Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L25/00—Baseband systems
  • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
  • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
  • H04L25/03891—Spatial equalizers
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B7/00—Radio transmission systems, i.e. using radiation field
  • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
  • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
  • H04B7/08—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station
  • H04B7/0837—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the receiving station using pre-detection combining
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B7/00—Radio transmission systems, i.e. using radiation field
  • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
  • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
  • H04B7/0413—MIMO systems
  • H04B7/0452—Multi-user MIMO systems
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L25/00—Baseband systems
  • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
  • H04L25/0202—Channel estimation
  • H04L25/0224—Channel estimation using sounding signals

Definitions

• TITLE Process for processing radio frequency signals received on R antennas, reception process, decoding process, computer program and corresponding system.
• the field of the invention is that of telecommunications.
• the invention relates to communications in the uplink, that is to say in the direction of the mobile terminals (or UE for “User Equipment”) towards a base station (or eNodeB, gNodeB, etc.).
• the invention proposes a new distribution of the functionalities implemented by a radio unit and by a baseband processing unit, for the decoding of radio frequency signals received on a plurality of antennas of a base station.
• the proposed solution applies in particular, but not exclusively, in the context of 5G NR (“New Radio” or “New Radio”) mobile networks.
• a radiofrequency signal received on an antenna undergoes analog processing, analog-to-digital conversion, then digital processing.
• the digital processing can be performed by a baseband processing unit, also called “Base Band Unit” (BBU) or “Distributed Unit” (DU) in English.
• BBU Base Band Unit
• DU Distributed Unit
• the active part of the analog processing can be performed by a “radio unit”, also called “Radio Unit” (RU) or “Remote Radio Head” (RRH) in English. It is recalled for this purpose that within the analog processing part, a passive part can be distinguished, comprising in particular the radiating antenna elements, and an active part, comprising in particular the filters, amplifiers, analog/digital converters, etc.
• RU Radio Unit
• RRH Remote Radio Head
• the base station BTS (“Base Transceiver station”) was connected to the passive elements of the antenna via a coaxial cable, via a limited number of antenna ports (maximum 4).
• the disadvantage of this architecture is the loss of power of the radio frequency signal between the antenna ports and the base station. This also limits the acceptable distance between the BTS and the passive antennas.
• the centralized RAN (Radio Access Network) architecture based on geographical separation of baseband (DU) computing capacities for digital processing, and radio transmitters (RU) for active analog processing, has developed.
• This type of architecture provides both functional benefits, thanks to better coordination between cells at the level of centralized units, and in terms of cost, by pooling the computing capacities of the different cells in common servers.
• several RUs can communicate with a DU.
• the interface between the DU and the RU is called “FrontHaul” in English, and can make it possible to deport the RU up to a maximum distance of 20km to centralize the DUs.
• TXRUs Transceiver Unit
• the evolution of base stations has consisted in bringing the TXRUs closer to the antenna, or integrated into the antenna in an RU radio unit.
• the baseband processing entity DU is thus connected to the RU by an optical fiber, carrying a digital signal, thus limiting the propagation losses associated with the use of a coaxial cable.
• the number of TXRUs has increased significantly over time and can now reach, for 5G, the value of 64 (massive MIMO).
• Open Fronthaul The xRAN Fronthaul Working Group, and more recently the O-RAN Standards Alliance, have supported the full specification of a single, open and interoperable interface between different vendors of RUs and DUs (“Open Fronthaul”).
• the 7.2x split is an adaptation of the 7.2 split specified in 3GPP and which makes it possible to reduce the complexity of the RU by moving processing functions up to the DU level.
• the DU includes the RLC/MAC/PHY-high layers
• the RU includes the layer called PHY-low.
• the PHY-low layer implemented in an RU includes, in addition to the active analog part of the antennas, certain baseband processing (close to the analog part) such as FFT/IFFT or digital beamforming. " in English. The resulting increase in bandwidth on the “open fronthaul” interface can be compensated by compression mechanisms referenced by the O-RAN Alliance.
• FIG. 2 illustrates more precisely the functionalities implemented by the RU 21 and by the DU 22, in the up direction, for the 7.2x split.
• the 7.2x split consists in moving the channel estimation 221 and RE-demapping 222 functionalities (extraction and separation of the resource elements - or "Resource Element (RE) - carrying the data and carrying the reference signals, in particular the DMRS (DeModulation Reference Signal)) in ORAN “Distributed Unit” (O-DU) 22.
• the 7.2x split includes a feature, implemented by O-RU 21, called port reduction 212 (“port reduction”) enabling the number of streams to be transmitted to the O-DU to be reduced.
