GB2545467A - DC-Free / nyquist-free line encoders based on n-ask modulation - Google Patents

Abstract

The invention relates to line code for terahertz communication. A method for. encoding comprises: splitting 440 a bitstrearn 310 into substreams b1-bn, encoding the sub­strearns using line encoders by .selecting. a codeword from codewords based on a current running digital sum RDS or running alternative sum RAS; and performing ASK- modulation using a weighted combination of the selected codewords. In an embodiment of the invention: selecting the codewords by the line encoders takes: into account the weights used for modulating. For instance, the: current RDS or RAS is updated as a weighted combination of the individual DS or AS .of the selected codewords. The same assignment between weights and codewords is used for modulation and for updating. Optimum encoding is based on minimizing the sum of the current RDS or RAS with all the weighted DS or AS of the codewords selected in parallel.

Description

DC-FREE ./ NYQUIST-FREE UNi ENCODERS BASED OH N-ASK MODULATION

PlEtD Of- THE INDENT SON 1¾ present Invention relates to line encoding, In; particular to DC free encoding.

The invention has applications in wfrefesi communication systems, in particular, in totohefte; tolecommynicatipn systems, i,e. in systems based on radio waves belonging to frequency band from 100 6Hz up to 10 THz. This “terahertz” band; dvedege radio eteeftpiAagnefib rad®

Document AX, Widmer et P.A. Framsmk f & DC-Balanced, Partitioned·· Bfcdki 88/10B Transmission Code”, ISM Journal of Research and development, ml 27, Ho. 5, Sep. 1983. pp. 440-451} discloses one of the most popular DC free code. The proposed scheme, celled “88108” encodes each 8 bits of a digital data stream into 10 bits plus a status information” obtained from the previous encoding. The idea is to continuously balance between the ones (“1”) and the zeros (“0") so that at each limp the encoded bitstream has exactly the same number of ones arid the number of zeros, yyito su# continuous balancing, toe bitstream has a null mean value (it has no; contlnuousdomponentlvit to be “DC free”.

The “DC free" encoding modifies the sppgtted bitstream info an output bitstream having a bounded digital sum variation, Tf% dpt|i sum variation (DSV) is computed from the running digital sum (RDS) which ip defined as follows (after modulating the bitstream with 1 and -1. meaning that each one in the bitstream counts for 1 and each zero in the bitstream counts for -1);

I and J are integers and yri are the bits of the bitstream.

The running digital sum (RDS) also represents the difference between the number of ones and the number of zeros in the bitstream part considered (from ! to J).

The digital sum variation (DSV) is computed according to the ioilowhg formula:

i and J are of the bitstream.

For the encoding; the bitstream is subdivided lets blocks of S bits (one byte) sod presented lb tie encoder. Each byte is then subdivided into twesyb--blocks: one sub-block ;eompdslng for instance the five first bits and one sub-block comprising for instance the last three bits.

The first'lye bid are encoded using a ib/bbicodeithereby obtaining 6 bits. The Sb/6b enc^dteg is :;fshlea:0r:*ehdeho ai the encoder. A first iable contains codewords having more ones than zeros (or the same number of ones and zeros for some codewords in the table). A second table contains codewords having more zeros than ones (or the same number of ones and zeros for some codewords in the table).

Tp fries used for the 5b/6b encoding disclosed inabove-aied document AX PA. Frena&mk are fisted below. In the first tabie referred to as 'iabie.^Sb Jb _,RD...pius''! codewords have a nprnber of zeros at least equal to the number of ones., in the second tabie referred; to fp the codewords have a number of ones at least equal to the number of zeros One may note that the difference between ine number of ones and the number of zeros is either 0 or -2 for table.,5b. 8b _RDj>ius. while it lt$ either 2 or 8 for the table 6b„.8b_RD. minus. This difference within a Sihpie codeword CVV ^ {yj is named digital sum (DS);

Also, each time a codeword has a null DS in ons et the tables, the same codeword Is used at the same table position in the other tabie. ft results that the two tables Implement a pure b|ecfwe conversion function. ~(0 1 1 Ο O 0; table 5b 8b RD minus " [1 0 0 1 11; 1 0 0010; 011101; 0 10 010; 101101; 1 1 00 01; 1 1 0 00 T; 0 0 10 1 0: 1 1 0 1 0 1: 101001; 10100 1; 011001; 01100 1: 0 0 01 1 1; 1 1 1 000; 0 0 01 10; 1 1 1 0 0 1: 10010¾ 100101; 010 101; 010 101; 1 1 0 1 0 0; 1 1 0 1 0 0; 00 1101; 001101' 10 110 0; 10110 Q; 0 1 1 1 0 0; 0 1110 0; 1 0 1 0 0 0; 010111 1 00 100; 011011; 1 000 1 1; 1 0001 1; 0100 11; 0 10011; 110 0 1 0; 1 1 0 0 1 0; 0010 11; 0010 11; 101010; 101010; 011010; 01101 0: 0 001 0 1; 111010; 00 1 1 00;: 110011; 10 0110; 100110: 0101 1 Οι 0 1 0110: 00 1001; 110 110; 0 01110; 001110. 0 1 00 01; 101 11 0; 1000 01; 011110; 0101 Ο 01; 10101 1j;

Simtiariy, the Iasi three bits i» encoded using a 3b/4b code thereby obtaining 4 bits. The 3b/4b encoding is else based on tables or ^odeboaks* lypiabie at the encoder a first table Tabw^ob^b^RDjslys* and a second table ‘table.. 3b„4b„ RD ...minus

The tabies used for the 3b/4b encoding disposed In document AX Wkim&r ei P.A Franaszek are listed below, in table Ktabie_3b./1b_RDj3ius", the codewords have a number of zeros at (east equal to the number of ones, in tabid "table.. 3b_4b._RDjn!nusv, the codewords have a number of ones at least equal to the number of zeros. Again, the digital sum (OS) of each codeword in tabla...3b...4b.RDplus is either 0 or -2, while the digital sum fDS} of each codeword in tabfe.„3b _4b„RP^mlnus is either 2 or 0. table 3b 4b RDjsius «[0100; table 3b -[1011; 100 1; 1001; D10 1; 010 1; 00 11; 1100; 0010; 1101; 1 01 0; 1 0 1 0; 0 1 1 0; 0 110; 0 00 1; 1110; 100O|; 0111};

Let's now consider encoding the following byte: 0 0 0 0 1 0 0 1. The ehqddbd: word would be: - 10 01 0 0 1 0 1 1 when an milsal running disparity is equal: » 10 0100 0 1 0 0 when an initial running disparity is equal: 1

The running disparity is defined in document AX, Wldmer et PA, Franesmk as being the number of ones minus Ole number of zeros in the bitstream due to each encoded 0 bits fend respectively 4 bits) at the output of the Sb/8b (and respectively 3b/4b) encoder. The atoning disparity may be seen as the running digital sunt (RDS) of the bistream as being encoded (successively each 6-hit sub-block and then each 4-bit sub-block. At each new encoding, the the digital sum of the codeword selected for encoding or 4-bit long).

To iusbdlelhe^above) an encoded byte to be encoded (e.g. 0 0 0 0 1 0 0 1) is divided snto two parts, five first bits (i.e. 0 0 0 0 1) and three last bits (i.e. 0 0 1).

In an example, the running disparity is equal to +1, then table.,5b..6b.,RQ...p!us is used (instead of tabie^ObjSb^RD^minusI, The Sb/6b encoder 0 01 0 0(since0000 1 isencoded using entry0.16 In the Sb/OS cedeword Table 3 of the above-mentioned AX Widmnr etP,A FmnBsmk publication) The running disparity is updated, and Its new value is Oiijs equal Id: #4 T (2 ~ 4) - -1 (+1 is the previcus/initial running disparity and ψ * 4) is the running dispiriy or digital sum of the encoded word (1 0 0 1 0 05),

Since the updated running disparity Is pgup) itl -A, Mbp tableJSbj4b„RD_rninus is used for the 3h/4b encoding. The output for encoding 0 0 1 is equal to: 1 10 1, and the running disparity is updated again (in the example its new value is -1 * ( 3 -1)A T.f,;

For each byte to be encoded two codewords are possible:depending on the initial polarity (positive or negative) of the running disparity.

