CN1613204A - Spread spectrum multiaddress coding method, its application method and coding device - Google Patents

Abstract

The invention provides a coding method, application method and coding device of extended frequency multiaddress, it relates to the wireless communications fields of extended frequency and code division multiple address (CDMA), repeats the elements on each of the sequence bits in the initial orthogonal sequence group by means of selecting the any orthogonal sequence groups as the initial orthogonal sequence group, and then obtains another sequence group, obtains a subcode portion of the DOC code by means of multiplying the each subcode in another subcode portion by each bit of the switching sequence Q one by one, obtains another subcode portion of the DOC code by means of multiplying the each subcode in the subcode portion by each bit of the switching sequence 1 one by one; obtains the DOC code groups by means of combining the subcode portion; since the invention can obtain the enough code groups, it's possible to average the disturbs that go through the different users by means of applying the form of skip code or disturb code; the said coding device of extended frequency multiaddress of the ivnention can produce the DOC code.

Description

Spread spectrum multiaddress coding method, its application method and coding device

A kind of spread spectrum multiple access code encoding method, application process and code device technical field

Concretely it is a kind of spread spectrum multiple access code encoding method, application process and code device the present invention relates to spread spectrum and CDMA (CDMA) wireless communication technology field.

Background technology

LS codes are a classes " complementary orthogonal, (Complementary Orthogonal Code) is announced, Application No. PCT/CN00/00028, inventor in PCT：Li Daoben, it is entitled：In a kind of patent application document of the spread spectrum multiple access coding method with zero correlation window, by the way that code word is divided into C and S two parts and a class is generated by tree with " zero interference window "（Interference Free Window, IFW) code word.

With Golay pairs（Golay Pair) it is similar, LS codes are generally divided into C codes part and the subcode part of S codes part two.In a LS code character, the auto-correlation function of each code word is preferable, and cross-correlation function has a zero correlation zone near origin.In fact, more generally, cross-correlation function is only on some ad-hoc locations（" window is stupefied "）There may be nonzero value.In the LAS-CDMA initial system schemes using LS codes, the C portion and S portion of LS codes are respectively at two at intervals of 4 chip (chips）Transmitted on the time slot of width, and employ LA codes（Different arrangements) distinguish different cell or sector.

Because available LS code characters number is less, therefore the application of LS codes is restricted.For example, if being negated even if each code word in code character and also regarding as and obtain another code character（Rotation without allowing non-180 degree of angles）, when spread spectrum coefficient is equal to 128, unilateral IFW width and is equal to 3, at most can only also construct totally 8 LS code characters.This is also why LA codes to be used in LAS-CDMA systems（Different arrangements）One of the reason for distinguish different districts/sector.On the other hand, the correlation properties between different LS code words pair are different, this also means that, distribute a fixed LS code if each user, when the traditional single user detector of use, interference that each user is undergone and its The specific code word used is relevant, that is, the performance between all users is non-average.Certain communication quality is provided because system design is necessary for each user, therefore system design must is fulfilled for using relevant requirement during the poor code word of correlation properties, this is a kind of wasting of resources for the preferable code word of those correlation properties.

If being in a cdma system, that each user distributes the fixed spread spectrum code word that a length is equal to spread spectrum coefficient (Spreading Factor), here it is " short code CDMA " --- situation during shortcode cdma systems.Many Multiuser Detections（Multiuser Detection) technology designs aiming at " short code CDMA " situation（Particularly in the uplink）.It is relative with " short code CDMA ", " frame hopping " can be used（Code Hopping) mode solves to disturb nonequilibrium problem between user.Now, the spread spectrum code word that each user is used constantly is changed by certain way, so that interference is equalized between users（Although total interference value does not change）.If using traditional Single-user detection means, using this " frame hopping, mode, which is designed, can obtain, more robust more preferable than foregoing " short code CDMA " mode（Robust system).Using long pseudorandom（PN) sequence pair short spreading code carries out scrambler

(scrambling) a kind of very cylinder just effective frame hopping mode, is considered as, corresponding system is referred to as

" long code CDMA " (longcode CDMA) system.

In existing commercial CDMA system, including IS-95, IS-2000 and WCDMA etc., this long PN sequences scrambler mode is employed, therefore these commercial CDMA systems are all " long code CDMA " systems.In these systems, Walsh orthogonal sequences are employed as channel code, and use long PN sequences（Such as m-sequence, Gold sequence）Carry out scrambler.In such systems, the effect of scrambler mainly includes：A) different cell/sections are distinguished；B) a kind of means of data confidentiality are provided；Autocorrelation value and cross correlation value randomization when c) making the relative skew non-zero of Walsh sequences in same cell/section；Notice that Walsh sequences non-zero offset correlation itself may be very big；D) the cross correlation value randomization of the Walsh sequences in different districts/sector is made.

And for LS codes, therefore, it is difficult to the interference for making to be undergone between different user using certain frame hopping or scrambler form equalization due to its code character number very little.Meanwhile, if directly disturbed to LS codes Code, then its original zero correlation window（IFW) it will be destroyed with the stupefied structure of relevant window.

The content of the invention

It is an object of the present invention to provide a kind of spread spectrum multiple access code encoding method, application process and code device.Wherein, spread spectrum multiple access code of the present invention can be referred to as DOC (Distinctive Orthogonal Code) code.The correlation function of DOC codes only has larger nonzero value on some ad-hoc locations, and it is stupefied that these ad-hoc locations are referred to as window（Window Frame DOC codes the interference introduced due to the non-ideal correlation properties between code word can be concentrated on the stupefied position of these windows.If interference path（Including ISI, MAI and ACI) exist only on those non-stupefied positions of window, then they can be completely eliminated by DOC codes；If just having interference path to be present on the stupefied position of window, certain simple mode can also be passed through（For example, the method such as balanced, Pre-RAKE and Interference Cancellation）Interference on the stupefied position of window is eliminated or is moved in advance other stupefied positions of non-window up, so that the purpose for eliminating interference can also be reached.Using DOC codes and other associated technologies, above-mentioned interference can be completely eliminated or partially removed in corresponding bi-directional synchronization CDMA (CDMA) system, make it possible the digital mobile communication system for setting up Large Copacity.This kind of new spread spectrum multiple access code（DOC codes）With with LS codes very similar characteristic, such as relevant IFW properties and the stupefied position property of window.LS codes are considered as a special case of this kind of new spread spectrum code word, and still, in addition to the LS code special cases, the new spectrum-spreading address code of a class proposed by the invention is typically no longer orthogonal complementary code.

