GB2430586A

GB2430586A - Transmission of data encoded using an irregular low density parity check (LDPC) code

Info

Publication number: GB2430586A
Application number: GB0518756A
Authority: GB
Inventors: Mohamed Rafiq Ismail
Original assignee: Toshiba Research Europe Ltd
Current assignee: Toshiba Europe Ltd
Priority date: 2005-09-14
Filing date: 2005-09-14
Publication date: 2007-03-28
Anticipated expiration: 2025-09-14
Also published as: GB0518756D0; US20070079213A1; JP2007089157A; GB2430586B

Abstract

Wireless transmission of data is effected across a communications channel defined by a communications medium by means of an encoder, operable to apply a low density parity check (LDPC) code to data for transmission 24. The LDPC code is irregular with respect to the degree of variable nodes, and so the transmitter 20 further comprises means for sorting encoded data 26 with respect to the corresponding variable node degree, and modulation 30 and distribution 28 means for allocating encoded and sorted data onto the communications medium.

Description

1 2430586 Wireless Communications Apparatus This invention relates to

apparatus, a method, and source code suitable for use in effecting wireless communications, and particularly directed to wireless communication involving channel coding.

Channel coding is employed in a communications system to afford greater protection from error to the transmitted data. Various channel coding systems exist, several of which result in an expansion of the amount of data required to be transmitted, thereby adding redundancy to the transmission.

A number of encoding-decoding techniques are well documented and commonly found in communication devices. Convolutional coding, turbo coding and parity check coding are a few of the schemes commonly used. "Low-Density Parity-Check Codes." (R. G. Gallager, PhD thesis, 1960, Massachusetts Institute of Technology) describes a method of coding which allows for a comparatively simple decoding scheme. Further refinement on much work has been done on Low Density Parity Check (LDPC) codes, for instance: "Good codes based on very sparse matrices" (D. Mackay, R. Neal, in BOYD, C (Ed):"Cryptography and Coding". IMA Conf., 1995, (Springer), pp. 100-111 (number 1025 in Lecture Notes in Computer Science)); "Near Shannon limit performance of low density parity check codes" (D. Mackay, R. Neal, IEEE Electronics Letters, vol. 33, No. 6, March 1997); "A linear time erasure-resilient code with nearly optimal recovery" (N. Alon, M. Luby, IEEE Trans. Information Theory, pp. 1732-1736, Nov. 1996); "Efficient encoding of low-density parity-check codes" (T. J. Richardson, R. Urbanke,, IEEE Trans. Information Theory, vol. 47, pp. 638-656, Feb. 2001); and "Design of capacity-approaching irregular low-density parity-check codes" (T.

Richardson, A. Shokrollahi, R. Urbanke, , IEEE Trans. Information Theory, vol. 47, pp. 619-637, Feb. 2001).

An LDPC code can be expressed as a factor graph, showing the relationship between variable nodes and check nodes. Factor graphs were introduced in " A recursive approach to low complexity codes" (R. M. Tanner, IEEE Trans. Information Theory, pp. 533-547, Sept. 1981) and used also in "An introduction to LDPC Codes" (W. E. Ryan, Handbook for Coding and Signal Processing for Recoding Systems (B. Vasic, ed.), CRC Press, 2004). Factor graphs are also known as Tanner graphs.

LDPC factor graphs can be either regular or irregular. The number of connections to each node is termed the degree of the node. The degree of a variable node determines the relative degree of protection conferred upon that node. The higher the degree of a node the better the capability for correction of an information bit associated with that node. This is because there is more information associated with that node from which the identity of the dropped bit can be determined.

For a regular LDPC factor graph, every variable node is of the same degree, as is every check node. It will be understood by the reader that this does not imply that check nodes and variable nodes are of the same degree - check nodes could be of higher or lower degree than variable nodes.

In contrast, fbr an irregular LDPC code the degree of the variable nodes and of the check nodes are each necessarily different.

"A Simple Coded Modulation Scheme based on Irregular Low-Density ParityCheck Codes" (K. Harada, M. Mukai, H. Tsurumi, Intl. Symposium on Info. Theory and its Applications, ISITA 2004, Parma, Italy, Oct. 10-13, 2004) describes a method for sorting an LDPC coded sequence based on variable node degree and then mapping to a signal constellation point using a set partitioning technique. This paper demonstrates the benefit of accounting for node degree in establishing an LDPC code.