In document D. Y Young Khv et a!. (“White paper on me M8810 Line Code for IDGbE", proposal for 10 Gigabit Ethernet (IEEE 802.3)), document S. Kang at at CM8810. Dual mode Encoder/Decoder, and M8810 code generating method'') and document U$2005/0012646, the authors disclosed a new line coding scheme providing a spectrum hawing null power at the null frequency and a null power at the Nyquidf frequency (i.e. half the sampling frequency). Such a line encoder: is called "DC lee*' add ^iyquisi. free" line encoder. A code parameter sailed ‘running alternate sum" (RAS) Is considered for dulldlng the corresponding tiescode:

I and J are ifriegora, are the ills of the bistream (after modulation with 1 or -1), The RAS is the sum of the coded output symbols within an arbitrary internal between n»f and n-d, wliit alfernating iplar ry

Similarly todhe Ronnln| Qipal Sum, .a parameter named ''Alternative Sum;; V&riation” (ASV) Is equation;

if the ASV is Unite, then the bitstream fyj has a spectrum pompooentmeif· ditha Nyquist frequency.

Similar to the digital sum (DS), each ^awotll:ln"fha' fsbies; MfSoiti 9S*0;' alternate sum (AS) as being the sum of its bits (modulated with 1 and -1) with alternating polarity.

For example, let s consider the binary sequence “1100“. the OS plglfdl Sum} computation gives: DS ~ 1+ 1 -1 -1 « 0 1¾¾ AS | Aharda te Sum) is: AS - -1 + 1 4-1-1 -0

The OS and the AS pf fhd above binary sequence are hull; In additlOm shifting this binary sequence, as explained below, provides the same values for AS and I3S: *1001” -» DS = t * 1 ~ f a» 1 ~0 AS ~-1 -1 -5-1+1-0 ‘001T ~> 03--1-1 + 1+1-0 AS -+1-1-1 4 1 ~ 0 *0110" => DS = -1 + 1+1-1=0 AS- + 1 + 1-1-1-0

The line code MB24 has the Mmm codebook (1100,1001,0011,0110). The MB24 Sine code is thus a DC free and Nyqulst free line code. The M824encoding is defined by the table below;

Decoding a MB24 Sine code is straightforward. Indeed, one can see from the above febte;; that tie two bits in the middle of each codawcrd ccrrespond to the encoded word. The decoder architecture is very siropie.

Similar construction can be done in order to design the MB 12 and the !itS34 Sine obbaa.

The following table illustrates the encoding process for the MB 12 cede;

The MB12 encoder is initialized to state S1 and then according to the input lit: and me table above two output bits are selected and the current state of the encoder is updated,;

The following table illustrates the encoding process for me MB34 code;

The M834 encoder is initialized to state 81 and then according to three input bits and the table abdwd, four output hits are selected and the oufteat stafe of the encoder is updated.

An example of the MBS 10 is illustrated and deeply discussed in above-cited document 8, Kang at at and document US200S/0012640.

Other codes are disclosed in the prior art that deal with the DC free, or DC Tee arid Nyquisl free characteristics.

The code disclosed In document J. G, Kim ("An improved DC free Nyquisi free Error Control Line OddlT ICCS’94, November 1994. Singapore) uses the extended Hamming code for building a “DC free'1 and "Nyquist free” error correcting code.

The (4,8) extended Hemming code has some interesting characteristics since 12 among 16 of the codebook codewords have a null running digitei sum and a bull alternate digital sum as shown in the table below.

tesfeciid dooumentd) G. ife^/theiiideaiieio build a naw ead^ ^lairtsng: from a paiioffheidodeworde listed in the above table;. Twelve of these codewords have a nuJ DSixand'aaoiySblS:, Thus, starting fromfll^s^ it^i^/ipoda^ords, a table tosh fee each of the twelve codewords of the labia.. 1¾ new;: Mie- contains 144 codewords of 18 bits, each one having a; nullfrunniop digital sum end a null running alternate sum. A third tafeiejs. ^en.':p|g|8in^i>y aafedting only 126 codewords out of 144 codewords of the second table. With this third table haying 128 codewords, one can: encode a seven-bit word into a i δ-bit codeword belonging to the third table {since 27 ~ 128). The coding rate that is obtained is ?/16 ~ 0.4375,

To transmit the encoded words {e.g. codewords) as obtained above for bitstream, they are used to modulate the amplitude of a earner wave.

Amplitude-shift keying (ASK) is a form of amplitude modulation that represents digital data as variations in the amplitude of the carrier wave. ASK uses a finite number of amplitudes, each assigned to a unique pattern of binary digits. Usually, each value of the amplitude relates to an equal .number of bits. Each pattern of bits forms the symbol that is represented bv the amplitude.

One property of ASK modulation is to provide reliability levels) |e, different levels of sensitivity to noise, for the source bits.

Usually, ASK constellation Is generated in order to space regularly the symbols {carrier amplitudes) and the source bits are mapped on the symbols.

Thus, whatever the mapping between the source bits and the symbols, the sensitivity to noise of each source bit. Is not tine seme.

For instance, in case of 4ASK modulation using the constellation 125 of Figure 1a, the MSB (most significant bit of a two-bit input) Is less seneltiye; to noise than the LSB (less significant bit of the two-bit input). In what follows, the part of the te|aiitE··iiNai-fis.‘‘the most reliable bit or bits'1.

The constellation of Figure 1a corresponds to the foiiowing combination of the input bits: 2 " (2 * MSB-1 )+(2 * ISB-1). One may notice that this combination im weighted with a higher weight provided to MS8 compared to LSB (twice as big).

The difference in sensitivity to noise is side is ffeei idlfference between the weights assigned to MSB and LSB, and thus between the values of the moan disfanee: when considering ail symbols, the distance being the. g^FMbss^ihfi» .two epnMSk

For tire lSB, the mean distance between Ψ and *ff is 3a and for MSB, the mean distance:between O' and T is 4a.

SUMMARY OF THE INVENTION

The inventors have found that using the known line DC free encoders with an ASK modulation is not optimal The present invention lies within this context, seeking to provide an improved uri terms of 8ER/SNR performance) DC free line encoding scheme involving ASK modulation.

According to embodiments of the invention, there is provided a device for encoding a bitstream for transmission over a communication drain composing; a split unit for splitting a bitstream part into a plurality of slipstreams of uncoded words; a plurality of line encoders for respectively encoding the plurality of sub-streams of encoded words, wherein each of one or more line encoders are configured to select a codeword from codewords based on a current running digital sum and/or a asi-. a selected codeword from each line encoder as input and for^medoiaffriig#e amplitude of a carrfer wave as a function of a weighted combination of the plurality of received codewords; wherein the select ion of the codewords by the one or more line encoders takes Into account one or more weights used in the weighted combination for modulating.

By taking into account the weights used for modulation when selecting the codewords, the resulting running digital sum and/or running alternate sum are also shaped or designed based on the weights. Thus the behaviour of the bitstream over the communication channel >s taken into account when selecting the appropriate codewords for encoding the bitstream.

The inventors have observed some improvements so BER/'SNR performance of the resulting ling boding scheme* compared to ebnyebtiddoi tine encoders used with an ASK modulator. In addition, these embodiments of the invention remain simple to implement at both the encoder and the decoder.

Correspondingly, embodiments of the invention provide a method for encoding a bitstream for transmission over a communication chain comprising; splitting a bitstream part Into a plurality of sub-streams of uncoded words; encoding the plurality of sub-streams of uncoded words using a plurality of respective line encoders, wherein encoding an encoded word in each of one or more line encoders includes selecting a codeword from codewords based on a current running digital sum and/or a-:iraa^^ii^atng;allda^to»Ti modulating (he amplitude of a: carrier wave as a function of a weighted cornynation: of selected codewords for thaigiuraiify Offline :esieoder«f; wherein select?n|iilhe.;cpdewa!^sfl^:ihe one or more: line encoders takes Into, account onaiiorimiorerwelghts used: -v^i^iNt%^<ipsrf!¥ bslin^^n/ti3r:-rrM»^yfeHn^ : The method has the same advantages:as the device defined above. Opsone! features of embodiments of the Invention are defined in the appended claims. Some of these &amp; device, and can be transposed into method features dedicated to: mf. method according to embodiments of the invention. in embodiments, the current running digital sum and/or running alternate sum are weighted running digital sum and/or weighted runmng aitbrnaie sum applying the same weights to the codewords as the weighted combination for modulating. Thus, as the bitstream is being encoded, the influence of the amplitude-based modulation (end thus of the transmission) -s taken into account.