Spread spectrum multiple access code encoding method of the present invention can obtain enough code characters, thus in spread spectrum multiple access code（DOC codes）Application in equalize using frame hopping or scrambler form the interference undergone between different user.

In spread spectrum multiple access code（DOC codes）Application in present invention also offers it is a kind of in cellular communication system network carry out DOC code character distribution method, in favor of the application of DOC codes.

Spread spectrum multiple access code encoding device of the present invention can produce described DOC codes.

Technical scheme is as follows：

A kind of spread spectrum multiple access code encoding method, it is characterised in that comprise the following steps： Any orthogonal sequence group is chosen for initial orthogonal sequence group；

Hand over the element of each sequence in Xu Lie Group on everybody to be repeated in described initial Zheng, obtain another sequence group；

Generate transform sequence Q;

By each sequence in described another sequence group respectively with described transform sequence Q wise multiplications, obtain first subcode part of DOC codes；

Generate a series of transform sequence I；

By each subcode in first described subcode part respectively with described transform sequence I wise multiplications, obtain another or multiple subcode parts of DOC codes；

Described subcode part is combined, DOC code characters are obtained.

Any orthogonal sequence group of described selection is that initial orthogonal sequence group refers to：The sequence group W that one group of orthogonal sequence containing N number of a length of N can arbitrarily be chosen is initial orthogonal sequence group；Wherein： W= {Wp w₂、 …、 w^}。

The described element by each sequence in described initial orthogonal sequence group on everybody is repeated in, and is obtained another sequence group and is referred to：

When described initial orthogonal sequence group is W={ W_n W₂..., when, can by each sequence w in W, (=1,2, elements of the Λ on everybody is repeated time respectively successively, obtains including N number of a length of another sequence group

A₂、 …、 A ;

Wherein, it can set：

= (w w,₂ … _W!.,w) i= l,2,";N and：

A, = (a_{u ¾2}...) (i=1,2, N) then have：

aiJR+l：^ai R+2 =… = ^i_tQ+i R ⁼ W,y₊₁

a= l,2,'",N; j = 0,l,"',N-l)

The step of described generation transform sequence Q, includes：

What described transform sequence Q can be obtained by basic cycle sequence P after repetition； Repeat the situation of R -1 time by turn corresponding to each sequence in described initial orthogonal sequence group, P can be set as basic cycle that length is 2, wherein it is some positive integer, it by it is front and rear two sections long be R subsequence Ρ, constituted, and each small tract for being by segment length is constituted again by Ρ, Ρ, that is,：

P={ P, P } and

...,^c

P²jai ⁼Wji, s²X ...,^s

Wherein, ^ and ^ a=ι, 2, TO be a length of R small tract, and（Following orthogonality relation is met between s：

(^) (^=(ζ=1,7 θ

Wherein, Τ represents the transposition carried out to row vector；With the sequence Ρ described in the structure_2ΒΊFor " complementary series "；Described transform sequence Q basic cycle must be one " complementary series ".

Further, basic cycle P can be constructed_{2 ¾}Former and later two subsequences and P, it is equally some " complementary series " to make them, and by they combine after basic cycle come tectonic transition sequence Q.

Above-mentioned construction process can continue to promote, it is possible to obtain a kind of make based on generation tree construction of transform sequence Q " basic cycle "；Contain one " basic cycle " on each node in the spanning tree, corresponding transform sequence Q obtains " basic cycle " after repetition；Arranged in the r of spanning tree（r = 0,1,2, …）Contain a length of 2^=2" basic cycle ", and corresponded to every repetition R-l=2 of each sequence in initial orthogonal sequence group^rSituation at-l times；If initial orthogonal sequence group contains N number of a length of N orthogonal sequence altogether, then obtained by transform sequence Q is after the length is repeated N/2 times for 2R " basic cycle ", resulting transform sequence Q length is NR.

A kind of mode for producing the spanning tree of transform sequence Q " basic cycle " can be described as follows：In spanning tree, the basic cycle contained by root node on the 0th row is（+1 +1 );Spanning tree can be constructed in the following manner：If some node in r row contains a length of

One " basic cycle " P °, it is by two sections long a ° of subsequence and b respectively for 2=R^QComposition, i.e.,：

P°= (a⁰, b°)

Then in spanning tree, by the node it is derivative, positioned at r+1 arrange two nodes, respectively containing a length of 2²=4iTwo " basic cycle " P¹And P², it is respectively：

P^J= (a¹, b¹)

= ( a⁰, b°, a⁰, -b° )

P²= (a², b²)

= (a⁰, -b°, a⁰, b°)

According to the recurrence relation, it is possible to obtain a kind of generation table structure of transform sequence Q basic cycle.

It is described by each sequence in described another sequence group respectively with described transform sequence Q wise multiplications, first subcode part for obtaining DOC codes refers to：Can arbitrarily choose one group containing N number of a length of N orthogonal sequence sequence group W be initial orthogonal sequence group, and $=^, W₂、 …、 W^};And by W each sequence W (i=l, 2 ,-, elements of the N on everybody is repeated R-l time respectively successively, obtain one group of another group of sequence group for including N number of a length of NR-, A₂、 …、 A^};

By each sequence Α in described another sequence group A, obtains the Shang subcode part of DOC codes with transform sequence Q wise multiplications respectively.¹^^¹^ ..., ^), that is,：

C^A^Q (ζ=1,2, Λ

Wherein, symbol " " represents the wise multiplication of two sequences.

The step of described generation transform sequence I, includes：Can be by described a length of 2 basic cycle（It is different from transform sequence Q basic cycle） X₂Described transform sequence I is generated after repeating NR/2Z times, and described transform sequence I may include I 1₂、 …；Wherein：

Described basic cycle X₂Can be by sequence 8_2£Composition, or can be by sequence-B₂Composition；Wherein：Sequence B₂By continuous L "+1 " immediately continuous L "-constitute, sequence that-immediately individual continuous Z "+constitutes by continuous L " -1 "； Successively using different transform sequence ft, 1₂..., I, when J carries out time conversion altogether, each conversion bank row I basic cycle can be in set { X_2i,=1,2, } in optionally, the basic cycle of every kind of length can only be chosen once；

Because wherein each transform sequence has two kinds of selections, i.e. B_2iOr-B_2£, so a total of 2 " ' plant between the series selection of possible transform sequence, the code word that they are generated and can also have certain IFW characteristic and the stupefied position characteristic of window.

