The invention referred herein is about a coding and decoding method and device for reducing interferences in simultaneous signal transmission and multiple-user systems. This method enables the reduction of crosstalk or interference in multi-access communication systems based on any means of transmission or image capture through simultaneously coded impulse transmission.
This invention is developed in several fields due to the wide spectrum of use of this technology. By way of example but without limitation, we shall mention the use of this invention in the field of the audiovisual industry, especially, in the telecommunication industry. However, this technology is likewise useful in the military and civil field as for instance in radar or sonar device communications. Another example of the relevance and versatility of this technology is its use in medical diagnostic devices based on images, such as magnetic resonance imaging and ultrasound.
In most communication systems, the spectrum is limited and must be shared by a number of users.
There are several spectrum sharing systems: by means of frequency division (OFDM, DMT, etc.), Frequency Hopping (FH), Code Division Multiple Access (CDMA), Wavelength Division Multiplexing (WDM), and their combinations.
For the last years, several studies and researches have been focused on the possibility of reusing the spectrum or, at least, interfering as less as possible. All of them try to obtain the maximum spectrum efficiency and, therefore, the best use of the transmission channel while enabling the simultaneous transmission of signals without mutual interference.
One of the biggest problems is the interference among users in current and future mobile telephony systems. The system based on code division or CDMA is a system that is based on low cross-correlation properties of different sequences used by different subscribers. Due to the fact that said cross-correlation is not null, there is interference resulted from the simultaneous access of several users called MAI (Multi-Access Interference), which prevents from increasing the number of subscribers above the limit related to said interference.
On the other hand, low correlation properties are not met when there is a power difference transmitted among several subscribers; that is why the network should be able to control the power transmitted by each subscriber in order to ensure MAI interference is as less as possible. The evolution of mobile telephony set out by the European Consortium 3GPP tends to the use of various technologies, among which the multiple access is suggested by means of frequency division and using OFDM (Orthogonal Frequency Division Multiplexing).
Moreover, the effect of sharing the same frequency band among subscribers or services is particularly harmful to xDSL cable broadband access systems, where the Far End Crosstalk (FEXT) makes that, when the number of subscribers sharing the same cable of pairs increases, there is a decrease in the speed of data capable of transmitting for each subscriber at a specific distance. This effect may be significant and reduce the coverage for a specific service up to 50% for medium speeds in relation to 12 Mbps, and getting to 2500% in the case of 20 Mbps speeds, passing from 1 Km to 200 m of coverage radius.
The coding of different carriers by using complementary sequence sets has already been proposed in several studies such as the one published by Hsiao-Hwa Chen et al in [“A Multicarrier CDMA Architecture Based on Orthogonal Complementary Codes for New Generations of Wideband Wireless Communications,” IEEE Communications Magazine, October 2001, pp. 126-135].
Another approach to the same solution is proposed by Zao Ying et al in [“Complex Orthogonal Spreading Sequences Using Mutually Orthogonal Complementary Sets,” MILKON International conference, 2006. 22-24 May, pp. 622-625]. Complementary sequences are used in such a way that each sequence and carrier requires four phases. Both methods are identical, except for small modifications as regards sequences employed.
Finally, there is a reference to Shu-Ming Tseng's work [“Asynchronous Multicarrier DS-CDMA Using Mutually Orthogonal Complementary Sets of Sequences,” IEEE Trans. On Comm., Vol. 48, No. 1, January 2000, pp. 53-59], where the same procedures of modulation and demodulation are repeated in relation to the previous ones with slight modifications.
One of the inconveniences of all previous implementations is that the maximum spectrum efficiency is 1 bit/s/Hz. That efficiency proves to be very low when it is used in high-capacity communication systems like current ones, which vary from 3 bps/Hz in radio systems to 12 bits/s/Hz in xDSL.
Moreover, those technologies are exclusively designed for CDMA-based systems; that is why they cannot be used by any other communication system to reduce interference among subscribers. Besides, the output signal bandwidth is greater than the basic signal bandwidth. Thus, it is necessary to completely modify current systems' transmission and reception phases in order to integrate said technologies.
All this leads to the deduction that a technology capable of emitting information efficiently and reducing interference among subscribers or services using the same frequency band is needed, while respecting the bandwidth parameters and power transmitted, regardless of the way of modulating data in basic band, be it OFDM, CDMA, QAM, WDM or other variant of them.
This technology shall be used in any system that requires independence or orthogonalization of information channels with each other without modifying neither transmission spectrum nor power transmitted. Among evident applications, we shall mention the reduction of crosstalk among simultaneous subscribers of xDSL services, the increase in the number of subscribers per cell in mobile telephony systems, the increase of fiber optic cable capacity using different wavelengths or RADAR or SONAR signal orthogonalization, and the generation of medical images, among others.
Neither background art nor patents or models with similar features to the ones proclaimed herein are known.
The invention referred herein is based on using M complementary sequences sets. Complementary means that the sum of their autocorrelation results in the Kronecker's delta.
Besides, the value of M also matches the number of complementary sequences sets that are orthogonal among each other.