In embodiments, the plurality of line encoders is further configured to jointly 'select the codewords for the plurality of sub-streams, meaning that:the abjection of a codeword in a line encoder depends on another codeword selected in another line encoder and used in the same weighted combination for modulating.;

This provision makes it possible to reduce the digital sum variation {DSV} and/or alternate sum variation {ASV} compared to pure parallel line encoders. This Is because the choice of a codeword In on line encoder (and thus potential digital/alfemate sum) may impact the choice of the codeword m another line encoder.

In specific embodiments, one or more line encoders are further configured to select the respedlyai colewofd from codewords based on a digital sum and/or an alternate sum of at ieashone other codeword1 selected by another line encoder and; used ih the same weighted combination for modulating. This is a more specific design that ensures the digital sum variation (DSV) and/or alternate sum variation (ASV) to be beef controlled (and thus bounded|:

In other specific emiodimehts, ail the line encoders ere configured to select their respective codeword ffem: codewords based: on the cumgrtt running digital sum end/dri runr-sing alternate' sum when encoding an uncodeo word:©! the respective sub-stream.

In particular, each, of.ail the line encoders may be further configured to select its respective codeword from codewords based on digital sums and/or alternate sums of the other codewords selected by the other lino encoders and used in the same weighted combination for modulating.

The above provisions mean that the joint selection of the codewords is global from amongst the tine encoders. This approach achieves the bast result in terms of low digital sum variation (DSV) ano/or alternate sum variation (ASV), and thus in terms of BER/SNR performance of encoding:, in specific embodiments: selecting the respective codeword from codewords based on digital sum or sums end/or alternate sum or sums of one or more other codewords includes applying the same weight to each digital sum and/or alternate sum of a codeword as the weight applied to the same codeword in the weighted combination for modulating. The Impact of amplitude-based modulation Is thus fully and precisely taken into account when selecting the codewords to encode the bifttraamt

In other specific embodiments* the plurality of line encoders Is further configured to update the cunwit; running weighted running digital sum andfer weighted running alternate sum using digltal/sums and/or alternate some # fhe cgdewo^ In the plurality of line encoders and using respective weights from the weighted combination lor modulating. same weights are applied to the codewords when updating and when modulating, This is to mirror the behaviour of the bitstream when being transmitted.

In particular, the plurality of line encoders may be further configured to u#e the Idddied current running digital sum and/or running alternate sum to encode a next bitstream part, in other words, this means that the current running digital sum and/or running alternate sum (before updating) may be a running digital sum and/or a running alternate sum obtained from the encoding of a previous bitstream part (and thus updated when encoding the previous bitstream part). in specific embodiment, the plurality of line encoders is further configured io select the codewords for the line encoders by minimizing the following formula: abs{RS+c;.S(CW,)*.. *a.„S(CWft)) wherein sbs(x) is lie absolute value of x; CW* is a codeword for line encoder i; ,01=) areilt^v^^ codewords fCW*in the weighted combination for modulating: RS is the current running digital sum and/or fanning atfsmafesum; and SCOW) Is a digital sum and/or atfernafe sum of codeword CW.

Such approach provides an optima! selection ;pf: rxteewords given the ampiltudeA>ased modulation used:,f he optimization is to be cohsiderad psminimidng thed|iM sum variation (DSV) andior alternate sum variation (ASV) for the bitstream.

Iff some embodiments, when encoding a first encoded word, a first line encoder is further configured to: determine at least one first nominal codeword associated with the first encoded word in a first codebook. and determine: a first substitute codeword from a second code book; select one of said at least: oneslirst comInal code word and first substitute codeword/ based onlhetourrehlrurming digital sum and/or running:alternate sum;

Updhte: thei corrontironning digital sum and/or mnning; aiternate: sum based on a:df|ltal :Sumiand/or:en alternate sum of the selected codcwvord: and iutput tbesaeleetedi eedeword as a first encoded word.

Ip; particular embodiments:,: when encoding a second uncoded word, a second line encoder of the plurality is configured to; determine at least one second nominal (tedeworf associated with the second uncoded word in another first codebook, and determine a second oobstltute codeword from another second codebook; select one of said at least one second nominal codeword and second substitute codeword, based on the update# digital: sum ahi/or tphhihg: alternate sum; ah$ output the selected codeword as a second encoded word.

The first encoded word and second encoded word (ar.d thus corresponding codewords) may be input to the amplitude modulation unit in some embodiments, the updating of the current running digital sum and/or running alternate sum is based on the weight applied to the selected codeword in the weighted combination for modulating.

According to a specific feature, the current running digital sum and/or n.rnning aternate sum is of an encoded fciisraeam comprising the selected codewords for the hitsfaam parts upIda current bitstream part,

According toanother^spaclio teature, the nomirmi ooiaword is a Hamming codeword.

According to yet another specife featere, the plurality of line encoders may t-e further configured to substitute codeword is outputted.

For exlfhpla, said encoded word further ooinprisds a parity hif code and; said parity bit code is indicative of whether a nominal codeword or a substitute codeword is outputted..

Tbifpfd#, when ellng, IdFexarapie, the digital sum and/or alternate sum values, the action of substitution can adjust one of the running sums without disturbing the other (no trade-off is required). This is the case, for example, since the digital sum and the alternate sum cannot be non-nuli at the same time.

For example, the substitute codeword has a negative digital sum and/or a negative alternate sum when the digital sum and/or the alternate sum of a previous codeword used for encoding a previous input word is positive;

According to embodiments, the substitute codeword has a positive digital sum and/or a positive aiternate sum when the digitai sum and/or the alternate sum of a previouscodewordused teencoding a previous input word is negative.

According to embodiments, said substitute codeword has a digital sum and/or an aiternate sum thatkeeps the current mnning di^faf sum and/or the running aiternate sum null.

For example, said subsfituto codeword has a same size as the nominal codeword; jin some embodiments, encoding an encoded word: In a line encoder includes determining available codewords froth two or more codebooks based on the value of the encoded word (In tact each codedook may associate codewords with uncoded word values, thereby making such determining quite easy), and selecting a codeword from codewords in the line encoder consists in selecting a codeword from the determined available codewords. This is usually the selection between a nominal codeword and an alternate codeword, depending on a current running digital sum and/or alternate sum. in some embodiments, each line encoder is a DC-iree and/or Nyquist-free |ine:'ehppder.

Embodiments of the invention also provide a ηαη-transitory computer-readable· medium storing a program which, when executed by a microprocessor or computer system in a device, causes the device to perform the method as defined above.

The non-iransiiory computer-readsble medium may have fiafures and advantages that are analogous to those set out above and below in relation to the methods and node devices.

Other embodiments of the Invented also pr$Mdb'-a hiethod;fbr erreodingsa bitstream for transmission over a ioommynicatiop: chain: substabteiiy as herein described with reference to......and as shown In, Figure 4 or 7 of the accompanying drawings.

At least parts of the methods according to the invention may be computer implemented. Accordingly, the present invention may false the form of an entirely hardware embodiment.; an entirely software embodiment (including firmware, resident software, micro-code, etc ) or an embodiment combining software and hardware aspects that may all generally be referred"*» IsysfettS.·

Furthermore, the present invention may take the form of a computer program product:: embodied in any tangible medium of expression having computer usable program code embodied m the medium.

Since the present Invention can be implemented in software, the present invention can be embodied as computer readable gode for provision to 1.: programmable apparatus on any suitable carrier medium. A tangible carrier medium may comprise a storage medium such as a floppy disk, a CD-ROM. a hard disk drive, a magnetic tape device or a solid state memory device and live like. A ifabetent earner medium may Include a signal such as # eiedfteet signal, an.....electronic signal, an optical signal, an acoustic signal, a magnetic signal or an electromagnetic signal, e.g. a microwave or RF signal.