The benefits of using multiple input and multiple output (MIMO) antennas at the transmitter and receiver are well documented, for example in "A Simple Transmit Diversity Technique for Wireless Communications" (S. M. Alamouti, IEEE JSAC, vol. 16, no. 8, Oct. 1998, pp. 1451-1458) and "Layered Space-time Architecture for Wireless Communication in a Fading Environment When using Multiple Antennas" (G. J. Foschini, , Bell Labs Tech. J., vol.! no 2, autumn 1996, pp. 4 1-59).

Further, "Design of Low-Density Parity-Check Codes for Modulation and Detection" (S. ten Brink, G. Kramer, A. Ashikhmin, , IEEE Trans. Comms, vol. 52, no. 4, April 2004, pp. 670-678) and "Best Mapping for LDPC Coded Modulation on SISO, MIMO and MAC Channels" (J. Wu, Fl-N Lee, Proc. IEEE Wireless Communications and Networking Conf., March 2004, pp. 2428-243 1) describe the use of codes which can he expressed on factor graphs (of which LDPC codes are an example) within a MIMO system resulting in enhanced performance in comparison with traditional channel coding/MIMO solutions.

Variable node degree distribution information is not currently utilised for assigning which transmit antenna(s), in a MIMO system, a modulated symbol will be transmitted from based on channel information. Similarly, in an OFDM based transmission scheme no account is taken of variable node degree distribution in bit loading the sub-carriers.

This is because existing processes require the use of channel state information from the receiver, which introduces further processing overhead to the system.

Therefore, according to the invention, there is provided a wireless transmitter for transmitting data in a communications channel defined in a communications medium, the transmitter comprising an encoder, operable to apply a low density parity check (LDPC) code to data for transmission, the LDPC code being irregular with respect to the degree of variable nodes, sorting means for sorting encoded data with respect to the corresponding variable node degree, and modulation and distribution means for allocating encoded and sorted data onto the communications medium.

According to another aspect of the invention, a transmitter, in use, uses a Tanner graph or the like to apply a parity check, the Tanner graph exhibiting irregularity with respect to variable node degree, and sorts the Tanner graph with regard to variable node degree, then modulates and distributes the data to the communications medium concerned.

Corresponding methods of transmitting, and receivers and methods of receiving, are also contemplated as being in accordance with aspects of the invention.

Further, it will be appreciated that all of these aspects of the invention can be fulfilled by provision of suitable computer programmable communications means, and a computer program, which may be in the form of a software product supplied on a storage medium or as a downloadable file or collection of files.

Further aspects, features and advantages of the invention will now be described by means of specific embodiments of the invention, provided by way of example only, with reference to the accompanying drawings in which: Figure 1 illustrates a communications system including a MIMO transmitter in accordance with an embodiment of the invention, together with a receiver suitable for use therewith; Figure 2 illustrates a transmitter encoding train in accordance with a second embodiment of the invention; Figure 3 illustrates a transmitter encoding train in accordance with a third embodiment of the invention; Figure 4 illustrates a transmitter encoding train in accordance with a third embodiment of the invention; Figure 5 illustrates a communications device configured in accordance with the transmitter of the first described embodiment.

With reference to the drawings, a first embodiment of the invention is illustrated in figure 1. A simplified MIMO communications system 10 is depicted, comprising a transmitter 20 and a receiver 40. The transmitter 20 and the receiver 40 are shown as units dedicated to transmission and reception respectively, but it will be appreciated that, in use, a communications device will comprise elements of both in order to effect bidirectional communication.

it will be appreciated that, though the invention is described in the context of MIMO transmission, the invention is equally applicable to other transmission modes.

The transmitter 20 comprises a data source 22, which may be any hardware and/or software components generating data for transmission, an irregular LDPC encoder 24, a VND sorter 26, a distribution unit 28, an array of modulators 30 and a corresponding array of anteimas 32.

The irregular LDPC encoder 24 determines an LDPC code and applies this to the data to be transmitted, in accordance with the LDPC code (which, as described above, can be expressed by means of a factor graph). Following application of the irregular LDPC code, the encoded data is passed to a variable node degree (VND) sorter, which sorts the bits of the encoded data by reference to the degree of the variable node on which each bit is encoded. This results in bits encoded on a low degree variable node being ordered ahead of bits encoded on a high degree variable node.