Orthogonal means that the sum of the cross-correlation of each complementary sequence set is zero.
These two properties are used in this patent to obtain the desired results. In the specific case of pairs (M=2) of orthogonal sequences, they are called Golay sequences, paying tribute to its discoverer.
The main property of sequences used in this invention is that they have an ideal autocorrelation feature, that is, it corresponds to a perfect Kronecker's delta without lateral lobules, and a mutual null cross-correlation among the families in an orthogonal sequence set.
For the proper implementation of the result, the system comprises two well-defined blocks:
a.—coding system in transmission, and
b.—decoding system in reception.
The appropriate employment of this process enables to totally cancel interferences.
First, we relate the elements comprising the drawings taking into account that identical references refer to identical elements.
-1-F(ω) consists of a band-pass filter bank adapted to the set of complementary sequences selected for said user.
-2-H(ω) corresponds to the channel between the transmitter and receiver point that can be modulated as the sum of N independent band-pass filters.
-3-F′(ω) consists of a band-pass filter bank adapted to the same set of:
-4-D1(ω), D2(ω) . . . DM(ω) correspond to the different data flow signals that are to be transmitted simultaneously.
-5-FA(ω), FB(ω) . . . FM(ω) correspond to the band-pass filter banks adapted to the families of orthogonal sequences used by each user in order to orthogonalize when receiving data from each of them in relation to the remaining flows.
-6-H(ω) Similar to -2-, corresponds to the means of transmission.
-7-F′A(ω), F′B(ω) . . . F′M(ω) correspond to the band-pass filter banks adapted to the families of orthogonal sequences used in the transmission by each user to orthogonalize when receiving data from each of them in relation to the remaining flows.
-8-Rx1(ω), Rx2(ω) . . . RxM(ω) correspond to signals retrieved by each user without mutual interference.
In order to better understand the invention, three sheets of drawings are attached, where the following is distinguished:
It presents the block diagram of a coding system for only one user.
It presents the block diagram for M users that are transmitted and received independently.
It presents the sketch of an xDSL communication system using the technology described in this patent.
The invention proclaimed here comprises two independent applications for the same united result.
On the one hand, a method is claimed.
And on the other, a device.
For the embodiment of said method, a device for signal coding and decoding is required.
The method uses sets of M complementary sequences. Complementary means that the sum of their autocorrelations results in a Kronecker's delta.
Besides, the M value also matches the number of complementary sequence sets that are orthogonal with each other.
Orthogonal means that the sum of the cross-correlation of each complementary sequence set is zero.
These two properties are used in this patent to obtain the desired results. In the specific case of pairs (M=2) of orthogonal sequences, they are called Golay sequences, paying tribute to its discoverer.
The device, as a communication system, is comprised of three main blocks:
An encoder -1- and -5-, a decoder -3- and -7-, and a channel -2- and -6-.
The encoder system is in charge of convolving the basic band signal to be transmitted with a set of complementary sequences. The decoder, on the other hand, is in charge of correlating signals received with the same set of complementary sequences used in the emission and of adding the results in order to obtain the original spectrum.
The main property of sequences used in this invention is that they have an ideal autocorrelation feature, that is, it corresponds to a perfect Kronecker's delta without lateral lobules, and a mutual null cross-correlation among the families in an orthogonal sequence set, complying with:
Where φii are the individual autocorrelations of each M complementary sequence selected with N-length, and Φ and Ωi are the response in frequency of autocorrelation and of complementary sequence i of the family Ω in the set of M-length orthogonal sequences in the bandwidth used, and * is the conjugated operator.
The generation of those sequences is performed based on the so-called basic kernel known up to date of 2, 10 and 26 bits (the rules of generation of complementary sequence families is discussed in the article titled “Complementary Sets of Sequences” by C.-C. Tseng and C. L. Liu, published in IEEE Trans. Inform. Theory, Vol. IT-18, No. 5, pp. 644-651, September 1972).
In order to understand the technology, it is convenient to observe the process block diagram (
Taking into account that the function of channel transference in bandwidth frequency B is:
H(ω)=H1(ω)+H2(ω)+ . . . +HN(ω) (5)
We will suppose that the bandwidth of each channel is B/N in order to facilitate the process.
The signal received through the channel will correspond to the convolution of the input signal with the channel response or, which is similar, to the product of their spectra:
Rx(ω)=D(ω)·H(ω)=D(ω)·[H1(ω)+H2(ω)+ . . . +HN(ω)] (6)
Where F1(ω), F2(ω) . . . FN(ω) are band-pass filters corresponding to frequency bands of channels 1, 2, . . . , N and unity gain convolved by complementary sequences in the following way:
F
1(ω)=Ω1(ω)
F
2(ω)=Ω2(ω)
F
N(ω)=ΩN(ω) (7)
Where Ωi is the element i of set Ω within the complementary set of sequences (A, B, C, D, . . . ) of N elements meeting property (4) among them, as it is explained in the article by Tseng mentioned above.