BRIEF OESCRIBTIOH OF THE DRAWINGS

Features and advantages of the invention will become apparent From the following description of non-limiting exemplary embodiments, with reference: fp $|§ appended drawings, in which:

Figure 1a schematically illustrates the architecture: of the conventional transmitter;

Figure 1b schematically Illustrates the architecture of a transmitter according to co-pending application GB 1417280.3;

Figure 2a illustrates an exemplary architecture of the transmitter of Figure lb, i<e. :acoordir;g to ca-persdmg afpilealon GB 1417280,3;

Figure. illustrates an exemplat^'dndvdchii^afeiar^ifeobire^&amp;co^i^fd;:.: embodiments of the invention;

Figure 3 schematically Illustrates an encoding device according to embodiments of the Invention, when using N - ASK modulation;

Figure 4 illustrates, using a flowchart, general steps of a joint encoding according fo embodiments of the invention, when using N-ASK modulation;

Figure S schematically illustrates a decoding device. t$: $Sl· encoding device of Figure 3, when using N-ASK modulation;

Figure 8 illustrates, using a flowchart, general |ί|$ίφ ϋ algorithm whan using N-ASK modulation;

Figure ? illustrates, using g flowchart, general steps of an encoding algorithm when using 4-ASK modulation according: to specific embodiments df the: invention;

Rgure B llusfrafes,; using a flowchart, general steps of a decoding algorithm correspndfhg. toliliureil;: when using 4-ASK modulation; and

Figure ii; schematlcaly illustrates a genera! architecture of a device; according fo enthpliments,::

DETAILED DESCRIPTION OF THE INVENTION A OC-free communication system uses a DC^reeltoe coding scheme In order to suppress the DC comoonent of bltsN^Ta',dat^.'^Mi:ddd^ not travel earnestly: la the communication chain and'or In Ihe transceivers.

The DC-free line coding maintains a running digital sum (RDS) value bounded within a digital sum variation (DSV) and uses some known technics like lid®, 048068, MB810 etc. to keep the bounding. These technics are already imgiemenfed in some staf||a?|le|POf Impress, GigahitEfheroet, DVI, HDMI, etc).

In addition, an ECO (Error Correction Coding) feature is applied to the bitstream data in order to perform the error correction. Hence, a combination of a line coding with an error correction coding is commonly used th order fo have both error correction and DC free capabilities.

The present Invention particularly regards the PGTree line coding when ysei in cooperation with an amplitude-based modulation, Eor instance, embodiments of the present Invention provide an improved (in terms of BER/SNR performance fhadte to a decreased DSV) DC-free line coding by performing joint encoding of different binary sub-streams before applying an amplitude shift keying (ASK) modulation.

Amplitude-based modulation, such as n-ASN modulation, is usually expressed as a Ihear comblnatkjn of οΛ the weight or coefficient assigned to each sub-stream represents the qualify or reliability ievei of the corresponding sub-

Stream.: Flgb reliability level is equivalent to high coefficient. and low reliability ievel is equivalent1 to low coefficient. Thus both representations are considered equivalent herein. Also, top :ooefilpiants and: weights::are considered equivalent terms, arid they can be used infemhsngaafely· Same equivalence applies between the terms quality and TOiiai:i%l

The variation of the quality of fmnsmissfe^ of bits belonging to the same bitstream may be due to characteristics of the modulation used for traosmlffng the bitstr^mdfW'l^feisqe·,· in case of a pure amplitude modulation wherein the modulated signal is a linear combination of the inputs, the Input bits which are multiplied by the highest weights are the less thus allocated to the bits encoded with the line encoder that better matches energy around nuii frequencies).

Other types of modulation (other than pure amplitude modulations) may also have uneven qualify of transmission for each bit

The present invention has thus application within this context.

For example, when encoding the bitstream with a 4-ASK modulation (2 bits symbol), 1 bit of the input symbol is more robustly transmitted than the other. According to embodiments of the present invention, selecting codewords by the line encoders takes into account the weights used in the weighted combination for modulating. According to embodiments of the present Invention, the current running digital sum and/or running alternate sum (used to select the correct codeword from two or more codewords) are weighted running digital sum and/or weighted running alternate sum applying the same weights to the codewords as fhei wfeighted combination for modulating.

This makes it possible to use the contributions of the amplitude-based modulation scheme to drastically reduce the DSV of the modulated bitstream, and thus to improve BER/SNR of the transmitted bitstream.

The ABICmodulation maybe modeled as a weighted linear combination of sub-streams.

In the figures and in wMf follows., identical elements and steps are designated with a same numerical reference.

With reference to Figures la and 1b, the amhitepfuresidf a conventional transmitter and a transmitter disclosed in co-pending application G8 141T280.3 ana described: in order to better assess the new aspect according to the present embodiments. fmm la schematically illustrates the architecture of the conventional frah§^!ttir.

Bits are provided by a data source 105 and are scrambled in a scrambler 110 in order to ieddbmiee 1! -pi Ί’ bits before being input to a line coder 116 which carries out an encoding process*

Bor examoie. the fine coder may carry out an encoding that removes the DC component (i.e. the mean value of the amplitude of the signal) of the incoming stream, l.e. the line coder is a DC-free coder The BhIOb or MB 12 encoding processes perform such removal for example.

In such encoding processes, encoding an unccced word m the line coder includes selecting a codeword from codewords based on e current running digital sum and/or a current running alternate sum, :':ΐΤΙΐϊ&amp;· -iSiSfNiieici^n.· a it to an ASK modulator lii^ich generates a; modulated wave based: or? the input bit values. For instance, for a 4-ASK moddlafe. bits e;re grouppifeiby ilisio.

The curves of graphs: 160 end 135 respectively snow examples of Input data and the: ralafed generated waveform when using a 4-ASK modulator and the cdoiteltatlbh $$$..

Figure 1b scharhotlosiiy iiustrates the erchffeclbfi of a transmitter according to ec-pending application GB 1417280.3.

As for the conventional transmitter, the bits provided by the dafi Sobrce 105 are scrambled in the scrambler 110 before being Input to a Spiff: module 160. in split module 188, I he bits ere split into two sub-streams, A first sub-stream is named “Dbf and a second sub-stream named "Die.

The split is performed based on the coding rates of the two PG-free line anopefs 170 and 180 which encode the two sub-streams.

Source bits are grouped by blocks. Considering a size "tf of the source bits block, "dbf is the size of the clocks of the first sub-,stream and Ait" the size of the blocks of the second sub-stream

The relationship between these block sizes is as follows: d~ dfvKdl, dt~ dM * (CR2/CR1}, where CR1 is the coding rate of the first line encoder and CR2 the coding rate of the second line encoder.

The first sub-stream DM is input to the first line encoder 170 and the second sub-stream OL Is Input to the second line encoder lilt Outputs of both line encoders are nonnested to the inputs: of a 4-ASK modulator lib, For.e^asihpfet output oi the:ilrstline encoder iT|:::ii':<oofPOoted ^siiiie MSB Input of modulator 191 and the output oi' the second line encoder liltis eonpeeted to theibBB input of modulator

In this coding schemeK;the. plurality of: sph^sireams of: encoded^ words'm encoded using s plurality of respective line rnmxjfam* ^-j^rt!he :c^vention§lf»!Sdd^?| encoding m uncoded word in each of one or more line encoders includes selecting: a codeword ffdm codewords based on a current running digital sum and/ορ a current:; running alternate sum.

The acronyms "MSB" and “ISO5' di hot refer issued from the data source. Instead, the MSB Inpdt df the 4-ASK mcbuidtfrr for the like) receives the bits which are the most reliable (less corrupted during the communication) as compared to the bits received by the LSB input

Input bits fed to the 4-ASK modu|dipr lid a# used to gpneffte: a modulated wave, usually by modulating the amplitude of a carrier wave as a function of a weighted combination of the input bits, i.e. of the codewords selected for the plurality of line encoders 1?0 and 180.

Figure 2a illustrates an exemplary architecture of the transmitter of Figure 1fe, he. according to co-pending application GB 141/280.3. in other words,, block 210 Illustrates the DC-free encoding as of Figure 1b.

The 4-ASK modulation is a linear combination of the bits encoded by the DC-free line coders 225 and 230. if both the MSB end the t.SB sub-streams are DC-free then the bitstream obtained by performing the linear combination (for modulation) is also DC free. This m because the digital sum of one block of output codewords (i,e. symbols) Is also a linear combination of the digital sum of a bloc* of MSB codewordi andithe digital sum of a block of 188 codewords;

Due to weighted ^pp^bioabon-tir' mcNdulilibn (assuming that If uses a weight equal to 2 foriteMS^'-and a LSB), tbs digital sum for the output modulated symbols is equal ίο 2*X+Y where X Is the digital sum of the MSB block and Y is the digital sum of the LSB block.

