number of binary data per symbol, which leads to the use of complex modulations such as phase shift keying (PSK) with several phase states, which is a phase modulation (or coding) or the so-called “M-ary Orthogonal Keying” (MOK) or order M orthogonal modulation.
A description of these modulations appears in two general works:
Firstly with respect to phase modulation, it is pointed out that this is more usually a binary modulation or BPSK or quaternary modulation or QPSK. In the first case it is possible to encode symbols with one bit (m=1) and in the second case symbols with two bits (m=2).
These modulations are more usually implemented in differential form (DBPSK, DQPSK) ensuring a good robustness in difficult channels, because no phase recovery loop is necessary. This differential form is also very suitable for processing the multiplicity of propagation paths.
On reception, a differential demodulator carries out the multiplication between the signal to be demodulated and its version delayed by a symbol period. In the case of quaternary modulation use is made of two signal channels, one channel processing the component of the signal in phase with a carrier and another channel which processes the component in quadrature with the carrier.
In the case of MOK modulation, it constitutes a technique in which with each symbol to be emitted is associated a signal taken from among a group of orthogonal signals. These signals can be spread codes of a same family of orthogonal codes. In this case, the modulation also implements the spread. However, these signals may also not be perfectly orthogonal and in this case the performance characteristics are less satisfactory.
If a symbol is constituted by m bits, there are two m possible configurations for the symbols. The number M of available codes must therefore be at least equal to M, with M=2m If the length of the codes is N, it is known that it is possible to find N orthogonal codes.
Thus, we obtain M=N and the number of bits per symbol is consequently limited to log2N. A known MOK receiver is illustrated in the attached FIG. 1, where it is possible to see a bank of matched filters 10 1, 10 2, . . . , 10 M, followed by the same number of samplers 12 1, 12 2, . . . . , 12 M, circuits 14 1, 14 2, . . . , 14 M for determining the energy (or amplitude) of the sampled signal, a circuit 16 for determining the highest energy (or highest amplitude) signal and which delivers the number of the channel corresponding to said signal and finally a circuit 18 which, on the basis of the number of said channel, restores the corresponding code, i.e. the transmitted symbol S.
The MOK technique has a variant called MBOK (“M-ary Bi-Orthogonal Keying”) consisting of adding to the set of orthogonal signals used in a MOK modulation their opposites in order to constitute a set of 2M signals, which are obviously not all orthogonal to one another. Demodulation uses M correlators, each adapted to M orthogonal codes, but also requires sign recovery means.
If, for increasing the spectral efficiency, there is an-increase by one unit of the number m of bits in each symbol, the number M of available codes doubles, which multiplies by 2 by the number of channels of the receiver. Thus, the complexity increases much more rapidly than the spectral efficiency, so that this technique has certain limits.
MOK and MBOK modulations are used in certain digital communications systems, in conjunction with a coherent reception structure, which requires the knowledge of the phase of the carrier. The emission of a preamble, prior to the emission of the useful data, is a standard process permitting the estimation of said phase. However, in channels subject to fading and/or multiple path, the phase of the Carrier undergoes variations, which can be very fast and which the reception system must detect and compensate. This is generally obtained by the periodic emission of preambles, which then occupy the channel and lead to a reduction in the useful data rate. In accordance with this diagram, the durations of the preamble and the useful data packet must be less than the channel coherence time (time during which the channel is considered to be stationary). Moreover, there is an increase in the complexity of the reception structure.
Therefore the expert prefers to use non-coherent demodulation diagrams or diagrams which are differentially coherent, which do not require the knowledge of phase information. These techniques avoid the need for long preambles, for phase estimators and for phase derotators, at the cost of a slight sensitivity loss. Moreover, non-coherent demodulation very significantly simplifies the processing of the multiplicity of propagation paths, because each path has inter alia its own phase (and therefore would not require its own phase estimator in a coherent diagram).
French patent application 98 11564 filed on Sep. 16, 1998 by the present applicant proposes a mixed modulation-demodulation digital transmission process combining the MOK technique and the DPSK technique. According to this document, the following procedure is adopted:
The attached FIGS. 2 and 3 respectively illustrate an emitter and a receiver implementing this process.
FIG. 2 shows an emitter comprising:
the corresponding receiver is shown in FIG. 3 and comprises:
The data subgroups MMOK and mPKS are then collected for reconstituting the symbol S. In this technique, the number of bits transmitted per symbol is consequently: m=mMOK+mPSK.