It is described by each subcode in first described subcode part respectively with described transform sequence I wise multiplications, another or the multiple subcode parts for obtaining DOC codes refer to：First described subcode part and a series of transform sequence I successively wise multiplication can be obtained other one or more subcode parts of DOC codes, described transform sequence I may include 1₂、 ...；

Wherein：When the transform sequence ^ wise multiplications of each subcode in first described subcode part respectively with an a length of NR, when obtaining another described subcode part, it can set：

First described subcode part is the 1st subcode part C¹, another described subcode part is the 2nd subcode part C²,：

By the described the 1st, 2 transform sequences 1 of subcode part respectively with another a length of NR₂Wise multiplication, and respectively obtain the 3rd, 4 subcode part C³And C⁴, that is,：

C^ C¹, - ^ ( = 1,2 ··,Λ0

C⁴!. = C I₂ G = 1,2 ",N)

Aforesaid operations can constantly go on, general, when having obtained individual subcode part：

CK C²、 …、 C^L, wherein： L = l^j, _/is nonnegative integer,

Then can by C C ...,（The transform sequence Ι wise multiplications of ^ respectively with an a length of NR, can obtain another subcode part Cⁱ⁺¹、 C^L+\ …、 C^2L, that is,：

(i = \,2,-,N, j = \,2,"',L) When aforesaid operations have been carried out time altogether, and transform sequence 1 is used successively₂、 …、 l_m, then the individual subcode parts of M=2' altogether " can be generated（¹、 C …、 C^M, wherein N number of a length of NR subcode code word is contained in each subcode part.

Described is combined to described subcode part, obtains DOC code characters and refers to：M subcode part can be obtained to DOC code characters { the DOC DOC containing N number of overall length for NRM code word by combination₂..., DOC^, wherein：

DOQ=(C, C, C^M _{ (z' = l, ",N)

Thus generate a DOC code character.

In DOC code characters of the described overall length for NRM code word, parameter used is that N, R, M can generate a series of DOC code characters comprising different number of codewords, code word size and related properties using different parameter combinations.

Described transform sequence Q and described transform sequence I can carry out a variety of conversion using multiple combinations, so that the DOC code words generated are met only non-zero correlation on the stupefied position of window, and in the presence of a zero correlation window near origin,

Said process at least needs some transform sequence Q or transform sequence I to carry out linear transformation.Xu Lie Group, and transform sequence Q and transform sequence I described in identical are handed over using the initial Zheng described in different, a collection of DOC code characters can be constructed, this batch of code character can be claimed to constitute a code character set；Each described code character set is corresponding Yu Yi Group transform sequences Q and I.

M described subcode part can be in different mutually orthogonal synchronous fading channels；I.e.：

M described subcode part can be sequentially allocated to M continuous time slots；In order to avoid protection period that is overlapping between each subcode part, being made up of respectively in their end addition full neutral element.

But above-mentioned condition it is not necessary to one by one because caused by subcode partly overlaps extra correlation it is smaller；Particularly when lap is smaller, introduced extra non-zero correlation can typically be ignored. Present invention additionally comprises a kind of application process of spread spectrum multiple access code, it is characterised in that carries out scrambler to the code word in DOC code characters, i.e.,：By the sequence wise multiplication of each code word in DOC code characters respectively with an equal length.

Described scrambler comprises the following steps：

A length of N scrambler section can be intercepted successively from long PN sequences；

It will be repeated in R-1 times per bit element in each described scrambler section, obtain an a length of NR sequence；

A described a length of NR sequence is integrally repeated M-1 times again, an a length of NRM sequence, i.e. scrambler sequence is obtained；Described scrambler sequence is that described scrambler section is constituted by M sections of identical sequences；

By the M sections scrambler section in described scrambler sequence, wise multiplication is distinguished with M subcode partial codeword in DOC code words respectively, so as to complete scrambler process.

In the construction process of described DOC codes, Walsh orthogonal sequences group can be chosen as initial orthogonal sequence group, the DOC code characters produced by this orthogonal sequence group based on Walsh can be described as " basic DOC code characters,；

Here the length applied to the actual scrambler section of each symbol is smaller than the code length of code word in described basic DOC code characters；Wherein：The length of described basic DOC codes is N ^, and the length of scrambler section is N, and N can be referred to as to scrambler section " actually active length,, its value is equal to the length of each sequence in initial orthogonal sequence group.

In a code character set, all code characters can be regarded as what is obtained by a basic DOC code character after different scrambler section scramblers, i.e.,：Basic DOC code characters after scrambler, remain in that the relevant IFW characteristics and the stupefied position characteristic of window of original code character.

When carrying out spread spectrum operation using DOC codes and long scrambler, " multiple spread spectrum " can be used（Multiple Spreading) mode.

Described multiple spread spectrum refers to：Two-stage can be divided into be spread：The first order is to use basic DOC Code character is spread, wherein each basic DOC codes define a code channel；The second level is to carry out scrambler to basic DOC codes using by the scrambler sequence obtained by the scrambler step.

In application process of the present invention, one kind can be set up using DOC codes " channel code+(long）PN scramblers " cascade " multiple spread spectrum " structure, and remain in that the IFW characteristics and the stupefied position characteristic of relevant window of original basic DOC code characters.Long PN scramblers can be used for distinguishing different cell or sector, and (Whitening utilizes spread spectrum multiple access code to the interference whitening for making caused by the non-zero correlation at the stupefied position of window, can partly it even completely eliminate in S-CDMA-Synchronous Code Division Multiple Access (CDMA) and spread spectrum system, the various interference caused by nonideal correlation properties, so as to greatly improve power system capacity, and it is simplified other relevant interference cancellation techniques.

A kind of DOC codes are additionally provided in an application of the invention to carry out in cellular communication system network

The method of DOC code characters distribution, it is characterised in that：It is that a DOC code character is distributed in each cell or sector.

The method of described DOC code characters distribution, it is characterised in that：For each cell or the sector distribution basic DOC code characters of identical one；Meanwhile, each cell or sector are used as scrambler using different long PN codes.

The method of described DOC code characters distribution, it is characterised in that：It is that several different basic DOC code characters are distributed in adjacent cell or sector；Wherein, these basic DOC code characters belong to different code character set；Meanwhile, each cell or sector are used as scrambler using different long PN codes.

Present invention also offers a kind of spread spectrum multiple access code encoding device, including：PN sequence generators, Walsh code generators, transform sequence generator, frequency divider, correlator；Wherein：

Output data after the output data of PN sequence generators is related to the output data of Walsh code generators again with transform sequence generator is related, obtains output data.