However, the DC-free encodings by line coders 228 and 23! are; pedormed Independently because they are parallelized. In particular, a respective current running digital sum (RDS) Is kept up-to-date for each line coder, it means that the maximum running digital sum reaches '2*DS\W+DSVusb< where DSV«§&amp;, Is tlm digital surd var-ation (DSV) for line coder feeding MSB input of ASK rnoduiat©r^|§:-|f|SeiiP#f:f^: for instance) and DSVtSB is the digital sum variation (DSV) for line coder feeding LS8 input of ASK modulator 235 (lino coder 230 for instance),

To substantially reduce such digital sum variation (DSV) for the whole S8* of line coders, embodiments of the invention provides DC-free encodings as schematically shown in Figure 2b.

The various DC-free iine encodings (for MSB and LSB docks in case of 4-ASK for instance) are performed jointly in block 250 when feeding the moduletor for the same output modulated symbol. Joint encoding means that the codewords for the plurality of sabr^^dd¥S-:^^:jdiNii^i:igie^®6^dlJ^ldtib©rvve5eBidis^:#i^-:^i©fection of a codeword in a line: encoder depends on ianofher codewoird eeletdeil lhianother line: encoder and: used In: the: eeme weighted Mmbination fer modulating (for the same outpufirnodulated symbol).

Joint encoding: makes It possible to reduce the digital sum al the output modulated symbols, it is beloved that: the smallest digitafeum for the output symbols is 2*DSVMS8*DSVl5s8. a conseguenoe io.^bb^fc^,€|:' the invention is that the power of the s||nalat; the null frequency of encoded bitstream 280 Is lower than the power of the signal at the puli^ h'eguehoysof encoded bitstream 240 (without the invention).

Figure 3 schematically illustrates an encoding device 300 according to embodiments of the invention, when using N~A.SK modulation (N being an integer power of two).

Here, an N-ASK modulation block 360 is considered which performs a linear combination of input symbols or codewords $·,, , s,„ using respective weights

Ci.....,cf;< to provide output symbol $ forming the encoded and modulated bitstream 370. The latter can then be transmitted over a communication channel using appropriate transmitter. To feed the n inputs, n~iog2(N) parallel DC-free line encoders are used, respectively 330,, 330:>, .,, 330,,. in operation, the input bitstream 310 is first split into several sub~bifsireams of encoded words by splitter 320, each being input to a Side encoder 330. In the example of the Figure, input bitstream |h}i «Is spilt Into unocsded bifsbr eaeh fid to hie input of a respective line encoder 330;.

Then the line encoders 330 perform joint DC-free encoding to obtain encoded words, here codewords or symbols %

As provided m embodiments '.of Me invention, such joint encoding to select codewords by the fine encoders takes IbSS: account &amp;m or more;:wef|fMi§-used in lies weighted comtMnatiom

Aigeeefailimethod for such joint encoding Is now described with reference to Figure 4.

Tie genera! method starts at stop 410, end step 420 is executed that initiaisd'es the running digital sum RDS value for the modulated bitstream 370 to aarm Thanks to step 470 described below, the :RDS ¥ajde:a$i^evcke ever lme,::then when the joint encoding is iteratively performed on successive input bitstream parts ffhanks to loop from step 470 to slap 430 described belpsdl, si new ourrept: RDS value? h :.Af step 400, an input bitstream 310 is receiipd; which M ^pNt at step 446 into n parallel sub-bitsipams: b.. Each sub-bitstream b= Is made of uncoded words or 'blocks’' B:.

At this stage of the process, each lift© encbdpf :33¾ haa-fil OWh uncbddd block B, to process.

Next step 450 is performed by each line encoder 33¾ based on ifs input hipek of bits B;, Step 460 consists for the line encoder in determining available codewords from two or more aodebooks based on: the dalop

Usually, two tabies labteJ^DjjJusf (positive; digital iprn labia) and ‘fable ,.RPjesous'1 (negative digital sum table) am implemented in conventional She encoders, wherein one table is selected based on the current RDB, !ahle_RDjdbs, used in case of positive RDS, contains codey^&amp;g[#:.'?«iyh, no?· dr su$

values; while table., RD_minus, used in case of negative RDS, includes codewords with nul or positive digital sum values. The tables era also known as “oodabooksT

Step 450 uses the same tables to determine for instance two codeworha CWupius and CWijninus from respectively iab!e„RD_pius and tabie_RD_minus. This avoids substantiaiiy?modifying the conventional line encoders. ^hile all the line encoders 330, have determined -their ovwtrli^y^iiis and DW^minus for current Bu they haw ©acM^issJecf datermihed available codewords. itnce: the encoding is jointly performed between ail the line encoders, each of them has to select; Is respective codeword from the two determined codewords: based on digital sums of the other codewords selected by trie other line encoders:, isra the codewords considered are only those determined at the same iterators of step 43¾

To ir# account the ASKmodulatlon in a proper way, the same weight c, is applied to each digital sum of a codeword GW, as the weight q applied to the same codeword (or equivalent symbol $;) In theweighted^comhiqaSon Jt ptodi*{stmg.

Furthermore, to optimize (close to i^tq| the ROB q| the imgddlated bitstream 370, the selection Is also based on the currant ROB,

At step 400, an embodiment of such joint encoding consists for the plurality of lino: encoders to seiect the codewords for the line encoders by minimizing the following formula: abs(RDS^Oi.OS(Cvy,)+. ..±c,.OS(C\Nn}) wherein absfx) is the absolute value of % DWGif a codeword for fine encoder: 330; selected from CW;j3ius and CWjjninus; (c, ,..,..0,,} are the weights applied respectively to codewords {CV\h„..,CWt} in the weighted combination for modulating; RQS is the current ranring digital suns; and DS(CW) is a digital sum of codeword CW.

As a result of step 460. π codewords {CW}s are obtained, which are the n Inputs for the N--ASK modulator 360 to generate an output moduiafed symboi B for currentencoded b&amp;treami370.

Thanks to this minimization, the digital suns of the output modulated symbol S is bbihdOd by lower values compared to known techniques.

Next to step 460, the RDS :$ updated at step 470 as a weighted running digital sum using the digital sums DS(Wt) of the codewords selected in the plurality of line encoders 330; and using respective weights c, from the weighted combination for modulating: RPSi:«mpS^P?,DS(CW5^.,:,*H;n.DS(CWrt},

As mentioned above, thanks to the loop from step 470 to step 430 to considers next pert in the input bitstream 310, the updated RDS is used to encode a next bitstream part.

Figure 5 schematically illustrates a corresponding decoding device 500 i e. when using N--.ASK modulation, and Figure 6 illustrates, using a flowchart, general steps of a decoding algorithm when using N-ASK modulation.

The process initiates at step 610. N-ASK demodulator 520 receives moduiafed bitstream 510 as Input., block;ialed bitstream Sill· Is for instance bitstream 370 of Figure 3 and includes successive subparts S, A subpart S is thus feci to N-ASR modulator 520 at step 620 god unmodulated, at step 630, into n symbols {S;Ws „ using conventional tectonics, with η~ίορ2(Ν}.

Each spobbl s, is fed into one corresponding DC-free line decoder 530( from:: niparaile! DGrtree line decoder si ttoo DC-free. decoding in each line decoder 530; is conventional at step 6-0. thanks to the overall bijeetivity of the tables, each symbol s. has only one corresponding.- This “decoded" word 8; is thud : output as;s part of corresponding sub-bitstream b; feedinp multipiexer 560,

The rv parallel sub-bitstreams : h,are then malbpfejred alstep 650 to provide the output bitstream 570 corresponding to the decoder bitstream,

Eigpres 7 and 6 illustrate a specific embodiment of the invention In case of using 4-ASK modulation. Joint encoding process of Figure 7 is for instance implemented in line encoder 260 of Figure 2b. yore particularly Ftg«M 7 :liustretes a 4-A8K |o|n|: encoding method 700 based on the extended Hamming (4^3} Oddi,

The method 700 starts at step 701, the running digital sum RDS value for One modulated bitstream: similarly to Figure 4. the loop from steps 725 , 726, 727 or 728 as cescribed below promdes an updated RDS to encode a next bitstream part.