• port reduction 212 port reduction
• the ORAN “Radio Unit” (O-RU) 21 would transmit a number of streams IQ equal to the number of reception branches R to the O-DU 22, whereas the number of spatial layers v to be detected is often much lower.
• these 8 streams can be distributed between a single PUSCH (SU-MIMO) or several PUSCHs (MU-MIMO) occupying the same time-frequency resource, each PUSCH i being transmitted from a different terminal (UE) and able to carry v i spatial layers with
• this pre-coding 212 (performed by the O-RU) cannot rely on the DMRS-based channel estimation 221 (performed by the O-DU), since this would require sending all of the R flow to the O-DU 22 (DMRS signals being carried by each PUSCH channel), which is contradictory with the aim of reducing the number of ports.
• the port reduction must be based on other reference signals, for example, the SRS transmitted in the up direction with a relatively high periodicity (of the order of 40 ms).
• the O-DU 22 estimates the channel based on SRS signals and sends it back to the O-RU (beam creation option called “channel information based beamforming” in the O-RAN standard ) or sends the “port reduction” pre-coding coefficients directly to the O-RU (beam creation option called “weights based beamforming” in the O-RAN standard).
• An advantage of the 7.2x split is to be able to co-locate, in the DU 22, the equalization 223 and decoding 224 functionalities, which allows the implementation of advanced receivers with interference subtraction involving a decoding feedback loop .
• a disadvantage of the 7.2x split is having pre-coding coefficients for port 212 reduction less up-to-date than if this pre-coding was based on DMRS channel estimation, since DMRS are part of the transmission of PUSCHs (they give an instant picture of the channel and the interference). A major difficulty therefore appears for the 7.2x split for the reception of the PUSCH(s) in the up direction.
• the invention proposes a solution which does not have all the drawbacks of the prior art, in the form of a method for processing radio frequency signals received on R antennas, with R 3 2, implementing a radio unit communicating with a baseband processing unit.
• such a method comprises, implemented by the radio unit:
• Such a method further comprises, implemented by the baseband processing unit:
• the proposed solution is therefore based on a new distribution of functionalities between a radio unit (located as close as possible to the antenna structure of a base station) and a baseband processing unit (located at the foot of the base station, or in a data center close to the base station, for example within a radius of 20 km).
• a first channel and covariance estimation of the noise plus interference is implemented by the radio unit. It can therefore be implemented from DMRS-based reference signals, which allows a more precise estimation of the transmission channel, and improves the quality of the projection. In particular, it offers an interesting solution for the reception of the PUSCH channel(s) in the uplink.
• a second estimation of the transmission channel of the radiofrequency signals, after projection, is moreover implemented by the baseband processing unit.
• a second estimate of the covariance of the noise plus interference can optionally be implemented. Such an estimate of the covariance of noise plus interference is not necessary if projection is followed by whitening.
• Equalization and decoding are for their part implemented by the baseband processing unit, which allows advanced reception processing, in particular iterative processing based on the subtraction of the estimated interferences.
• the solution notably proposes a projection technique implemented by the radio unit, making it possible to reduce the quantity of signals intended for the baseband processing unit.
• the projection is implemented on a vector of R complex samples obtained from the R frequency representations (one sample for each frequency representation).
• the sample corresponding to this resource element i.e. to this subcarrier of an OFDM symbol
• the R frequency representations are identified in each of the R frequency representations.
• the projection is implemented for the useful resource elements and for the reference resource elements (i.e. before de-mapping). In this way, the transmission of control information from the radio unit to the baseband processing unit is dispensed with.
• a useful resource element carries one or more data symbols
• a reference resource element carries one or more reference symbols.
• a reference signal identifies the set of reference symbols that can be used for channel estimation.
• the covariance matrix K I therefore represents the noise plus interference before projection.
• the matrix K 1 therefore represents the noise plus interference resulting after projection.
• the vector y 1 of L samples projected at the output of the projection is expressed in the form: with a size vector with I L an identity matrix of size L x L, representing noise plus interference resulting after projection and bleaching.
• the second channel estimation implemented by the baseband processing unit implements the estimation of the channel matrix G b H after projection, from the reference signal after projection, by DMRS example, to rebuild the model
• L v
• the projection matrix G can be with the matrix equal to This projection has the particularity of being lossless information on signal x.
• the projection matrix can be equal to with a matrix carrying L vectors of dimension R corresponding to L arrival directions in reception.
• the method further comprises the transmission, from the radio unit to the baseband processing unit, of a type of projection implemented.
