EP2389741A1

EP2389741A1 - Method for encoding, for a radio infrastructure, data with a double interleaving of parity symbols, and associated codec

Info

Publication number: EP2389741A1
Application number: EP10707323A
Authority: EP
Inventors: Bessem Sayadi; Yann Leprovost
Original assignee: Alcatel Lucent SAS
Current assignee: Alcatel Lucent SAS
Priority date: 2009-01-23
Filing date: 2010-01-19
Publication date: 2011-11-30
Also published as: US8677210B2; FR2941580A1; US20120008706A1; WO2010084283A1; JP2012516099A

Abstract

The invention relates to a method dedicated to encoding data to be sent via a transmission infrastructure using waves, which includes: i) a step comprising simultaneously creating M first matrices with T rows and C columns with data subsets of B consecutive bursts received, the data subsets of each burst being distributed in at least two first consecutive matrices; ii) a step comprising simultaneously creating M second matrices with T rows and N columns with parity symbols resulting from encoding data respectively contained in the rows of each one of the M first matrices; iii) a step comprising simultaneously creating M third matrices with K rows and C columns with parity symbols resulting from encoding data respectively contained in the columns of each of the M first matrices; and iv) a step comprising distributing by interleaving J subsets of parity symbols of each second matrix in J consecutive sets, as well as P subsets of parity symbols of each third matrix in P of said consecutive sets, and placing the respective data of the consecutive bursts received in each one of the consecutive sets.

Description

METHOD FOR ENCODING DOUBLE INTERLEAVING DATA OF PARITY SYMBOLS, FOR RADIO INFRASTRUCTURE, AND ASSOCIATED CODEC

The invention relates to the transmission of data by wave transmission infrastructures.

By "wave transmission infrastructure" is meant here any communication infrastructure in which the data transmission takes place by means of waves. Therefore, it may be both a radio communication network and a broadcast network. As non-limiting examples, it may therefore be a wireless broadcast network (for example a terrestrial network of DVB-H type (for "Digital Video Broadcasting - Handhelds" - mobile TV), or a hybrid network (that is to say both satellite and terrestrial (for example a DVB-SH type network (for "Digital Video Broadcasting - Satellite Services to Handhelds") (or DVB-SSP) - coupled satellite channel to a terrestrial radio relay channel))), or a cellular or mobile network (such as for example a GSM / EDGE or UMTS type network) or even a metropolitan network or MAN (for "Metropolitan Area

Network "- such as a WiMAX type network).

It is recalled that the DVB-SH network is a hybrid variant of the DVB-H network (for "Digital Video Broadcasting - Handhelds") which has been developed for mobile television, ie for the one-way broadcasting of content. , such as television programs, in broadcast mode (broadcast / point-to-point) or in multicast mode (point-to-multipoint). In a DVB-SH network, the satellite channel is intended to provide global coverage while the terrestrial relay radio channel is intended to provide the ground with a cellular type coverage. It is also recalled that the transmission of content in broadcast mode or in multicast mode is done through dedicated services, which may possibly be time multiplexed. As is known to those skilled in the art, the radio signals transmitted by an infrastructure, for example of the hybrid type, are subject to degradation, in particular when they borrow the band S (between about 1. 55 GHz and about 5.2 GHz). The level of these impairments may vary depending on the environment of the communication terminals which are the recipients of the transmitted data (possibly content). In fact, the propagation channel can be in different states depending on the level of weakening of the direct signal induced by a local shadow zone (for example because of the presence of tree (s) or building (s). )). For example, by applying a Markov model to a DVB-SH type infrastructure, it can be shown that the channel can be in three states called LOS ("for" Line Of Sight "). the transmitter)), shaded (or "shadowing") and blocked or not passing (or "blockage"). In the presence of a satellite angular elevation of about 40 °, the Markov model reproduces large or very small signal attenuation variations in two main environments called "intermediate shading by trees" (or ITS for Intermediate Tree Shadowing ") and suburban (or SUB for" suburban ").