Based on the diagram of
Rx(ω)=D(ω)·[F1(ω)H1(ω)F′1(ω)+F2(ω)H2(ω)F′2(ω)+ . . . +FN(ω)HN(ω)F′N(ω)] (8)
For expressions (8) and (6) be equaled, all channel responses should be identical and equal to the unit. This process is called equalization and may be achieved through a variety of conventional processes.
Therefore, in basic band, we will suppose that channels have been previously equalized to this process, obtaining, finally, this expression:
Rx(ω)=D(ω)·[F1(ω)F′1(ω)+F2(ω)F′2(ω)+ . . . +FN(ω)F′N(ω)] (9)
Where F1(ω), F2(ω) . . . FN(ω) are band-pass filters corresponding to frequency bands of channels 1, 2, . . . , N and unity gain convolved by complementary sequences in the following way:
F
1(ω)=Ω*1(ω)
F
2(ω)=Ω*2(ω)
F
N(ω)=Ω*N(ω) (10)
Where * is the conjugated operator.
Replacing in (9) and applying the property of complementary set of sequences (4), it is proved that:
Rx(ω)=D(ω)·cte (11)
From this result, and based on
Rx(ω)=[D1(ω)+D2(ω)+ . . . +DM(ω)]·[H1(ω)+H2(ω)+ . . . +HN(ω)] (12)
In that way, all users are independent from each other. If a set of complementary sequences from a family of orthogonal sequences is assigned to each user, it will be proved that they are independent and that they can be retrieved without mutual interference. As regards clarity, it will be proved with a pair of users using an orthogonal set, among them A and M. In that way, equation (12), assuming channel equalization and replacing (7) and (10) in (9), and eliminating variable ω by simplicity, results in:
Rx=Rx
1
+Rx
2
=D
1
·[A
1
A*
1
+A
2
A*
2
+ . . . +A
N
A*
N
]+D
1
[A
1
B*
1
+A
2
B*
2
+ . . . +A
N
B*
N
]++D
1
·[B
1
A*
1
+B
2
A*
2
+ . . . +B
N
A*
N
]+D
2
[B
1
B*
1
+B
2
B*
2
+ . . . +B
N
B*
N] (13)
Due to the properties of the sets of families of orthogonal complementary sequences, cross terms of (13) are null and the resulting expression is as follows:
Rx=Rx
1
+Rx
2
=D
1
·[A
1
A*
1
+A
2
A*
2
+ . . . +A
N
A*
N
]+D
2
·[B
1
B*
1
+B
2
B*
2
+ . . . +B
N
B*
N]=cte·(D1+D2) (14)
It can be showed that the previous process generalized for N users can be expressed as follows:
That is to say that the sum of signals received is equivalent to the sum of data transmitted, multiplied by a constant and without mutual interference. This means that users are orthogonal and independent.
In another embodiment of the invention, each user's channel may be different, as it is the case of some radio systems, satellites, and RADAR or xDSL systems. In this case, the channel model for two users is the following:
Rx(ω)=D1(ω)·[H11(Ω)+H12(ω)+ . . . +H1N(ω)]+D2(ω)·[H21(ω)+H22(ω)+ . . . +H2N(ω)] (16)
Where D1 is the transmitted signal, D2 is the transmitted signal by the interfering source, H1 is the transference function of channel between the generation point of signal D1 and the receiver, and H2 is the transference function from the generation point of signal D2, or interfering user, and receiver 1.
In this case, where channels are not identical, it is necessary to independently equalize each channel H1, H2, . . . corresponding to each user and interfering for the orthogonalization property to be met; however, the property is still useful for applications mentioned in this document though its complexity is greater.
There are other cases where the transmission point of all users is the same, such as the downstream channel of a mobile telephony basic station towards subscribers, a satellite-Earth link, or xDSL channels. See
Rx(ω)=[cte1·D1(ω)+cte2·D2(ω)]·[H1(ω)+H2(ω)+ . . . +HN(ω)] (17)
Where H1=H2=H and cte1, cte2 are constants. Thus, (17) mainly matches the expression (12) and, therefore, all users are orthogonal among each other once channel H is equalized in the receiver.
It should be highlighted that the signal emitted D has been considered to have modulation, power and bandwidth remaining unaffected in the orthogonalization process and independent of it, which represents a great advantage in front of the above mentioned proposals.
Moreover, we should consider that in the case of xDSL communications, (see
In conclusion, it can be stated that the advantages of this technology are, on the one hand, the capacity of building independent and orthogonal channels in time for different users using the same band of frequencies and, on the other, the ability to maintain elevated spectrum efficiencies regardless of the process described. Therefore, the invention described herein constitutes a powerful system of orthogonalization of channels, which improves current technologies using complementary codes increasing spectrum efficiency in communication systems, or increasing the amount of information obtained in RADAR, SONAR, or medical imaging systems.
Number | Date | Country | Kind |
---|---|---|---|
P 200800188 | Jan 2008 | ES | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/ES08/00734 | 11/24/2008 | WO | 00 | 7/13/2010 |