• the radio unit transmits to the baseband processing unit an indicator indicating whether the projection is a projection followed or not by whitening, etc.
• the baseband processing unit knows the type of projection implemented.
• the baseband processing unit can inform the radio unit of the functionalities implemented by the baseband processing unit. For example, if the baseband processing unit implements DMRS-based channel estimation, it can inform the radio unit, which knows that it is not necessary to transmit control information in this case.
• the invention also relates to a corresponding method for receiving radio frequency signals on R antennas, implemented by a radio unit, comprising:
• - de-mapping frequency representations identifying useful resource elements, carrying data, and reference resource elements, carrying at least one reference signal, one useful resource element carrying v data symbols and one resource element reference carrying v reference symbols, including at least one non-zero reference symbol, with the number of spatial layers used for transmission,
• reception method could of course comprise the different characteristics relating to the processing method as implemented by the radio unit, which can be combined or taken separately.
• the characteristics and advantages of the reception method are the same as those of the processing method as implemented by the radio unit described previously.
• the invention relates to the corresponding radio unit.
• the invention also relates to a corresponding method for decoding radiofrequency signals received on R antennas, implemented by a baseband processing unit, including:
• Such a decoding method could of course comprise the different characteristics relating to the processing method as implemented by the baseband processing unit, which can be combined or taken separately.
• the characteristics and advantages of the decoding method are the same as those of the processing method as implemented by the baseband processing unit described previously.
• the invention relates to the corresponding baseband processing unit.
• the invention also relates to one or more computer programs comprising instructions for the implementation of a processing, reception or decoding method as described above when this or these programs are executed by at least one processor .
• the invention finally relates to a system comprising at least one radio unit, configured to process radiofrequency signals received on R antennas, with ff>2, and at least one corresponding baseband processing unit.
• the radio unit comprises:
• the baseband processing unit comprises: means for receiving the vector of L samples projected transmitted by the radio unit , means for de-mapping the L projected samples, identifying at least one reference signal after projection, means for estimating the transmission channel of the radio frequency signals, after projection, from said at least one reference signal after projection, delivering a second channel estimate, means for equalizing the L projected samples, taking account of the second channel estimate, means for processing the equalized symbols.
• FIG. 1 illustrates different options for the separation of functionalities between the radio and baseband processing units, according to 3GPP;
• FIG 3 shows the main steps implemented by a radio unit and a baseband processing unit according to a particular embodiment of the invention
• FIG 4 shows the simplified structure of a radio unit according to a particular embodiment
• FIG. 5 shows the simplified structure of a baseband processing unit according to a particular embodiment.
• the general principle of the invention is based on a new distribution of functionality between the RU and the DU, according to which the RU implements a pre-coding / port reduction functionality based on an accurate channel estimation, and the DU implements implements equalization and decoding functionality, allowing advanced reception processing.
• the RU implementing the active part of the analog processing
• the DU implementing the digital processing
• the DU can be located at the foot of the antenna structure, or in a remote data center, for example located 15-20 km from the RU.
• a DU can serve several RUs (“pooling of resources”).
• An embodiment of the invention is described below in the context of a 5G network, according to which one or more terminals can share the same time-frequency resources.
• a time-frequency resource is a frequency (sub-band) and time (one or more OFDM symbols) granularity.
• a sub-band can range from a resource element (a subcarrier of an OFDM symbol), also called “Resource Element” or RE in English in the 3GPP standards, to a resource block (12 REs), also called “Physical Resource Block” or PRB in English, or to several PRBs.
• the base station receives different radiofrequency signals on R antennas, corresponding to the transmission, by at least one terminal or UE, of a physical channel PUSCH (“Physical Uplink Shared Channel”).
• a PUSCH channel can include several spatial layers v.
• a slot having a duration of 0.5 ms for a spacing between sub-carriers of 30 Khz in the “New Radio” standard several physical PUSCH channels can in particular be transmitted from different terminals. These can be multiplexed in frequency, time, space (MU-MIMO).
• physical channel is meant here a channel of the physical layer originating from a specific user who provides the means of transmitting by radio data/reference signals originating from the MAC layer (or from transport channels).
• FIG. 3 presents the main steps implemented by an RU 31 and by a DU 32 according to one embodiment of the invention, in the uplink.