In order to guarantee mobile terminals a high quality of service in the presence of difficult radio conditions (for example of the deep attenuation or misalignment type with respect to the transmitter), certain networks, such as for example those of the DVB-SH type, implement a large-scale temporal diversity. The latter can be provided by interleaving (or "interleaving") performed at the level of the physical layer (for class 2 terminals) or the link layer (for class 1 terminals). Interleaving at the link layer is an IP interleaving called MPE-IFEC (for "MulitiProtocol Encapsulation - Inter-bursts Forward Error Code" - "Multiprotocol Encapsulation - Forward Error Encoding") ). It is recalled that the technique called FEC consists of adding the emission side of the redundancy to data to enable correction on the receiving side of part of the errors introduced by losses on the transmission channel of these data. The MPE-IFEC technique is based on the paralleling of the first M (encoding / decoding) matrices, called ADTs (for "Application Data Tables"), and made up of subsets of data of at least one burst (possibly IP (Internet Protocol)) received, the subsets of data of each burst being distributed in at least one encoding / decoding matrix, then the paralleling of M second matrices, called FDT (for "FEC Data Tables") and consist of parity symbols resulting from an encoding (Reed-Solomon type) of the data contained in the first M matrices, and finally the distribution by simple interleaving subsets of parity symbols of each second matrix in S successive sets containing at least the respective data of successive bursts received. It will be noted that the S FEC data sets are sent with the associated data in a time slice generally called burst time slice (or "time-slice burst"). The MPE-IFEC coding technique is also described in DVB Bluebook A.131, entitled "MPE-IFEC (draft TS 102 772 V1.1.1)", published in November 2008 by the ETSI (European Telecommunications Standard Institute). . An ITS environment imposes an increase in the length of consecutive erroneous bursts, and a reduction in the length of consecutive non-errored bursts is reduced. On the other hand, a SUB environment imposes a reduction in the length of the consecutive erroneous bursts, and an increase in the length of the consecutive non-erroneous bursts. However, it can be shown that the MPE-IFEC technique is truly efficient in the presence of an ITS environment only if the depth of the interleaving is important (typically 30 seconds, which corresponds to a variable S of value high), while it is truly efficient in the presence of a SUB environment only if the depth of the interleaving is low (typically 10 seconds, which corresponds to a variable S of low value). The operator of a network type DVB-SH is therefore forced to configure the latter according to the worst environment, namely the ITS environment. The invention therefore aims to improve the situation in a wave transmission infrastructure.

To this end, it proposes a method dedicated to the coding of data to be transmitted by means of a wave transmission infrastructure, and comprising the following steps:

constituting in parallel M first matrices of T lines and C columns with subsets of data of B successive bursts (or "bursts") received, the subsets of data of each burst being distributed in at least two first successive matrices in parallel M second rows of T rows and N columns with parity symbols resulting from an encoding of the data which are respectively contained in the lines of each of the first M first matrices,

constituting in parallel M third matrices of K rows and C columns with parity symbols resulting from an encoding of the data which are respectively contained in the columns of each of the first M matrices,

interleaving, on the one hand, J sub-sets of parity symbols of each second matrix in J successive sets, and, on the other hand, P subsets of parity symbols of each third matrix in P of these successive sets, and place in each of the successive sets the respective data of the successive bursts received.

The method according to the invention may comprise other characteristics that can be taken separately or in combination, and in particular:

the variable M can be defined by the equation M = B + max (0; max (J; P) - D) + max (0; D - B), where max () is the maximum function and D is a a delay, expressed in number of bursts, between the instant of reception of the data of a burst and the instant of transmission by the data infrastructure of this same burst;

- the variable B can be equal to 3;

the variable J can be equal to 2; the variable P can be equal to 4;

the first M matrices can be constituted using a technique called "multiprotocol encapsulation (MPE)";

the M second matrices and the third M matrices can be constituted by means of a technique called "direct error correction"

(FEC) ".

The invention also proposes a codec (or "codec-decoder") intended to equip communication equipment adapted to connect to a transmission infrastructure by means of waves, and comprising: coding means responsible for constituting parallel:

• M first matrices of T rows and C columns with subsets of data of B successive bursts received, the subsets of data of each burst being distributed in at least two first successive matrices, • M second matrices of T lines and N columns with parity symbols resulting from an encoding of the data contained respectively in the lines of each of the first M matrices, and

M third matrices of K rows and C columns with parity symbols resulting from an encoding of the data contained respectively in the columns of each of the first M matrices, and

interleaving means arranged to interleave, on the one hand, J subsets of parity symbols of each second matrix in J successive sets, and, on the other hand, P subsets of parity symbols of each third matrix in P of these successive sets, and to place in each of the successive sets the respective data of successive bursts received.