Described device, it is characterised in that described PN sequence generators refer to：There can be N number of PN sequence generators, described Walsh code generators refer to：There can be N number of Walsh codes generator, described transform sequence generator refers to：There can be N number of transform sequence generator；Wherein： Output data after the output data of 0th PN sequence generator is related to the output data of the 0th Walsh code generator again with the 0th transform sequence generator is related, obtains the 0th output data；Output data after the output data of n-th PN sequence generators is related to the output data of n-th Walsh code generators again with n-th transform sequence generator is related, obtains n-th output data；The 0th described output data is related to described n-th output data to obtain output data.

Device of the present invention can generate DOC (Distinctive Orthogonal Code) spread spectrum multiple access code, and the code character inner code word has zero interference window（Interference Free Window, IFW) characteristic, and only on some specific positions（Referred to as " window is stupefied "）There may be larger non-zero correlation.Brief description of the drawings

Fig. 1 is the flow chart of DOC code generating process in the present invention；

Fig. 2 in the present invention between two DOC code words typical cross-correlation function characteristic schematic diagram；A kind of spanning tree structural representation of transform sequence Q " basic cycle " Fig. 3 is is encoded in the present invention when；

The schematic diagram for the different choice that transform sequence I may be used Fig. 4 is encodes in the present invention when；Fig. 5 for the present invention in spread using basic DOC codes and long PN sequences progress scrambler, the schematic diagram for the multiple spread spectrum structure realized；

Fig. 6 is the DOC code characters in the present invention（N=32, R=2 ,=2) in two basic correlation function values of the DOC code words after scrambler mean-square value figure（Computer artificial result, 10⁵Individual sample)；Fig. 7 is the DOC code characters in the present invention（The mean-square value figure of two basic correlation function values of the DOC code words after scrambler in N=32, R=4, M=2)（Computer artificial result, 10⁵Individual sample)；Fig. 8 is two in the present invention DOC code characters for belonging to different code character set（The mean-square value figure of two basic correlation function values of the DOC code words after scrambler between N=32, R=4, M=2)（Computer artificial result, 10⁵Individual sample);

Fig. 9 carries out the signal of system group network for a kind of use DOC codes in the present invention as spread spectrum multiple access code Figure 10 be it is bright in another ^J DOC 5 hesitate to spread many ^1^ rows system ^ and net figure；Figure 11 a are the structural representation of device of the present invention；

Scheme another structural representation that l ib are device of the present invention.

Embodiment

The present invention gives a kind of new spread spectrum multiple address code --- the coding method, application process and code device of DOC (Distinctive Orthogonal Code) code.Referring to Fig. 2, the correlation function of DOC codes only has larger nonzero value on some ad-hoc locations, and these ad-hoc locations are referred to as " window is stupefied " (Window Frame).DOC codes can concentrate on the interference introduced due to the non-ideal correlation properties between code word on the stupefied position of these windows.If interference path（Including ISI, MAI and ACI) exist only on those non-stupefied positions of window, then they can be completely eliminated by DOC codes；, whereas if just having interference path to be present on the stupefied position of window, certain simple mode can also be passed through（For example, the method such as balanced, Pre-RAKE and Interference Cancellation）Interference on the stupefied position of window is eliminated or is moved in advance other stupefied positions of non-window up, so that the purpose for eliminating interference can also be reached., can be in corresponding bi-directional synchronization CDMA using DOC codes and other associated technologies（CDMA above-mentioned interference) is completely eliminated or partially removed in system, makes it possible the digital mobile communication system for setting up Large Copacity.

(1) DOC codes coding step

In order to realize foregoing invention purpose, the building method for the DOC codes that the present invention is designed includes following steps（The general process of DOC codes is generated, as shown in Figure 1）：

Bu Sudden I：Any sequence group W={ W for choosing one group of orthogonal sequence containing N number of a length of N_r

W₂W^} ;When N is 2 some power, the orthogonal sequence group is certainly existed, for example Walsh sequences group；For the ease of narration, below, the orthogonal sequence group is called " initial orthogonal sequence group "；Step II：By each sequence W in initial orthogonal sequence group W_zThe Yi Fen Do of element of (i=1,2, N) on everybody repeats R -1 time, obtain one group of sequence group A comprising N number of a length of NR=A, A₂、 …、 A^} ;Wherein, R is some positive integer, can typically be chosen for 2 some power；That is, if it is assumed that： W_f = (w_uw … w_!yv) "=1,2, Λ

And,

A_z = (a _¾2… a, ）（ζ·=1,2 .,Λ

Then have：

aiJR+l ~ ^aiJR+2 ^{= = a}W+l)R ^{= W}ij+\

(z'=l,2,〜,N; j = 0,l,〜,N— 1)

Walk Pei III:(Q building method is individually provided one a length of NR of generation transform sequence Q later）.Then each sequence in sequence group A is entered₂The 1st subcode part of DOC codes is obtained with transform sequence Q wise multiplications respectively¹ = {< ¹"=1,2, ^ }, that is,：

C^A^Q (i=l, 2 .., N)

Wherein, symbol " " represents the wise multiplication of two sequences.

Step IV：By the 1st subcode part C¹In each subcode transform sequence ^ wise multiplications respectively with an a length of NR, obtain the 2nd subcode part (²={ ²=1,2,7 ^, that is,：Then, can by the 1st, 2 transform sequences 1 of subcode part respectively with another a length of NR₂Wise multiplication, and respectively obtain the 3rd, 4 subcode part C³And C⁴, that is,：

C³

, Λ

C = C² I₂=1,2 .., N)

Aforesaid operations can constantly go on, general, if it is nonnegative integer to have been obtained for the j of L=2）Individual subcode part C¹ C² C^L, then can be by their transform sequences 1 respectively with an a length of NR_£Wise multiplication, can obtain another subcode part Cⁱ⁺¹、 C^L+2 C^2L, that is,：

The Ι '=1,2 of 0^=0, N, _ /=1,2 ,-,）If aforesaid operations have carried out secondary altogether（Transform sequence I I are used successively₂ l_m), then can generate the individual subcode parts of Af=2' altogether " (¹、 C² C^M, wherein N number of a length of NR subcode code word is contained in each subcode part. Transform sequence I (including I 1₂、 …）Building method also will individually provide later.