• R antennas or R reception branches, with R 3 2, each receiving a different version of the same signal, corresponding to the combination of signals transmitted by at least one user terminal UE, for example a PUSCH.
• Each antenna r, with R 3 r 3 2 therefore receives a radiofrequency signal.
• the RU 31 performs a processing 311r on the radiofrequency signal received on each antenna r, making it possible to obtain a frequency representation of the radiofrequency signal received on each antenna r.
• Each frequency representation is formed from a set of complex samples.
• reception of the radiofrequency signal is understood to mean anything that corresponds to the active part of the analog processing (filtering, amplification, frequency transposition), without the analog-to-digital conversion, the analog-to-digital conversion 3112 of the radiofrequency signal received, the suppression 3113 of the 'guard interval if a guard interval was inserted before transmission, the transformation 3114 from the time domain to the frequency domain of the digital signal, delivering a frequency representation of the radio frequency signal received.
• the RU 31 also implements a de-mapping 312 of the R frequency representations, also called RE-demapping.
• de-mapping makes it possible to separate the resource elements carrying data, called useful resource elements, from the resource elements carrying the reference signals, called reference resource elements.
• a useful resource element can carry v data symbols, with v > 1 the number of spatial layers used for data transmission.
• the RU 31 also implements a first estimate 313 of the transmission channel of the radiofrequency signals and of the covariance of the noise plus interference affecting the radiofrequency signals, from said at least one reference signal extracted from the de-mapping 312, for example a DMRS.
• the data carried by the useful resource elements and the reference signals carried by the reference resource elements can, before RE-demapping 312, be filtered by the RU to reduce the dimension of the signal, taking into account the first channel estimate and noise plus interference covariance 313.
• the R frequency representations can be filtered by RU 31.
• the RU 31 performs a projection 314 of the R complex samples associated with this resource element (i.e. of the R complex samples obtained respectively from each of the R frequency representations corresponding to this sub-carrier - one sample per frequency representation) onto L complex samples, called projected samples , taking into account the first channel estimate and the covariance of noise plus interference 313, with R > L 3 v. More precisely, the same resource element k (associated with a particular OFDM symbol) is considered for each reception branch, i.e. the same time-frequency position in the R radiofrequency signals received, to construct a vector y of R complex samples.
• the projection step 314, also called precoding or port reduction, is described below in more detail.
• the matrix representative of the transmission channel associated with N PUSCH transmitted on the same resource element can be written likewise the Spatial layers used to transmit data or reference symbols can be written as a dimension vector.
• the vector x corresponds to a reference resource element, this carries symbols known to the receiver for channel estimation.
• the first channel and noise plus interference covariance 313 estimation from DMRS reference signals allows to determine the channel matrices and covariance
• the covariance matrix K represents noise plus interference before screening.
• the projection matrix G makes it possible to reduce the dimensions of the vector of the samples received y while attempting to keep sufficient statistics (without loss of information) on the transmitted symbols x for their reception.
• the matched filter is known to provide sufficient statistics in the presence of white noise by projecting the received signal and the noise onto the subspace of the useful signal.
• the projection 314 is not followed by whitening of the noise.
• the projection matrix is applied to the vector y, by the RU 31, to reduce the model to a dimension L, and to obtain at the output of the projection 314, a vector of complex samples, called projected samples, with
• the matrix therefore represents the noise plus interference resulting after projection.
• the projection matrix G can be this projection has the particularity of being without loss of information on the signal x.
• the projection matrix G can be which turns out to be a good approximation of the whitening matched filter without the complexity of inverting the covariance K I when the matrix K I approaches a multiple matrix of the identity.
• the vector y 1 of L projected samples can thus be transmitted to the DU 32 for the user data, via the DU/RU (“fronthaul”) interface, for example by an optical fibre.
• the projection 314 can be followed by whitening of the noise.
• the projection applied to vector y, by RU 31, is followed by bleaching to reduce the model to an L dimension without noise.
• a vector of complex samples, called projected samples is obtained, with where I L is an identity matrix of dimension L x L.
• the projection matrix G can be followed by whitening, such that this projection has the particularity of being without loss of information on signal x.
• the projection matrix G can be followed bleaching, such as which turns out to be a good approximation of the whitening matched filter without the complexity of inverting the covariance K, when the matrix K, approaches a multiple identity matrix (no or little spatial correlation of noise plus interference).