Of course, it will be appreciated by the reader that the reverse could alternatively be provided, that the VND sorter could be configured to sort the bits of the encoded data by reference to the degree of the variable node, such that the bits on a high degree variable node are ordered ahead of the bits on lower degree variable nodes.

The ordered data is then passed to the distribution unit 28 which acts, under the instruction of distribution configuration information stored in the receiver to allocate bits, or groups of bits, to modulator antenna pairs (30, 32). It will be appreciated that this is but one of a number of ways in which this can be achieved. Alternatively, distribution configuration information could be derived from the receiver or some other source. Further, it will be appreciated that the modulator and antenna do not necessarily have to be in pairs: the modulator can in the general case be configured to pass information to antenna processing means in whatever form is appropriate in the circumstances.

The distribution unit 28 makes allocation decisions based on prior information such as, in this example, spatial path gain.

A transmission is thus effected onto the MIMO channel, and this is detected at the receiver 40 by means of an antenna array 42. The signals received at the antennas 42 are then demodulated using a corresponding bank of demodulators 44. A deformatter 46 extracts side- or configuration information from the received signals. It will be appreciated that the side-information may include information about many different aspects of configuration of the communication channel, but in this example, the only information of interest pertains to the sort order established by the VND sorter 26 of the transmitter 20. This side information, together with the main information stream, is passed to a VND de-sorter 48, which reassembles the original order of the bits sorted by the VND sorter 26 in the transmitter 20.

The dc-sorter 48 then passes the re-assembled information to an LDPC decoder 50 which extracts the original information and passes this to a data sink 52 where the data is received and is available for user interaction.

Further configurations of the encoding train used in the transmitter of figure 1 will now be described with reference to figures 2 and 3 for the assistance of the reader in understanding the scope of the invention. Where components have substantially the same function as those illustrated in figure 1, they are given the same reference numbers. For reasons of clarity, the overall structure of the transmitter, such as the data source 22, is omitted.

In figure 2, the series of modulators 30 in figure 1 is replaced by a single modulator 130, interposed between the VND sorter 26 and the distribution unit 28. The distribution unit 28 allocates symbols to one of n subcarriers 131 defined on an array of antennas 32 based on external information, such as channel frequency response. In a particular example of implementation of this embodiment, the subcarriers can be defined by an array of 2 antennas.

In figure 3, the distribution unit 28 allocates bits to one of a series of m distinct modulators 230, preceding n subcarriers 231. m is less than or equal to n. The distribution unit 28 again operates on the basis of channel conditions.

It will be understood that distribution configuration information could alternatively be derived from sources other than the receiver.

It will be further understood that the pairing of modulator and antenna is but one arrangement within the scope of the invention. Alternative arrangements are also possible. For example, a beamforming block could be provided to interpose between modulators and antennas, so that one to one correspondence is not required. Figure 4 illustrates this in further detail - the first four stages of the arrangement illustrated in figure 2 are replicated but then a beam-former 332 interposes between the modulator and the antenna array. In that way, the subcarriers defined in the transmitter are translated to the antennas.

The combination of VND sorter and distribution unit (DU) facilitates exploitation of the irregular protection offered to information bits by the distribution of degree among the variable nodes of the Tanner graph defining the LDPC scheme.

Figure 5 illustrates schematically a wireless communications device configured as the transmitter 20 according to the first embodiment of the present invention. The device comprises a processor 302 operable to execute machine code instructions stored in a working memory 304 and/or retrievable from a mass storage device 306. By means of a general-purpose bus 308, user operable input devices 310 are in communication with the processor 302. The user operable input devices 310 comprise, in this example, a keyboard, but could include a mouse or other pointing device such as a touchpad, a contact sensitive surface on a display unit of the device, a writing tablet, speech recognition means, haptic input means, or any other means by which a user input action can be interpreted and converted into data signals.

Audio/video output devices 312 are further connected to the generalpurpose bus 308, for the output of information to a user. Audio/video output devices 312 include a visual display unit, and a speaker, but can also include any other device capable of presenting information to a user.

A communications unit 314 is connected to the general-purpose bus 308, and further connected to a series of antennas 32. By means of the communications unit 314 and the antennas 32, the device 20 is capable of establishing wireless communication with another device. The communications unit 314 is operable to convert data passed thereto on the bus 308 to an RF signal carrier in accordance with a communications protocol previously established for use by a system in which the device 20 is appropriate for use.