Bu Sudden V：This M subcode part is obtained into DOC code characters { the DOC^ DOC containing N number of overall length for NRM code word by combination in some way₂DOC, wherein：

DOC, = ( C¹,., C , C^. ) ( z' = l,2 .. ,N)

A DOC code character is thus generated, wherein parameter used is（N、 R、 M).A series of DOC code characters comprising different number of codewords, code word size and related properties can be generated using different parameter combinations.In step III and IV, various conversion can be carried out using the transform sequence Q and transform sequence I of various flexible combinations, only there is non-zero correlation on the stupefied position of window meet generated DOC code words, and in the presence of a zero correlation window near origin（Here correlation function is defined as the sum of each subcode part correlation function）.Note, in said process, at least（By some transform sequence Q or I) carry out linear transformation.

Using different initial orthogonal sequence groups, and identical transform sequence Q and I, a collection of DOC code characters can be constructed, for the ease of narration, claim this batch of code character to constitute one " code character set ".Each " code character set " is corresponding with one group of transform sequence Q and I.

In actual applications, this M subcode part should be in different mutually orthogonal synchronous fading channels.For example, this M subcode part can be sequentially allocated to M continuous time slots.Now, in order to avoid protection period that is overlapping between each subcode part, being made up of respectively in their end addition full neutral element.But, this it is not necessary to, although in fact, the overlapping of each subcode part can destroy the full zero correlation value at the original non-stupefied place of window, only can enter less correlation by I over these locations, their influence can be ignored in general.

Transform sequence Q and I construction are one of cores of DOC codes, and their generation method is provided individually below.

Transform sequence Q generation method：

Transform sequence Q is obtained by some so-called " basic cycle " sequence P after repetition.Each sequence repeats the situation of R-1 times by turn in the initial orthogonal sequence group of correspondence, if P_{2 ¾}It is 2KR for length One " basic cycle "（Wherein K is some positive integer）, it by it is front and rear two sections long be KR subsequence P, constituted, and P, P²^ is again each to be made up of segment length for R small tract, that is,：

And

^KR ⁼ C²R, ■,^C

s²,^s

Wherein, c and s (i=s, 2 ..., K) be a length of R small tract, and following orthogonality relation is met between c and s：

(c^)(s^)^T= 0 "· = 1,2 . , ）

Wherein, Τ represents the transposition carried out to row vector.For the ease of narration, weighing-appliance has the sequence of the structureFor " complementary series ".Transform sequence Q " basic cycle " must be one " complementary series ".It can illustrate, based on above-mentioned form Shuo " basic cycle " P_2ffliThe transform sequence Q constructed, in (2j l) XR, (wherein j is integer to the subcode partial codeword generated）On point position from/cross-correlation function value be zero.

Further, " basic cycle " P can be constructed_{2 ¾}Former and later two subsequence P and make them equally be some " complementary series ", now by they combine after " basic cycle " P construction transform sequence Q, the subcode partial codeword generated, (it is wherein positive integer）On point position from/cross-correlation function value be zero.

Above-mentioned construction process is continued to promote, a kind of make based on generation tree construction of transform sequence Q " basic cycle " can be obtained（Such as Fig. 3).Contain one " basic cycle " on each node in the spanning tree, corresponding transform sequence Q obtains " basic cycle " after repetition.Arranged in the r of spanning tree（r = 0,1,2, …）" basic cycle " containing a length of 2^=2, and corresponded to every repetition R-l=2 of each sequence in initial orthogonal sequence group^rSituation at-l times.If initial orthogonal sequence group contains N number of a length of N orthogonal sequence altogether, then obtained by transform sequence Q is after the length is repeated N/2 times for 2R " basic cycle ", resulting transform sequence Q length is.Accordingly, it can be explained as in the subcode part produced by this transform sequence Q, in addition to odd point, Only in J2R, (j is nonzero integer to the auto-correlation function value of each code word）There may be a non-zero correlation on point, and only in j2R, (j is integer to the cross-correlation function value between any two code word）There may be non-zero correlation on point.

The mode for producing a kind of transform sequence Q " basic cycle " based on spanning tree can be described as follows.In spanning tree, the basic cycle contained by root node on the 0th row is（+1+1).Spanning tree can be constructed in the following manner：If some node in r row contains a length of 2 2R " basic cycle " P⁰, it by it is two sections long respectively for 2=Subsequence ^ and！^ is constituted, i.e.,

Ρ⁰= (a⁰, b°)

Then in spanning tree, by the node it is derivative, positioned at r+1 arrange two nodes, respectively containing a length of 2²=4Two " basic cycle " P¹And P², it is respectively：

P^J= (a¹, b¹)

= (a⁰, b⁰, a⁰, -b°)

P²= (a², b²)

= (a⁰, -b°, a⁰, b°)

According to the recurrence relation, so that it may with obtain Bian Office row Q the cycle one kind.

On the other hand, when difference " basic cycle " Ρ using equal length is come tectonic transition sequence Q, between resulting different subcode code words, it is also possible to certain IFW characteristic and the stupefied position characteristic of window.

Transform sequence I generation method：

Transform sequence I (including I 1₂、 …）Equally be by some " basic cycle " after repetition obtained by（Such as Fig. 4).A length of NR transform sequences I is by some a length of 2L " basic cycle " X_2LObtained after repeating NR/2 times（ = 1,2, …）.Wherein, " basic cycle " X_2iContinue L in succession by continuous Ge Tight and (be set to sequence_2L) composition, or by continuous L "-immediately continuous "+Γ (i.e. sequence-B₂) composition.For example：

Length is 2 basic cycle X₂: B₂=(+1-1) or-B₂ = (-1 +1)

Spend the basic cycle X for 4₄: B₄=(+1+1-1-1) or-B₄ = ( -1 -1 +1 +1 )

Length is 8 basic cycle X₈: B₈=(+1+1+1+1 1-1-1-1) or

-B₈=(- 1-1 11 one 1+1+1+1+1)

It can illustrate, based on " basic cycle "₂The transform sequence I of construction, the two cross-talks code code word generated (is wherein positive integer in (2 Η) ￡）On point position from/cross-correlation function value and be zero.

Successively using m different transform sequence ft, 1₂1, } when carrying out common m time and converting, each transform sequence I basic cycle can gather { X_2£,=1,2 ... } in optionally, but " basic cycle " of every kind of length can only be chosen once.Because wherein each transform sequence has two kinds of selections（That is B₂Or-B_2i), so a total of 2 ', " certain IFW characteristic and the stupefied position characteristic of window can also be had by planting between possible transform sequence series selection, the code word that they are generated.