As shown in line encoder 250 of Figure 2b, the input bitstream ;s provided to the two DC-free line encoder 265 end 270 through two sub-streams hDM” and lib”. The proposed method first performs the line encoding of the input block intended to feed the MSB Input of 4-ASK modulator 275. For instance, this inpui block comes mom sub-stream DM, and is named "MSB block'’ in the following description.

The proposed method then performs the line encoding of the input block Intended to feed the LS8 input of 4-ASK modulator 275. For instance, this input block: comes from sub-stream DL and is earned 1.88 block” in the following description.

The line encoding of MSB block corresponds to steps 703-715, it. is followed by the line encoding of LS8 block corresponding to steps 716-728.

The line encoding of MSB block includes: determining^ atleast one first nominal codeword associated with the MSB block in a first codebook {stored ih the line encoder considered), determining a first substitute codeword from a second eodebobk; selecting one of said at least one first nominal codeword and first substitute codeword^ based on the current RDS; updating the current RDS based on a digital sum andfor an alternate sum of the selected codeword The updating preferably is preferably based on the weight applied to the selected codeword in the weighted combination for modulating: for Instance the D8 of the selected codeword is weighted using said weight before being added to the current ROS. The line encoding of MSB block ό£^·$Ι§. outputting the selected codeword as an encoded word encoding the MSB block,

The line encoding of LSB block includes quite similar operations: determining at least one second nominal codeword associated with the LSB block in another first codebook (stored -n the line encoder considered for encoding tie: LSB block), determining a second substitute cxxfewonii frpft d^ebook;. select one o? said at least one second nominal codeword and second substitute codeword, based on the updated RDS; and outputting the selected codeword as an encoded word encoding the LSB block (Si ddbrsei the R.DS may be updated anew during the line encoding of LSB block In order to have an up-to-date ROS when encoding a: new part of the Input bitstream (be: ne>d MSB and LSB blocks)

To be more precise, the Sine encoding of MSB block starts at steps 703 and 704 which check whether or not the MSB block equals QOQQ or 1111 respectively, if the four bits of MSB block are different from the two tested sequences: (1000 and 1111). MSB block is encoded at step 715 using She extended hamming code presented .earlier in the “background of the invention'’ section the codeword CW«s8 obtained is 8-bit long including a parity bit, and ltd digital sum DS(CWmss) is equal to aero. Any updating of the current ROS iwith DS(CW)wss) thus does not modify RDS. if the MSB block Is equal to 0000 (output “no” in test 708). two codewords are available. A first available codeword is the nominal codeword 00000000 associated with 0000 In the Hamming (4,8) code. A second available codeword is a substitute codeword inverting the value of the Hemming (4,7) code (i,e, 1111111 instead of 0000000) but keeping the parity bit (0) of the extended code: 11111110.

One may note that the parity bit code Is Indicative of whether a nominal codeword or a substitute codeword la selected and thus outputted

Ibis inversion thus makes it possible to have two Hamming cede tables:: teeconveni^ (4,8) code and the inverted Hamming (4,8) code (which keep unchanfed the parity bis).

The selection between the first and second avdiabie codewords is thus made based on the current RDS. Test 705 checks whether the RDS value is positive (including zero}, in which mm 'W® nominal) codeword 08000000 is sd^fed:: m C\NmH at step 70S, if the RDS value is strictly negative, the second(aodsohstltyfe) codeword 11111110 is selected as 077¾ a t step 707.

Similarly, fi lie MSBhiock is aqua! io 1111 (output “no* in tost 704), two iOhdewpriteareayatiabte, A tifMs^ifeiie ^dewonl is thenominai codeword 11111111 a^ootated With 1111 in tie Hamming (4,8) code. A seoendavaliabie codeword is a substitute codeword inverting tie value; of the Hamming (4,7) coda (i.e. 0000000; instead of 1111111) but keeping the parity hit (1) of the extended coda: 00000001.

The selection between the first and second avaiialie codewords is thus made based on the current RDS. Test 708 checks whether the RDS value is positive (including aero), case the second (and substitute) codeword 00000001 is sat|bt#d as CWfcssa at step 710, If the RDS vaiue is strictly negative, the first (and nominal) codeword 11111111 is selected as CVVW< at step 709,

This selection scheme thu# sal#i1^ digifalsum in case the current RDS is positive, and vice verse. This is to reduce the vaiue of the RDS to be close to 0 to satisfy the DC free constraint.

Next to any of selecting steps 707, 70S, 709 and 710, th&amp;eyrmnfRDS is updated with the weighted digital sum of the detected same as the corresponding weight used by the 4-ASK modulator, he. weight equate 2; for the MSB block

Thus following step 707 where a cdbewhfd witboDS-Tdi &amp; detected, step" 711 updates the current RDS with 2*8~*12.

Following step 708 where a codeword with DS--8 Is selected, step 712 updates Ihecurreol RDS withO-S-ie.

Following step 700 where a codeword with DS~*-8 is selected, step 713 updates the current RDS with 2*S~+10.

Following step 710 where escedeword with 8f§e-8 is selected;, step 714 updates the current RDS with 2*--8~-12

This Is the end of line encoding of the MS8 block, it is foliowed by the line encoding of the L.S8 block, wherein the RDS updated wth is telen tntefocsttetl A similar algorithm as to tire one for the MSB block is pedVxmed for the USB block. The line encoding of LS8 block starts at steps 716 and 717 whether or not the LSB block equals 0000 or 1111 respectively. if the four bits of LSB blbdl 'φ.fsp; the two tested sequences ($0QO::agd 1111), LSB block is encoded at step 720 using the extended hemrethl obde presented earlier in the “background of the invention” section.

The codeword CW;.<..S obtained is 8-bit long including a parity bit, ami its digital sum D$(€VYLSS) is equal to zero. Any updating of the current ROS with OS(CvVuSij) thus does not modify RDS.

if the LSB block is equal to 0000 (output l,no:' in test 710;, two codewords are available, A first 'available codeword is Ibetheminal codeword 00000000 associated: with 0000 in the Hamming (4,8} coda, A second available codeword is a substitute codewadi: tte-Yaibe of tbs Hamming :(1 ,7} code (>.e. 1111111 instead M 0000000} butfteeping thaiptiy bit (0) oflis extended code. 11111110.

Again, this Inversion makes It psslbfato have two Hammi:h|vbode tailee; The selection between the first and: second available codewords is thus made based an the currerilii; Test fit ebaekstwhether the current (and updated) RDS value is positive (secluding zero), in wh-ch case the first (end nominal): codeword 00000000 Is selected as CVVl3B at Step 722 If the current RDS value is strictly negative, the second (and substitute) codeword 11111110 is selected 721.

Similarly, if the LSB block is equal to 1111 (output “no* In test 71?}.. two codewords are available. A first available codeword is the nominal codeword 111111ft associated with Ti l l in the Hamming (4,8} Code, A second availsbie codeword Is 1 substitute codeword Inverting the value of the Hamming (4,7) code (i.e 0000000 Instead of 1111111 }but keeping the parity bit (1} of the: extended code: 00000001,

The selection between the first and second available codewords is thus made based on the current ROS. Test 719 checks whether the RDS value is positive (including zero}, in which case the second (and substitute} codeword 00000001 is selected as CW^ at step 724, If the RDS value is strictly negative, the first (and nominal) codeword 11111111 is selected as CVV^s at step 723.

This selection; s#ieme fhys selects codewords with negative digltai sum in case the current ROS is positive, and vice versa. This is lo reduce the value dffhe RiS to he close to 0 lb satisfy the DC free eonstre Ini

Next Ip any of selecting steps 721, 722, 723 and 724, the current: RDS is updated with the weighted digital sum of the: selected codeword. The weight used is the same as the corresponding weight used by the 4-ASK modulator, i.e. weight equals 1 for the LSB block.

Thus following step 721 wharfi.a codeword With: DS~*8 ; 72t:updates fhecurrent RDS with 1*6«-*#,

Following step 722 where a codeword with OS~-8 is selected, step 726 upda^^ IHe current RDS with iMtesJb

Following slop 722 where: a codeword with DS~*8 la: selected, step :227 spdates the oerreot RDS with ΐ'8~*&amp;,

Fbliowing step 724 where a icodeword with DS--6 fe selected, step 721 updates the current RDS with 1*-8~vS,

This Isrthecepd of line encoding ofthe LSB block. The method may thus :fpgp hack to step 703 when considering s next part of the input bitstream, ie, next 4-Pit MSS and LS8 blocks.