The coding means may for example be responsible for forming the first M matrices by means of a technique known as "multiprotocol encapsulation (MPE)". Alternatively or in addition, the coding means may for example be responsible for constituting the M third matrices by means of a technique called "direct error correction (FEC)" The invention is particularly well suited, although not exclusively, to hybrid networks of the DVB-SH (or DVB-SSP) type, as well as to all their evolutions. However, the invention generally applies to any type of wave data transmission infrastructure. Other features and advantages of the invention will appear on examining the detailed description below, and the attached drawings, in which:

FIG. 1 very schematically and functionally illustrates a hybrid transmission infrastructure that makes it possible to implement the invention, and

FIG. 2 partially and schematically illustrates an exemplary IP burst data coding according to the invention.

The attached drawings may not only serve to complete the invention, but also contribute to its definition, if any. The object of the invention is to propose a data coding method for the transmission of these data by a wave transmission infrastructure.

In what follows, it is considered as a non-limiting example that the wave transmission infrastructure is hybrid (PFC, SAT, RA), and more specifically that it is a network. DVB-SH type (or

DVB-SSP). However, the invention is not limited to this type of wave transmission infrastructure. It concerns indeed any type of infrastructure for transmitting data (possibly content, possibly multimedia), by wave, to radio communication terminals. Therefore, it may be both a radio communication network and a broadcast network. By way of nonlimiting examples, it may also be a wireless broadcast network (for example a terrestrial network of DVB-H type (for "Digital Video Broadcasting - Handhelds" - mobile TV), or a cellular or mobile network (such as for example a network type

GSM / EDGE or UMTS) or even a metropolitan network or MAN (for "Metropolitan Area Network" - such as a WiMAX type network). Furthermore, it is considered in the following, by way of non-limiting example, that the radio communication terminals (TC) are mobile phones (or cellular) or communicating digital personal assistants (or PDAs). However, the invention is not limited to this type of radio communication terminal. It concerns indeed all communication equipment, fixed or mobile (or portable or cellular), capable at least to receive data over the airwaves via an infrastructure of the aforementioned type. Therefore, it may also be a fixed or portable computer, a multimedia content receiver (such as a decoder, a residential gateway (or "residential gateway") or a STB ("Set-Top Box"). ")), Provided that it is equipped with means of communication by waves able at least to the reception of data.

In addition, it is considered in the following, by way of non-limiting example, that the data broadcast to the terminals (TC) are multimedia content data such as television programs. But, the invention is not limited to this type of data. It concerns any type of content, including videos, file data (or "data"), signaling data, radio programs, and audio content.

As is schematically illustrated in FIG. 1, the implementation of the invention requires the existence of a wave transmission infrastructure, here of the hybrid type and therefore comprising a satellite transmission channel and a communication channel. terrestrial radio transmission.

The satellite transmission channel comprises a PFC satellite platform (or gateway) for the provision of encoded contents and at least one SAT communication satellite coupled to each other by means of waves.

The satellite platform PFC is for example coupled to a content server SC which feeds it into contents in the form of bursts (or "bursts"), here of IP type (it could indeed be packet data type NAL (video) or RTP, for example). It is responsible for encoding these received contents by means of a coded CD (implementing a method according to the invention) before transmitting them by waves to the (communication) satellite SAT, which then takes care of the retransmit (broadcast) to TC terminals, either directly or indirectly via the terrestrial radio transmission channel.

This terrestrial radio transmission channel comprises at least one radio access network RA which may for example be part of a mobile (or cellular) communication network. This is the assumption that is retained in the following, by way of non-limiting example. For example, this radio access network RA is of the UTRAN (or 3G) type. It therefore mainly comprises N base stations (called Node Bs in the case of a UTRAN) and radio network controllers CR (called RNCs in the case of a UTRAN), connected together, as well as satellite communication PS which is coupled by waves to the satellite SAT (to receive the encoded contents) and wired to the radio network controllers CR (to feed encoded content received.

The TC terminals can receive the encoded contents either directly from the satellite SAT, when they are not located in a shadow zone, or N base stations, when they are located in a shadow zone.

Each TC terminal includes a CD codec, similar to that which equips the satellite platform PFC, so as to be able to decode the encoded contents it receives.

The invention proposes implementing, within the CD codec of the PFC satellite platform, a method of coding the content data received in the form of IP bursts (and more precisely encapsulated in

RTP / UDP / IP. It is recalled that we call IP datagrams the data that are contained in an IP burst.

The method according to the invention comprises four main steps that are performed by the CD codec when its satellite platform PFC receives bursts of content data (s) from the content server SC.

A first main step of the method according to the invention consists in forming in parallel M first matrices of T lines and C columns with data subsets of B bursts IP. It will be understood that we operate here by means of sliding windows.