In the generating process of foregoing DOC codes, in step III and IV, the various conversion made using transform sequence Q and I can use various flexible combinations.It is zero that various transform sequence Q and I reasonable selection, which will make from/value of the cross-correlation function on some positions,（If the correlation of each subcode part is carried out respectively, and total correlation function is the sum of each subcode part correlation function）.And the correlation being left on those positions is generally nonzero value these positions is exactly so-called " window is stupefied " position one by one.Particularly, if being zero in the correlation function value near origin in +/- W, gained code character has the zero interference window that One-sided Widths are W（ IFW ).

For example, an ancient weapon made of bamboo will construct such a DOC code characters, it contains the DOC code words of N=16 a length of 128, and its correlation function value is only for integer）There may be non-zero correlation on point position, and unilateral IFW width is equal to 7.The relevant coding method described using the present invention, can generate different DOC code characters, different transform sequence Q and I combination has been respectively adopted in they using following several ways：

Mode 1:N=16, R=4, M=2, wherein using linear transformation sequence Q (basic cycle P₈, produced by spanning tree）With linear transformation sequence I (basic cycle X₂ );

Mode 2:N=16, R=2, M=4, wherein using linear transformation sequence Q (basic cycle P₈, Complementary series）(basic cycle is respectively transform sequence I twice₂With)；Mode 3:R=2 of N=16., M=4, wherein using linear transformation sequence Q (basic cycle P₄) and twice transform sequence I (basic cycle is respectively₂With₈)；

Mode 4:N=6, R=l, M=8, wherein using cubic transformation sequence I, (basic cycle is respectively

X₂、 X₄And X₈ );

(2) the scrambler process of DOC codes

Due to based on any one initial orthogonal sequence group, by relevant generation method described in the invention, a DOC code character can be generated；And the orthogonal sequence group of the sequence comprising N number of a length of N has 1^NIt is individual（Discounting for code character inner code word sequential arrangement or negate conversion etc. equivalence）, so the quantity of DOC code characters is very more.Can be that a DOC code character is distributed in each cell or sector in actual cellular CDMA communications system.Which has corresponded to " short code CDMA " mode.

On the other hand, if using frame hopping mode, can equalize the performance between each user.It is a kind of very cylinder just effective frame hopping mode to carry out scrambler using long PN codes, and which has corresponded to " long code CDMA " mode.

In the generation step of DOC codes, Walsh orthogonal sequence groups can be first chosen, and each code word in the Xu Lie Group is multiplied by the sequence of an equal length by turn（The sequence is derived from one section in long PN codes, and its length is equal to the length of each sequence in Walsh Xu Lie Group）, so generate an initial orthogonal sequence group.Then, then fainted by other relevant steps and construct DOC code characters.

All it is to repeat and wise multiplication operation due to what is be related in the construction process of DOC code characters, so scrambler Bu Sudden can also be delayed to final step progress.Now, DOC code characters are generated by Walsh sequences group as initial orthogonal sequence group, and corresponding DOC code characters are referred to as " basic DOC code characters "；Then, each subcode part former scrambler section carried out after repeated transformation again with the basic DOC code characters carries out wise multiplication, so as to complete code-scrambling operation.In a code character set, all code characters can be regarded as what is obtained by a basic DOC code character after different scrambler section scramblers.Basic DOC code characters after scrambler have remained in that original IFW characteristics and the characteristic about the stupefied position of window. Scrambler can be carried out to the code word in " basic DOC code characters ", will each sequence wise multiplication of code word respectively with an equal length.Wherein, be N, R, M for parameter) DOC code characters, relevant scrambler process is divided into following steps：

Step A:Intercept one of a length of N successively from long PN sequences " scrambler section, D;

Step B:It will be repeated in R -1 times per bit element in each " scrambler section ", and obtain an a length of NR sequence E;

Bu Sudden C:Sequence E is integrally repeated M-1 times again, obtains an a length of JVRM sequence F, i.e. scrambler sequence；Scrambler sequence is by M sections of identical sequences（That is " scrambler section "）Constituted；

Step D:By the M sections " scrambler section " in scrambler sequence E, wise multiplication is distinguished with M subcode partial codeword in DOC code words respectively, so as to complete scrambler process；

It is noted that with existing cdma system（Such as IS-95 lengthens PN code scrambler modes using Walsh sequences）Difference, here applied to each symbol actual scrambler section code length of the length than code word in basic DOC code characters（That is processing gain or spread spectrum coefficient）It is small.Herein, the long NRM of basic DOC codes, and therefore the length of scrambler section is, below when describing relevant scrambler process, for the ease of narration, N is referred to as to " the actually active length " of scrambler section, its value is equal to the length of each sequence in initial orthogonal sequence group.

When application DOC codes and long PN sequences scrambler carry out spread spectrum operation, so-called " multiple spread spectrum " can be used（Multiple Spreading) mode, referring to Fig. 5.Multiple spread spectrum mode has been obtained for extensive use in existing commercial CDMA system.It can be divided into two-stage progress for the multiple spread spectrum of DOC codes：The first order is to be spread using basic DOC codes, wherein each basic DOC codes define a code channel（Code Channel), each the code word in basic DOC code characters is one " channel code " (Channelization Code);The second level is use（After repeated transformation）Long PN codes carry out scrambler (Scrambling).The purpose of scrambler is that the interference for introducing the non-ideal correlation properties in same cell due to code word between cell carries out albefaction（Whitening), and code character used is can be used to distinguish in different districts or sector, realizes " reuse " of the same basic DOC code characters in different districts or sector ( reuse )。

(3) coded system is illustrated：

Below by example and accompanying drawing, the present invention will be described in detail.During previous constructions, the selection of different R and M parameters can produce the DOC code characters of different qualities, several more typical examples are chosen here the make of DOC codes is described in detail, and briefly provide the relevant characteristic of corresponding DOC code characters.