As mentioned above, CWmss and CWis® are Input: to the: 4-ASi· modulator lb produce a modulated symbol for the encoded and modulate bltetfeam.

Figure 8 Illustrates, using a flowchart, general steps of a 4-ASK decoding .method. It is for instance adapted to decode a bitstream g&amp;mr^0^·#%mettod'.of figure 7.

Note that the 4-ASK decoding method does not require knowipdge of tie RDS value. This is because, the decoder starts by receiving an encoded MSfTblgoiA noted E8s*5g, and an encoded LSB block, noted E8l8s. from the 4-ASK dernoduiatpr {after demodulation of the received encoded and modulated bitstream) at step 602.

Step 803 performs a conventional Hamming decoding in order to correct possible errors in the bitstream considered. The Hamming decoding is applied only on the first seven bits of ESmss and EBtss, meaning that the parity bit EB{7) Is not modified The results of the Hemming decoding Is €A%so and CWtsa knowing that: CfWrfO- FB^b(7) end CWLS8{7}* EB>SS(7).

Again the decoding process first decodes the MSB part (CWmss) through; steps 604 to 812, and then decodes the LSB part (CVVLSS) through steps 813 to 821.

Tests 804 and 805 check whether or not the first seven bits of CV'4i3£1 are; equal to 0000000 or 1111111. If it is the case, tests 806 and 807 check whether the parity bit Dvy^SB{7) is equai fr> 1 or 0, in. order to determine if the encoded bite were inverted or^pot which. if the parity bit CW^i;b(7) is different from all the bits €Vt>W{0:8), the encoded bis were Inverted (l.e. the substitute codeword was selected by the encoder). i:: suoh::a oese:{output:::%o:· in either step 808 or 837), CW^^0:6} is inverted again and the resulting ONim (Having oniy Os or isj is used to retrieve the sbfrqbt: 4T}it IIBB bieek from the Hamming (4,6) 0}¾ This is step 810 Pr8l2. jn case the parity hit CVVWT?} is the same as ail the bits CW^Sb(0'6), no inversion was performed by the encoder. Thus, CW&amp;iSB (having only Os or 1s) is used to retueve the correct:4-blt yS8 block from the Hamming (4,8) code. This is |tqp i09 or 811. if the first seven bits of CWuss are not equal to 0000000 or 1111111 (output xm* attest 606), a conventional Hamming (4.8) decoding is MSB biock is retrieved from the Hamming (4,6) code at step 808.

This is the end of iine decoding of the MSB pail CW^g. it is followed by the line decoding of the LSB part CWt$a< still regardless of any RDS value. A similar1 algorithm: as to the one for the MSB pari GVv^s is performed lor the LSB pari OWf..?*.

Tests 813 and 614 cbedfe «ytier or noli:the first seven Oils of 0¾% are equal To; 0000000 or limn, if it is the ease, tests 816 and 617 chaeferwhel;her::::tte .parity bit CW^aC?) is equal to 1 or C, in order to determine I the encoded hits were Inverted or not which, if the parity bit CWlsbC?) is different from ail the bits CWtss(Q;6)i the encoded bits were inverted (Ιο,^ iie soyitdte codewordn^s sile#dd by the encoder). In such a case (output "no" in either step 818 or 81?), €VvVsg(0:8} Is inverted again and the resulting CWtss (having only Os or 1s) Is used to retrieve the correct 4-bit LSB block from iheHamming (4,8) code. This is step 819 or 821.

In base the parity bit CWW?) is hie same as all th$ inyerslob :^s parfermed by the encoder. Thus, CW^S8 (having only tls dr Is) is used to retrieve the correct 4~bit LSB block from the Hamming (4,8) code. This is step 618 or 820.

If the first seven bits of CW^a are not equal to 0000000 or 1111111 (output ’’no" at test 814), a conventional Hamming (4,8) decoding is performed, A correct 4-bit LSB block is retrieved from the Hamming (4,8) code at step 815,

Ail the examples above are mainly based on providing Pi-free encodiing and decoding. This is done by focusing on the value of the weighted rurmingdigif^ sum,: in variants, hiyquist-free encoding anti decoding may do sought, in which case the above algorithms can be implemented in a similar way while considenng a weighted running alternate sum (RAS) Instead of the weighted RDS, in diet: case, the codewords may be searched In tables buiit for alternate sums, for instance a hableJ^Ajslus'1 table (positive alternate sum tabiei ooetaihlnf: codewords with negative or null alternate sums, and; a ^bleJxA^minus* table (negative alternate sum table) containing doitdWQrds.with positive or nul alternate sums, psefonrsuia to minimize at step 460 may thus be changed into; abs( RAS * c< ,AS(CW >} h. >ο».Α§Ρ%}) wherein AS(CW) is an alternate sum of codeword CW, and the updating of step 470 may be correspondingiy changed Into RAS "** RAS^Ci. AS(CWi)a-... JtS(CVi£

Of course, a specific variant may seek to provide a DC-free and Nyquisi-free encoding, in which case the weighted-R PS-based approach is combined with the weighted-RAS-based approach.

Figure 9 is a schematic block diagram of a general archilopture of a device §00 for implementing of one or more embodiments of the invention. The device 900 comprises a communication bus connected to; - a centra; processing unit 901, such as a microprocessor, denoted CPU; - aiha«lbm,aedess: M«8$ibry 982, denoted RAM, for storing the executable code of the method of: embodiments of the invention as well as the registers adapted to record variables and parameters necessary for Implementing a method according to embodiments* the memory capacity thereof can be expanded by an optional RAM connected to snexpans-on port for example; » a read only memory 100* denoted; ROM, for storing computer programs fer imptemerilng: embodiments of the invention; - a network interface 904 Is typically connected to a communication network over which digital data to be processed are transmitted or received. The network interface 904 can be a single network interface, or composed of a set of different network interfaces (for instance wired and wireless interfaces, or different kinds of wired of wireless interfaces). Data are: written to the network interface for transmission or are read from the network Interface for reception under the control of the software application running in the CPU §01; user fits liar:; receiving inputs from a g|bh|br to display infer^allpnitaia user;

~ a hard disk 008 denoted HD ~ an I/O module 90? for resawing/sending data from/to externa; devices such as a video source or display

The executable code may be stored either in read only memory' 983, on the hard dfsti: 988 dr on a removable digital medium such as for example a disk; According to a variant, the executable code of the programs can be received by means of a communication: network, via the network interface 904, lb order tobestored in one of the storage means of the device 000., such as the hard disk 906, before being executed.

The central processing unit 001 is adapted to control and direct the execution of the tnsthjodens or portions of software code of the program or programs according to embodiments of the invention, which Instructions are stored In one of the aforementioned storage means. After poweft^: or*», iiNr &amp;PU 081 is capable of executing Instructions from main RAM memory 802 relating to a softvram applicbfloh after those instructions have been loaded fr#m tip program ROM 003 or the hambdlsc (HD) 000 for example. Such a software application, when Bmcut^d by the CPU 981, causes the steps of a method according to embodiments to be performed.

While the invention has beer; illustrated and described fh dels! iff; ftii drawings and foregoing description, such illustration ahd description ire; to bp considered illustrative or exemplary and not restrictive:,, the Invention being hot: restricted to the disclosed embodiment. Other variations to the disclosed embodiment can be understood and effected by thopf skiiied In the art in practicing: the cfalpied invention, from s study of the drawings, the disclosure and the appehdbdfbiaims.

In the cieims, the word “oomphstfrg* does not exclude oiher eiements of· steps, and the indefinite article “a* or !‘an:< does not exclude a plurality. The mere fact that different features are recited in mutually different dependent claims does not indicate that a combination of these features cannot be advantageously used. Any reference signs in the claims should not be construed as limiting the scope of the invention.