For example, the variable M can be defined by the following equation: M = B + max (0; max (J; P) - D) + max (0; D - B), where max () is the maximum function, J and P are two variables representative of two interlace depths on which will be discussed later, and D is a delay which is expressed in number of bursts S 1 and which represents the time difference between the instant of reception of the data of a burst Si and the moment of transmission by the hybrid infrastructure (and more precisely by the satellite platform PFC) data of this same burst If, once encoded by the coded CD.

Each first matrix can be considered as a block of B sub-blocks which each consist of a number of columns of data from a burst IP Si (and constituting a subset of the latter).

Variable B is at least two. Consequently, the subsets of data of each burst Si are distributed in at least two first successive matrices (Si and Si + 1). For example, the value of B can be three or four or even more.

The number C of columns of each first matrix is, for example, equal to 191. The number T of rows of each first matrix is, for example, at most equal to 1024. It will be noted that the first M matrices can for example be constituted by means of the technique called "multiprotocol encapsulation" (or MPE). Note also that when the encoding technique is that which is called MPE-IFEC, each first matrix is what one skilled in the art calls an application data table (ADT). )).

A second main step of the method according to the invention consists in forming in parallel M second matrices of T lines and N columns with parity symbols which result from an encoding of the data which are respectively contained in the T lines of each of the first Ms. matrices formed during the first main step.

It will be understood that each second matrix is derived from (and therefore associated with) one of the first M matrices.

For example, the second M matrices can be constituted at the means of the technique called "direct error correction" (or FEC). In this case, the line encoding is of the Reed-Solomon (or RS) type and each second matrix is what the person skilled in the art calls a FEC data table (or FEC Data Table). ). The number N of columns of each second matrix is for example equal to 64 (in the case DVB-SH).

A third main step of the method according to the invention consists in constituting in parallel M third matrices of K rows and C columns with parity symbols which result from an encoding of the data which are respectively contained in the C columns of each of the first Ms. matrices formed during the first main step.

It will be understood that each third matrix is derived from (and therefore associated with) one of the first M matrices.

For example, the M third matrices can be formed using the so-called "direct error correction" (FEC) technique. In this case, the line encoding is of the Reed-Solomon (or RS) type and each third matrix is what a person skilled in the art calls a data table.

FEC (or FDT (for "FEC Data Table")).

The number K of rows of each third matrix is for example equal to 64 (in the case DVB-SH).

It is important to note that the second and third main steps can be performed substantially simultaneously.

Note that the second and third matrices that come from

(and therefore associated with) one of the first M matrices can be considered as two complementary parts of a "product matrix" consisting of parity symbols that can be called here "product codes" Cij (i and j denote respectively a row and a column).

Each row of this product matrix constitutes, for example, a codeword

RS of the Galois field GF (q) having an error correction capacity t1, and each column of this product matrix constitutes for example a codeword

RS Galois field GF (q) having an error correction capability t2.

It should be noted that the first, second and third main steps can be implemented by an encoding module MC of the coded CD which equips the PFC satellite platform.

A fourth main step of the method according to the invention consists in particular of distributing by means of a double interlacing (or "interleaving"), on the one hand, J subsets of parity symbols of each second matrix in J successive sets Ei and, on the other hand, P subsets of parity symbols of each third matrix in P of these successive sets Ei.

As previously indicated, the variables J and P are representative of the two interleaving depths of the encoding method according to the invention. The term "interleaving depth" here means the number of successive sets of data Ei in which the subsets of parity symbols of a second or third matrix are distributed.

One of the two variables J and P may be greater than or equal to one, while the other must be greater than or equal to two. For example, the value of J can be equal to two or three or even more, and the value of P can be three or four, or even more.

Note that J may be equal to P, but this is not mandatory.

The fourth main step also consists in placing in each successive set Ei the respective data of successive bursts Si received. Each set Ei produced by the coding method according to the invention is thus finally constituted by at least data of one of the received bursts S 1, subset J subsets of parity symbols from J second matrices and P subassemblies parity symbols from P third matrices.

It will also be noted that the fourth main step can be implemented by an interleaving module ME of the codec CD which equips the satellite platform PFC and which is coupled to the coding module MC of the codec CD. FIG. 2 schematically shows an exemplary IP burst data coding performed by means of a CD coding implementing the method according to the invention. In this example, the variable M is equal to 4, the variable B is equal to 3, the variable J is equal to 2 and the variable P is equal to 4. Moreover, the "upper" part of FIG. 2 shows eight salvos S1 to S8 successively received by the satellite platform PFC, the "intermediate" part of FIG. 2 represents the four (M = 4) associations. parallel of a first matrix (ADT) ₁ of a second matrix (FDT1), resulting from this first matrix (ADT), and a third matrix (FDT2), resulting from this same first matrix (ADT), and the "lower" part of Figure 2 materializes five sets E4 to E8 constituted successively ("over water") from bursts Si previously received and second (FDT1) and third (FDT2) matrices of the aforementioned associations (in accordance with unidirectional arrows).