Example 1:DOC Ma Group（N=4、 R = 2、 M=2)

Illustrate the coding step of the present invention first since a simply example.If choosing initial orthogonal code sets for the Walsh sequence groups containing the sequence of 4 a length of 4（Correspond to coding step 1), they are respectively：

(+1 +1 +1 +1)

(+1 -1 +1 - 1 )

(+1 +1 -1 - 1 )

(+1 -1 -1 +1 )

Then each bit of above-mentioned each code word is repeated in once（Correspond to coding step II), obtain the sequence group A of the sequence of 4 a length of 8:

(+1 +1 +1 +1 +1 +1 +1 +1 )

(+1+1 one 1-1+1+1-1-1)

(+1 +1 +1 +1 -1 -1 -1 -1 )

(+1 +1 -1 -1 -1 -1 +1 +1 )

Then, the basic cycle is selected in the spanning tree of transform sequence Q basic cycle.Two basic cycles are had in the 1st row of spanning tree, are respectively（+ 1+1+1-1) and（+ 1-1+1+1), the two basic cycles are repeated twice respectively, the transform sequence Q of two optional a length of 8 are obtained, if they are respectively Q1 and Q2:

Q1=(+1+1+1 one 1+1+1+1 one 1) Ql is chosen as transform sequence in Q2=(+1-1+1+1+1-1+1+1), and each sequence in foregoing sequences group A is multiplied by after sequence Q1 by turn respectively（Correspond to coding step III), obtain the sequence of 4 a length of 8：

c = (+1 +1 +1 -1 +1 +1 +1 -1)

c = (+1 +1 -1 +1 +1 +1 -1 +1)

c = (+1 +1 +1 -1 -1 -1 -1 +1)

C=(+1+1-1+1-1-1+1 one 1)

They constitute the 1st subcode part of DOC codes.

In order to which the basic cycle for generating another subcode Partial Transformation sequence may be selected to be（Or (- 1+1) ,+1-1) now corresponding transform sequence is respectively：

I_t=(+1-1+1-1+1-1+1-1) or

I_t= (-1 +1 - 1 +1 -1 +1 -1 +1)

Previous ^ is chosen as transform sequence, each sequence in above the 1st subcode part of gained is multiplied by after sequence ^ by turn respectively（Correspondence coding step IV), obtain the sequence of another 4 a length of 8：

c\ = (+1 -1 +1 +1 +1 -1 +1 +1)

c\ = (+1 -1 -1 -1 +1 -1 -1 - 1)

C² ₃ = (+1 -1 +1 +1 -1 +1 -1 - 1)

\ = (+1 -1 -1 -1 -1 +1 +1 +1)

They constitute the 2nd subcode part of DOC codes.

By the 1st subcode part and the 2nd subcode part combination（Correspond to coding step V), one has just been obtained containing DOC code character of 4 overall lengths for 16 code word：

DOC1 = (C¹!, C²! ) = ( +1 +1 +1 -1 +1 +1 +1 -1 , +1—1 +1 +1 +1 -1 +1 +1 ) DOC2 = ( C¹,, C² ₂) = (+1+1- 1+1+1+1- 1+1, +1-1-1-1+1 1-1-1) DOC3= (C^, C² ₃)=(+1+1+1-1-1-1-1+1 ,+1-1+1+1-1+1-1-1) DOC4=(C^! ₄, C² ₄) = (+1+1—1+1— 1—1+1—1, +1—1—1—1—1+1+1+1) Given in table 1 between this 4 code words from/cross-correlation function value（Here correlation function is the sum of each each auto-correlation function in subcode part, without overlapping between each subcode part of consideration）, wherein R^ represents the correlation function between numeral DOCX and DOCY（It is auto-correlation function as X=Y, is cross-correlation function as Χ ≠ Υ）.From the table, all correlation functions only may have nonzero value on +/- 4 points, and these points are exactly " window is stupefied " (window frame) position.Particularly, exist（- 3 ,+3) it is interval in, except auto-correlation function is in addition to origin gets maximum, on remaining point is zero from/cross-correlation function value, therefore the DOC code characters have zero interference window（), IFW its One-sided Widths is equal to 3.The DOC of table 1 is suitable（N=4, R=2, the related tool of M=2) just arranging Walsh sequence, Bian Office row Q the cycle into（+ 1+1+1-1), Bian Office Lie the cycle into (+1-1)

Because above-mentioned code character is generated by initial orthogonal sequence group of Walsh sequences group, therefore it is one " basic DOC code characters " as defined above.Scrambler is carried out to the basic DOC code characters, because " the actually active length " of scrambler section here is 4, so the initial scrambler section that length is 4 can be chosen, can for example choose scrambler section is（+1 +1 -1 +1 ).First, every of scrambler section is repeated once, obtains sequence： (+1 +1 +1 +1-1 -1 +1 +1 )

Then the sequence is entirely repeated once again, obtains a length of 16 4 outstanding yards of sequences：

(+1+1+1+1-1 one 1+1+1 ,+1+1+1+1-1 one 1+1+1)

By the scrambler sequence respectively with after each code word wise multiplication in the basic DOC code characters that are previously obtained, obtaining containing 4 DOC code characters by the code word of scrambler with next：

DOC1=(+1+1+1-1-1-1+1 one 1, + 1 one 1+1+1-1+1+1+1) DOC2=(- 1-1-1+1 of+1+1-1+1, + 1-1-1-1-1+1-1-1) DOC3=(+1+1+1-1+1+1-1+1, + 1-1+1+1+1-1-1-1) give in DOC4=(+1+1-1+1+1+1+1-1 ,+1-1-1- 1+1-1+1+1) table 2 this 4 by between the code words of scrambler from/cross-correlation function.It is identical with the correlation properties of basic code character from the code character after the table, scrambler, that is, only positioned at +/- 4 points " window is stupefied, place there may be non-zero correlation.Both differences are that the correlation at the stupefied position of window is different, in fact, after long PN codes scrambler, can make correlation function value albefaction and Gaussian of the DOC code characters at the stupefied position of window.

Table 2 The correlation of (N=4, M=2 of R=2.)), positive row group be Walsh sequence, Bian Office row Q the cycle into（+ 1+1+1-1), Bian Office Lie the cycle into（+ 1-1), disturb (+1+1-1+1)

It is relative to move

- 7-6-5-4-3-2-1 01234567：

0 0 0 0 0 0 0 16 0 0 0 0 0 0 0

0 0 0 0 0 0 0 16 0 0 0 0 0 0 0 i?₃₃(r) 0 0 0 0 0 0 0 16 0 0 0 0 0 0 0

^44 (r) 0 0 0 0 0 0 0 16 0 0 0 0 0 0 0

0 0 0 -8 0 0 0 0 0 0 0 -8 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 -8 0 0 0 0 0 0 0 8 0 0 0

In addition, in above-mentioned construction process, other optional basic cycle difference tectonic transition sequence Q and I can also be taken, to generate the code character in other DOC code characters set.It can examine them that there is identical IFW width and the property about the stupefied position of window with upper example.Here no longer praise and state.