In the device 20 of Figure 3, the working memory 304 stores user applications 316 which, when executed by the processor 302, cause the establishment of a user interface to enable communication of data to and from a user. The applications 316 thus establish general purpose or specific computer implemented utilities and facilities that might habitually be used by a user.

Conventional LDPC based MIMO or OFDM systems do not explicitly account for the non-uniform distribution of variable node degree. Specifically, existing methods do not marry irregular variable node degree with other factors that may affect the error probability of a transmission.

In particular, whereas the channel coding mechanism of conventional systems does not account for the vulnerability or sensitivity of the incoming data to errors, the present invention does. Similarly, conventional channel encoding schemes do not account for the transient conditions of the communications channel.

Figure 3 shows the use of variable degree node to allocate which antenna the encoded information is transmitted on. The lower degree node symbols are transmitted on spatial channels that are deemed to be more reliable, and higher degree node symbols are transmitted on less reliable spatial channels, for example.

Furthermore, while the invention is described in the context of a MIMO system, it will be understood that the invention is not limited in scope to application to MIMO arrangements.

Moreover, while the invention has been described above with regard to a specific embodiment employing OFDM technology, it will be appreciated that this is not essential to the delivery of the invention, and the determination of sub-carriers in the communications medium is also not an essential element of the invention.

Claims

CLAIMS: 1. A wireless transmitter for transmitting data in a

communications channel defined in a communications medium, the transmitter comprising an encoder, operable to apply a low density parity check (LDPC) code to data for transmission, the LDPC code being irregular with respect to the degree of variable nodes, sorting means for sorting encoded data with respect to the corresponding variable node degree, and modulation and distribution means for allocating encoded and sorted data onto the communications medium.
2. A transmitter in accordance with claim I wherein the modulation and distribution means is operable to allocate encoded and sorted data based on a reliability criterion applied to the channel.
3. A transmitter in accordance with any preceding claim and employing OFDM, subcarriers thus being defined in the communications medium.
4. A transmitter in accordance with claim 3 when appended to claim 2, wherein the reliability criterion is applied in respect of separate subcarriers, and the modulation and distribution means is operable to take account of relative reliability of separate subcarriers.
5. A transmitter in accordance with claim 4 wherein the modulation and distribution means is operable to allocate encoded and sorted data associated with a relatively low degree variable node to a subcarrier with relatively high reliability criterion.
6. A transmitter in accordance with claim 1 wherein the modulation and distribution means is operable to allocate encoded and sorted data to an antenna beam forming means, wherein said sorting means is operable to sort in accordance with a criterion of the antenna beam forming means.
7. A method of transmitting data in a communications channel, the method comprising applying a low density parity check (LDPC) code to data for transmission, the LDPC code being irregular with respect to the degree of variable nodes, sorting encoded data with respect to the corresponding variable node degree, and allocating encoded and sorted data onto the communications channel.
8. A method in accordance with claim 7 wherein the step of allocating includes allocating encoded and sorted data based on a reliability criterion.
9. A method in accordance with claim 7 or claim 8 employing a multisubcarrier communications mode, and wherein said step of allocating comprises allocating encoded and sorted data to subcarriers in the channel.
10. A method in accordance with claim 9 when appended to claim 8, wherein said step of allocating comprises allocation on the basis of reliability criteria for each said subcarriers. -
11. A method in accordance with claim 10 employing OFDM.
12. A method in accordance with claim 11 wherein step of allocating includes allocating encoded and sorted data associated with a relatively low degree variable node to a subearrier with relatively high reliability criterion.
13. A receiver for receiving encoded information transmitted across a communications channel, the receiver comprising reassembly means for reversing, in accordance with reversal information, a sort order of data encoded by means of an irregular low density parity check (LDPC) code, the sort order being with respect to the degree of variable nodes of the LDPC code, and decoding means for decoding the encoded data.
14. A method of decoding data received across a communications channel, the receiver comprising reversing, in accordance with reversal information, a sort order of data encoded by means of an irregular low density parity check (LDPC) code, the sort order being with respect to the degree of variable nodes of the LDPC code, and decoding the encoded data in accordance with the LDPC code.
15. A computer program product comprising processor executable steps which are operable to configure computerised communications apparatus to perform the method of any of claims 7 to 12 or claim 14.