On the reception side (that is to say in the TC terminals), the erroneous bursts are corrected progressively since the above-mentioned sets Ei, which are transmitted successively via the satellite SAT and via the radio access network RA, comprise complementary subsets of parity symbols, and therefore it is necessary to wait to have received at least (B + max (JP) - D) time slots of salvo (sets Ei and FEC associated) to have the completeness parity symbols (FEC) needed to correct the content data of a salvo if initially received by the satellite platform PFC. On the reception side, the decoding of the content data (which includes any corrections) is ensured by the CD codecs of the TC terminals.

The invention is particularly advantageous because it makes it possible to perform an effective correction both in an ITS environment and in a SUB environment. In addition, it reduces the bandwidth used for error correction redundancy because the coding rate is reduced, and therefore it increases the number of services broadcast due to the higher performance offered by the codes. of product regarding error correction. In addition, the invention offers operators more flexibility to size their hybrid infrastructures, as it improves the quality of services and saves bandwidth. Finally, since the interleaving depths (J and P) may be small, the perceived quality of the contents is improved and the switching time between different channel contents (or "zapping time") is reduced (in case of error it is sufficient to wait for a delay equal to min (J, P) to change the channel).

The invention is not limited to the embodiments of encoding method, coding and platform (or gateway) described above, only by way of example, but it encompasses all the variants that may be considered by the man of art within the scope of the claims below.

Claims

A method of encoding data to be transmitted by means of a wave transmission infrastructure (PFC ₁ SAT, RA), said method comprising the steps of:

constituting in parallel M first matrices of T lines and C columns with successive subsets of data of B successive bursts, the subsets of data of each burst being distributed in at least two first successive matrices, - constituting in parallel M second matrices of T lines and N columns with parity symbols resulting from an encoding of the data respectively contained in the lines of each of said M first matrices,

constituting in parallel M third matrices of K rows and C columns with parity symbols resulting from an encoding of the data contained respectively in the columns of each of said first M matrices,

- Interleaving, on the one hand, J subsets of parity symbols of each second matrix in J successive sets, and on the other hand, P subsets of parity symbols of each third matrix in P of said successive sets. , and placing in each of said successive sets the respective data of said successive bursts received.

2. The method according to claim 1, wherein M is defined by the equation M = B + max (0; max (J; P) -D) + max (0; D-B), where max () is the maximum function and D is a delay, expressed in number of bursts, between the instant of reception of the data of a burst and the moment of transmission by said infrastructure (PFC, SAT, RA) data of this same burst.

3. Method according to one of claims 1 and 2, B is equal to 3.

4. Method according to one of claims 1 to 3, wherein J is equal to 2.

5. Method according to one of claims 1 to 4, wherein P is equal at 4.

6. Method according to one of claims 1 to 5, wherein said M first matrices are formed by means of a technique known as "multiprotocol encapsulation (MPE)".

7. Method according to one of claims 1 to 6, wherein said M second matrices and said M third matrices are constituted by means of a technique called "direct error correction (FEC)".

Coded (CD) for a communication equipment (PFC) of a wave transmission infrastructure (PFC ₁ SAT, RA), said codec (CD) comprising:

Coding means (MC) arranged to constitute in parallel i) M first matrices of T lines and C columns with subsets of data of B successive bursts received, the subsets of data of each burst being distributed in minus two first successive matrices, ii) M second matrices of T lines and N columns with parity symbols resulting from an encoding of the data respectively contained in the lines of each of said M first matrices, and iii) M third matrices of K lines and C columns with parity symbols resulting from an encoding of the data respectively contained in the columns of each of said M first matrices, and

Interleaving means (ME) arranged for interleaving, on the one hand, J subsets of parity symbols of each second matrix in J successive sets, and, on the other hand, P subsets of symbols of parity of each third matrix in P said successive sets, and to place in each of said successive sets the respective data of said successive bursts received.

9. The code according to claim 8, wherein said coding means (MC) are arranged to constitute said first M matrices by means of a technique called "multiprotocol encapsulation (MPE)".

10. The code according to one of claims 8 and 9, wherein said coding means (MC) are arranged to constitute said M third matrices by means of a technique called "direct error correction" (FEC) ".