Claims (14)

1. CLAIMS 1i: A device for encoding a bitstream tor transmission over a cdilmdnteafion chain comprising: a split ahS tor splitting a bitstream pari into a plurality of sub-streams of uncoded words; a plurality of line encoders for r|apegii#ty encoding the ploraiity of substreams of encoded words, wherein each of one or mom line encoders are configured to select a codeword fromcodewords based on a current running digital sum and/or a current running aiternate sum, when encoding a respective uncoded word; an amplitude modulation unit for receiving a selected codeword from each iine encoder as input and for modulating the amplitude of a carrier wave as a function of a weighted^ s^abihatlon M the piuralifp#^ wherein the selection of the codewords by the one or more fine encoders tabes info account one or mere weights used in the weighted combination for modulating. % The device ef Claim lt wherein the currentrunning dig® sum andhx running aiternate sum are weighted running digital sum and/or weighted running alternate sum applying the same weights to the codewords as the weighted combination for modulating,;
2. 3. The device of Claim 1, whemin the pluplfy of line encoders"Is further configured to jointly select the codewords for the plurality of sub-streams. 4 The device of Claim 1, wherein one or more iine encoders are ft Μίκ* configured to select the respective codeword from codewords based on a digital sum : and/or an alternate sum of at least one other codeword selected by another lihe encoder and used in the same weighted combination for modulating. m The device of Claim 1, wherein al! the line encoders are configured to select their respective codeword from codewords based on the current running digital sum and/or running alternate sum when encoding an unsgdbd word of the respective sub-stream. It The device of Claim S, whireih each Of all the line encoders is fedher configured to select its respective codeword from codewords based on digital sums iand/or alternate sums of toe other codewords selected by the other line encoders and used in the same weighted combination for modulating. ?. The device of Claim 4 or .£, wherein selecting the respective codeword from: codewords based on digital sum or sums and/or alternate sum or sums of one or more other codewords includes applying the same-wfi&amp;felto each digital surn and/or alternate sum of a codeword as the weight applied to the same codeword in the weighted combination for modulating. S. f he device of Claim 1, wherein the plurality of line encoders is further configured to update the current running digital sum and/or running alternate sum as a weighted running digital sum and/or weighted running alternate sum using digital sums and/or alternate sums of the codewords selected in the piurality of line encoders and using respective weights from the weighted combination for modulating,
3. 9. The device of Claim 8, wherein the plurality of liril encoders if further configured to use the updated current running digital sum and/or rtinnihg/eiternite fum to encode a next bitstream part..
4. 10. The device of Claim 1, wherein the plurality of line encoders is further cohfigursdsfOi seiecf the codewords for the line encoders by minimizing the following formula;/ ahs{RS+e,. S(CWi )*.. ,*sjyS(CW,gi wherein abs(x) is the absolute value of x; CW; is a codeword for line encoder I; {c<... ,cj are the weights appileiTespesflvely to codewords the weighted c»?^ji>ift^pvl^/ffiOduiafingr.MS islbe: current running digital sum and/or running;/alternate serl;uapdlSpW^ Is a digital sum and^ alternate sum of codeword CVY.
5. 11. The device of Claim 1, whemln/When: a ftet ine encoder is further configured to; determine at least one first nominal codeword associated with the first unceded word in a first codebook, and determine a first substitute codeword from a second codebook; select one of said at least one first nominal codeword and first substitute codeword, based on the current running digital sum and/or running alternate sum; update the current running digital sum and/or running alternate sum based on a digital sum and/or an alternate s»m of the selected codeword; and output the selected codeword as a first encoded word.
6. 12. The device of Claim 11,. wherein when encoding a second uncoded word, a second line encoder of the plurality ss configured to; determine at least one second nominal codeword: associated with; the second encoded word in another first oodeboofe end determine a second subsfcte codewom from another second godahook; select one of said at least one second nominal codeword end sdfpnd:i substitute codeword, bpsed on the updated running digital sum and/pr funnjfigt alternate sum; and output the selected codeword; as a second encoded word; :13, The device of Claim 11, wherein the updating of the current running digital ssum and/or running alternate sum is based on the weight applied to the selected codeword in the weighted combination for modulating, 14 The device of Claim I, wherein encoding an uncoded ward Ih a line encoder include® determining available codewords from two or mare; oodebooks baaed on the value of the unooded word, end selecting a radeword from codewords in the line encoder consists in selecting s codeword from the determined available codewords, IS. The device of Claim 4 wherein each line encoder is a DC-free and/or Nyguist-free line encoder.
7. 18, A method; for encoding a bitstream for transmission over a splitting a bitstream part into a plurality ofsub-sireams ofuncodedwords; encoding the plurality of sub-streams of encoded words using a plurality of respective line encoders, wherein encoding an uncoded word in each of one or; more line encoders includes selecting a codeword from codewords based on a cuffenl; running digital sum and/or a current running alternate sum; modulating the amplitude of a carrier wave as a function or a wdighfed combination of selected codewords; for the plurality of line encoders; wherein selecting the codewords by the one or more line encoders takes into account one or more weights used in the weighted combination for modulating,

   17. The method Of Claim 18, wherein the current running digital sum and/or running alternate sum are weighted running digital suniand/or weighted running alternate sum applying the same weights to the codewords as the weighted combination for modulating.

   18. The method of Claim 16, wherein the codewords are jointly selected for the plurality of sub-streams.
8. 19. The method of Claim 18, wherein selecting the respective codeword from codewords is further based on a digital sum and/or an alternate sum ot at least, one; Other codeword selected by another line encoder and used In the same weighted combination for modulating. 2$, The method of Claim IB, wherein when encoding an uncoded word of each sub-stream, a respective codeword is selected from codewords based on;: the current running digital sum and/or running alternate sum. 21» The method of Claim 20, wherein selecting a respective codeword fern codewords when encoding an unceded word of each sub-stream isl further based bn digital sums and/or alternate sums of the other codewords selected by the other line encoders and used In the same weighted combination for modulating.
9. 22. The method of Claim 19 or 21, wherein selecting the respective codeword from codewords based on digital sum or sums and/or alternate sum or sums of one or more other codewords Includes applying the same weigh? to each digital sum and/or alternate sum of a codeword as the 'weigh? applied to the same codeword in: the weighted combination for modulating.
10. 23. The method of Claim 16, further (comprising updating the current running digital sum and/or njnntng alternate sum as a weighted running digital sum and/or weighted running alternate sum using digital sums and/or alternate sums of the codeword^ selected In the piurality of line encoders and using respective weights from the weighted o^
11. 24. The method :ef Claim 23:* further ^mpnsihg using the updated current running digital;: sum and/or running alternate sum to encode a next bitstream part.

    21. The method of Claim 16, wherein selecting the codewords for the line encoders includes minimizing the following formula: a bs( RS * o;, S{ CVY! }-r.,,. T£d>S|(h/T^: wherein abs(x) is the absolute value of x; CW, is a codeword for line encoder I; fc·,,...,c,Tare the weights applied respectively to codewords {CW<,...,CW,,} In the weighted combination for modulating; RS Is the current running digital sum and/or running alternate sum; and SlCW) Is a digital sum and/or alternate sum of codeword CVV. 26» The method of palm 16, wherein encoding a first uncoded word ih; i; first line encoder includes: determining at least one first nominal codeword assoc mod with the first uncodec word in a first codebook, and determining a first substitute codeword from a second codebook; selecting one of said at least one first nominal codeword and first substitute codeword,; based on the current running digital sum and/or running alternate sum; updating;: the current running digital sum and/or running alternate: sum based on a digital :;som and/or ansaltams fesum ofthe selected codeword; and ' -wo:rist>·: |T«, The method of Ciaim 26, wherein encoding a second unoodadi woof in a second inasenoodar of the:: pidrslit| includes: determining at Idesf one second nominal codeword associated with the secondencoded word In another first eodobooh, and determining e second substitute: codeword from another second codebook; selecting oae of aaidi at least one second nominal; sodeword and second substitute codeword, based on the updated running digital sum and/or running alternate sum; and outputting the selected codeword .as;-.a second encoded fed,
12. 28, The method of Claim 28, wherein updating the current running digital sum and/or running alternate sum is based on the weight applied to fha: sefecfed codeword in the weighted combination for modulating,
13. 29, The method of Claim 16, wherein encoding an encoded word in;aline encoder includes determining available codewords from two or more codebooks based on the value of the uncoded word, arid selecting a codeword from codewords in the line encoder consists in selecting a codeword from the determined available codewords.
14. 30, The method of Claim 13, wherein each line encoder Is a QC-froa; pnd/br bfyqmsi-free lino encoder. 31« A non-bansltory computer-readable medium storing a program which, wbon executed by a microprocessor or computer system tn a device, causes the device to perform the method of Claim 16, 32« A method for encoding a bitstream for transmission over a communication chain, substantially as herein described with reference to, and as shown in, Figure 4 or T of the accompanying drawings,