As stated hereinbefore, the largest family of orthogonal codes of length L contains L codes (it is said that the family is cardinal N=L). However, as stated hereinbefore, the signals may not be perfectly orthogonal and in this case the performance characteristics are less satisfactory. In practice, the increase in the number of codes increases the complexity of the receiver to a very significant extent. This complexity problem imposes a limitation to the number of usable codes. Thus, full advantage is not taken of the increase in the spectral efficiency theoretically permitted by MOK modulation. As N increases this phenomenon becomes more critical and this is typical of the spread spectrum applications when it is wished to have a robust transmission system.
The aim of the present invention is to propose a solution to this problem. To this end, the invention proposes modifications to the receiver described hereinbefore in such a way that said receiver can constitute a receiver module (or elementary receiver), which can be cascaded (or connected in series) with other identical modules. This leads to the formation of a receiver constituted by several modules operating with a number of codes exceeding the number inherent in each module, but without any increase in the complexity of each module. In exemplified manner, a sequence length of L=32 is assumed, which corresponds to a code number N=32.
It is also assumed that there are two DPSK modulation phase states, i.e. mPSK=2 (case of QPSK modulation with four phase states).
For a receiver module use is made of a number Nc=8. The maximum number of bits transmitted in a symbol is:
The maximum number of bits accessible to a receiver module for one symbol is:
For this example, the number of bits transmitted per symbol as a function of the number of receiver modules is given in the following table, which makes it possible to compare the bit rates and the spectral efficiency.
Number of bits
Bit rate for a
Number of cascaded
60 MHz chip
a 2 Mbit/s link
The modular character of the receiver according to the invention offers a very great flexibility in the design of a receiver and makes it possible to obtain high bit rates without increasing the complexity of the circuits. In addition, the standard character of the basic module makes it possible to reduce costs and improve fabrication yields.
The receiver module according to the invention uses certain means of the receiver described relative to FIG. 3 and is characterized in that it is modified so as to be cascadable with other similar modules. To this end, the selection means also deliver, on a second output, the maximum energy (or amplitude) value. Moreover, the receiver module comprises supplementary inputs and outputs, with appropriate interconnections within the module, in order to permit cascading. With regards to the inputs, besides the first input receiving the signal to be processed, the module comprises:
The supplementary outputs comprise:
In addition, the module has the inputs and outputs necessary for the exchange of control signals, particularly for the mutual synchronization of the modules.
The present invention also relates to a receiver constituted by a plurality (at least 2) such receiver modules. Each receiver module operates with a group of M particular codes, the first, second and third inputs of a receiver module of rank or order i being connected to the first, second and third corresponding outputs of the receiver module of the directly lower rank or order (i−1). The final receiver module fulfills a particular function and is known as the master module, said master module receiving on its fourth input all the code numbers delivered by the fourth outputs of the (n−1) preceding receiver modules, all these numbers forming a global code number. This master module deduces from said global number the corresponding spread code and restores a first subgroup of (MMOK) data. The phase demodulation means of said master module receive the last switched signal and carry out demodulation in order to deliver a second group of mPSK data, said master module then reconstructing the transmitted global symbol. The master module also determines the signal or signals necessary for the synchronization of the other modules.
In such a receiver, the phase demodulation means of the (n−1) of the receiver modules preceding the master module are not used.
FIG. 1, already described, illustrates a MOK receiver.
FIG. 2, already described, is a diagram of a MOK-DPSK emitter.
FIG. 3, already described, is a diagram of a corresponding receiver.
FIG. 4 illustrates a receiver module according to the invention.
FIG. 5 shows a receiver constituted by several cascaded receiver modules.
The receiver module shown in FIG. 4 comprises means already shown in FIG. 3 and which carry the same references, namely the matched filters 40 1, 40 2, . . . , 40 M, the selection means 44, the switching means 45, the decoding means 46 and the demodulation means 58, 60. For simplification reasons, the samplers 42 1, 42 2, . . . , 42 M are not shown.
The module shown comprises four inputs E1, E2, E3 and E4 and four outputs S1, S2, S3 and S4. The input E1 is connected to the output S1 across a delay means 61. The input E2 is connected to the input of the selection means 44. The output 45s of the switching means 45 is connected to the output S3 across a delay means 63. The selection means 44 comprises a second output 44's, which delivers the energy (or amplitude) of the highest energy signal. This second output 44's is connected to the output S2.
The signals applied to the inputs of such a module are as follows:
The signals delivered by the outputs are as follows:
This receiver module functions in the following way.