• the network of reception antennas of dimension R, is linear and that the antennas are uniformly spaced (for example by half a wavelength).
• a direction of arrival/departure of the signal can be associated with a DFT vector of dimension R (where each coefficient of the DFT corresponds to a multiplicative factor to be applied to a different reception antenna for the formation of a reception beam in a given direction).
• the set of orthonormal DFT vectors forms an orthonormal basis of the received signal commonly used to analyze the directions of arrival of the signal in reception.
• a direction (DFT vector) u t is better than another direction u 2 if and only if
• the projection then consists, according to this example, in the succession of the following two steps:
• the projection matrix G can therefore be of dimension L x R.
• the vector y 1 of L projected samples can thus be transmitted to the DU 32 for the user data, via the DU/RU (“fronthaul”) interface, for example by an optical fibre.
• the DU 32 can thus receive the L projected samples transmitted by the radio unit 31 for a resource element k, for each channel PUSCH.
• the DU 32 implements a de-mapping 320 of the L projected samples (carrying data and reference signals), making it possible to separate the resource elements carrying the data, called useful resource elements, from the resource elements carrying the signals of reference, called reference resource elements.
• the reference signals carried by the reference resource elements for example DMRS, after projection, are used by the DU 32 to perform a second estimation 321 of the transmission channel after projection.
• DMRS-based channel estimation is therefore performed once by RU 31, once by DU 32.
• the second estimate only estimates the transmission channel after projection G b H.
• the interference for example the variance or the covariance of the noise plus interference
• the second estimate can also estimate the interference after projection K 1 .
• the DU 32 can then implement an equalization 322 of the L projected samples, taking into account the second DMRS-based channel estimate 321.
• the second channel estimate makes it possible in particular to estimate that GH, as well as if the projection is not followed by bleaching.
• the DU 32 implements a processing 323 j of the equalized symbols, for each user j, 1 £
• the purpose of the equalization is to process the interference between spatial layers in order to estimate/detect the transmitted symbols.
• the equalizer is for example an LMMSE-IRC equalizer according to 3GPP, ie a filter as estimated minimizes with the variance of the variable random X.
• / is proportional to the ith row of the matrix
• the DU receives the vector y 1 .
• Different techniques can in particular be implemented to inform the DU 32 of the type of projection implemented by the RU 31.
• the DU knows that the mapping implemented by the RU is implemented on the useful and reference resource elements by configuration, or because it receives a message from the RU signaling this information to the DU, or because it does not receive control information.
• the DU thus knows that a RE-demapping and a second channel estimation must be implemented by the DU.
• the DU can choose an option (projection implemented on the useful resource elements only, or on the useful and reference resource elements), and inform the RU of the chosen option.
• the processing of equalized symbols is a conventional processing.
• a feedback loop with equalization 322 can be provided.
• equalization can be performed jointly or disjointly by PUSCH.
• a feedback loop between the decoding of all users and the equalization of users is possible in the case of an advanced receiver.
• such a radio unit comprises at least one memory 41 comprising a buffer memory, at least one processing unit 42, equipped for example with a programmable calculation machine or a dedicated calculation machine, for a processor P, for example, and controlled by the computer program 43, implementing steps of the reception method according to at least one embodiment of the invention.
• the code instructions of the computer program 43 are for example loaded into a RAM memory before being executed by the processor of the processing unit 42.
• the processor of the processing unit 42 implements steps of the reception method described previously, according to the instructions of the computer program 43, for:
• each frequency representation being formed of a set of complex samples
• such a baseband processing unit comprises at least one memory 51 comprising a buffer memory, at least one processing unit 52, equipped for example with a programmable calculating machine or a dedicated calculation, for example a processor P, and controlled by the computer program 53, implementing steps of the decoding method according to at least one embodiment of the invention.
• the code instructions of the computer program 53 are for example loaded into a RAM memory before being executed by the processor of the processing unit 52.
• the processor of the processing unit 52 implements steps of the decoding method described above, according to the instructions of the computer program 53, to:

Landscapes

• Engineering & Computer Science (AREA)
• Computer Networks & Wireless Communication (AREA)
• Signal Processing (AREA)
• Power Engineering (AREA)
• Radio Transmission System (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=77710916

Family Applications (1)

Country Status (3)

Family Cites Families (2)

• 2021
  • 2021-06-04 FR FR2105931A patent/FR3123774A1/fr not_active Withdrawn
• 2022
  • 2022-06-03 EP EP22735014.7A patent/EP4348874A1/de active Pending
  • 2022-06-03 WO PCT/FR2022/051059 patent/WO2022254161A1/fr active Application Filing

Also Published As

Similar Documents

Legal Events