The selection means 44 compare the energies of the M+1 signals, namely the energies of M output signals of M matched filters and the value of the energy applied to the second input E2 of the module and corresponding to the highest energy from the receiver module of the preceding rank (or zero if it is the first module). Two cases can be envisaged:
If the receiver module is the sole module (a zero signal being applied to the inputs E2 and E3), the demodulation means 58-60 function normally and the module delivers the reconstructed symbol mMOK+mPSK. If the receiver module is followed by other modules, said reconstruction is transferred to the final module (master module) and the demodulation means 58-60 are not used.
FIG. 5 illustrates a receiver formed from a plurality of n modules R1, . . . , Ri−1, Ri, . . . , Rn, which are cascaded. The inputs E1, E2, E3 of a module Ri of rank i are connected to the outputs S1, S2, S3 of the preceding module Ri−1 of rank i-1. The outputs S4 of each module are connected to the input E4 of the final module R by a connection 70. These outputs deliver the numbers of the channels and said numbers constitute a global number as from which the means 46 of the final module Rn restore the data subgroup
The demodulation means 58, 60 of said final module restore the data subgroup mDPSK. These two subgroups enable the master module Rn to reconstruct the symbol S.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5164964 *||Mar 11, 1991||Nov 17, 1992||Mitsubishi Denki Kabushiki Kaisha||Diversity circuit and frame phase (or sampling timing) estimation circuit using the diversity circuit|
|US5583884 *||Dec 16, 1994||Dec 10, 1996||Nec Corporation||Spread spectrum modulation and demodulation systems which accelerate data rate without increasing multilevel indexing of primary modulation|
|US5719852 *||May 29, 1996||Feb 17, 1998||Interdigital Technology Corporation||Spread spectrum CDMA subtractive interference canceler system|
|US5799035 *||Nov 12, 1996||Aug 25, 1998||Commissariat A L'energie Atomique||Digital circuit for differential receiver for direct sequence spread spectrum signals|
|US6094449 *||May 8, 1998||Jul 25, 2000||Nec Corporation||Spread spectrum communication synchronization acquisition decoding apparatus|
|US6259688 *||Mar 25, 1999||Jul 10, 2001||Interdigital Technology Corporation||Spread spectrum CDMA subtractive interference canceler system|
|US6335947 *||Oct 5, 1998||Jan 1, 2002||Commissariat A L'energie Atomique||Process for the processing of an information transmission signal by spread spectrum and the corresponding receiver|
|US6973144 *||Sep 12, 2000||Dec 6, 2005||Lucent Technologies Inc.||Apparatus and method for channel estimation used for link adaption with error feedback|
|US7133456 *||Jul 24, 2002||Nov 7, 2006||Kamilo Feher||Modulation and demodulation format selectable system|
|US20030058786 *||Oct 17, 2002||Mar 27, 2003||Kabushiki Kaisha Toshiba||Method and apparatus for receiving diversity signals for use in OFDM radio communication system|
|U.S. Classification||375/330, 375/152|
|International Classification||H04L27/22, H03D3/22, H04B1/707, H04J13/00|
|Cooperative Classification||H04B1/7093, H04B1/707, H04J13/004|
|European Classification||H04B1/707, H04B1/7093|
|Aug 28, 2006||AS||Assignment|
Owner name: COMMIAARIAT A L ENERGIE ATOMIQUE, FRANCE
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:NOGUET, DOMINIQUE;LEQUEPEYS, JEAN-RENE;LATTARD, DIDLER;AND OTHERS;REEL/FRAME:018189/0062
Effective date: 20000608
|Feb 2, 2007||AS||Assignment|
Owner name: COMMISSARIAT A L ENERGIE ATOMIQUE, FRANCE
Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE SPELLING OF THE RECEIVING PARTY. PREVIOUSLY RECORDED ON REEL 018189 FRAME 0062. ASSIGNOR(S) HEREBY CONFIRMS THE CORRECT SPELLING IS COMMISSARIAT A L ENERGIE ATOMIQUE.;ASSIGNORS:LEQUEPEYS, JEAN-RENE;NOGUET, DOMINIQUE;LATTARD, DIDIER;AND OTHERS;REEL/FRAME:018849/0699
Effective date: 20000608
|Jun 14, 2011||CC||Certificate of correction|
|Jun 22, 2011||FPAY||Fee payment|
Year of fee payment: 8
|Jun 24, 2015||FPAY||Fee payment|
Year of fee payment: 12