Example 2:DOC code characters（N=4、 R=l、 M=4)

Initial orthogonal code sets are chosen first for the Walsh sequence groups containing the sequence of 4 a length of 4（Correspond to coding step I), they are respectively：

(+1 +1 +1 +1 )

(+1 -1 +1 - 1 )

(+1 +1 -1 - 1 )

(+1 -1 -1 +1 )

Due to R=l here, that is, not using transform sequence Q accordingly to be converted（That is, skip coding step II, III), therefore, above Walsh sequences group just constitutes the 1st subcode part of DOC codes, that is,：

c = (+1 +1 +1 +1 )

c ⁼ (+1 -1 +1 - 1 )

^3 = (+1 +1 -1 - 1 )

^₄ = (+1 -1 -1 +1 )

In order to generate the 2nd subcode part, transform sequence I, basic cycle may be selected to be (+1-1) or（- 1+1) the former basic cycle, is chosen, now corresponding transform sequence is respectively：

1】Each sequence in 1st subcode part is multiplied by after sequence ^=(+1 one 1+1 one 1) by turn respectively（Bian Ma Bu Sudden are corresponded to IV), the sequence of another 4 a length of 4 is obtained, they constitute the 2nd subcode part of DOC codes:

C = ( +1 -1 +1 -1)

c² ₂= (+i +1 +1 +1)

-1 -1 +1)

C² ₄=(+1+1-1 one 1)

In order to generate the 3rd, 4 subcode parts, transform sequence 1₂Basic cycle may be selected to be（+ 1+1-1-1) or（- 1-1+1+1) the II former basic cycles, are chosen, now corresponding transform sequence difference

+

For：

1₂= (+1 +1 -1 -1)

Each sequence in 1st subcode part is multiplied by sequence 1 by turn respectively₂Afterwards（Correspond to coding step IV), the sequence of another 4 a length of 4 is obtained, they constitute the 3rd subcode part of DOC codes： c³^ (+1 +1 -1 -1)

c³ ₂= (+1 -1 -1 +1)

(+1 +1 +1 +1)

c³ ₄ = (+1 -1 +1 -1)

Each sequence in 2nd subcode part is multiplied by sequence 1 by turn respectively₂Afterwards（Correspond to coding step IV), the sequence of another 4 a length of 4 is obtained, they constitute the 4th subcode part of DOC codes： c\ = (+1 -1 -1 +1)

c\ = (+1 +1 -1 -1)

c\ = (+1 -1 +1 - 1)

C⁴ ₄ = (+1 +1 +1 +1)

4 subcode parts are combined and (have corresponded to coding step V) by more than, have just obtained one containing DOC code character of 4 overall lengths for 16 code word：

DOC C C²!, C³!, C⁴ ₁) = (+1+1+1+1, +1-1+1-1, +1+1-1-1, +1-1-1+1) DOC2=(C¹ _?5 C²,, ₂, C₂) = (+1-1+1-1, +1+1+1 +1, +1-1-1+1, +1+1-1-1)

(+1+1-1-1, +1-1-1+1, +1+1+1+1, +1-1+1-1) DOC4=(C¹ ₄, (, C₄, ()=(+1-1-1+1 ,+1+1-1-1 ,+1-1+1-1 ,+1+1+1+1) can verify, this group of DOC code is preferable from/cross-correlation function, i.e., the auto-correlation function of each code word only gets nonzero value in origin（It is complete zero equal to the cross-correlation function 16), between any two code word（Here correlation function is the sum of each each auto-correlation function in subcode part, without overlapping between consideration subcode part）. '

Basic DOC code characters scrambler can be carried out to more than, due to scrambler section here " actually active length, it is 4, so the initial scrambler section that length is 4 can be chosen, can for example choose scrambler section is（+1 +1 - 1 +1).Whole by scrambler section repeats M-l=3 time, obtains the scrambler sequence that an overall length is 16：(+1+1-1+1 ,+1+1-1+1 ,+1+1-1+1 ,+1+1-1+1)

By the scrambler sequence respectively with after each code word wise multiplication in the basic D0C code characters that are previously obtained, obtaining containing 4 D0C code characters by the code word of scrambler with next：

D0C1 = (+1 +1—1 +1, +1-1-1—1, +1+1 +1-1, +1—1+1+1 )

D0C2 = (+1-1-1-1, +1+1-1+1, +1-1+1+1, +1+1+1-1 )

D0C3=(+1+1+1-1 ,+1-1+1+1 ,+1+1-1+1 ,+1-1-1-1)

D0C4=(+1-1+1+1 ,+1+1+1-1 ,+1-1-1-1 ,+1+1-1+1) can equally verify, the group be still from/cross-correlation function by the D0C code characters of scrambler it is preferable,

In addition, in above-mentioned construction process, other optional basic cycle difference tectonic transition sequence ^ and ^ can also be taken, to generate the code character in other D0C code characters set.It can examine them that there is identical IFW width and the property about the stupefied position of window with upper example.Here repeat no more.

Example 3:D0C code characters（N=32、 R = 2、 M=2)

As a more complicated example, the Walsh sequences group containing the sequence of 32 a length of 32 is first chosen as initial orthogonal sequence group, to generate a basic D0C codes Group containing the code word of 32 a length of 128.Here Selecting All Parameters are R=2, also will initially each sequence is weighed by turn successively in orthogonal sequence group Multiple R -1=1 time.Transform sequence Q basic cycle is chosen in the 1st row of its spanning tree.Here selection transform sequence Q basic cycle is（+ 1+1+1-1), and it is repeated 16 times and obtains transform sequence Q.And transform sequence I basic cycle is chosen for（+ 1-1), and 32 times are repeated obtain transform sequence so, by the generation method of above-mentioned DOC codes, the available DOC code characters for containing 32 code words are as follows：

Basic DOC code characters（IV=32, R=2, M=2), transform sequence Q basic cycle is（+ 1+1+1-1), the basic cycle of transform sequence is（+1—1 )

Code word 1

(+++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++ _ +++-+++-+++-,+- +++ _ +++-++-+++-+++-+++ _ +++-+++-+++-+++ _ +++-+++-+++ _ +++-+++-++) code word 2

(++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++-+++ -+,

+ -+-++ -+-+-+-+-+-+-+-+-+-+-+-+-) code word 3

(+++ ++++ ++++ ++++ ++++ ++++ ++++ ++++ + ₅

+-++ -+one by one+- ++ -+-+- ++ -+one by one+- ++ -+one by one+- ++ -+one by one+- ++ -+one by one+- ++ -+one by one+- ++ -+---) code word 4

(++ -+one by one+- ++ -+- mono-+- ++ -+one by one+- ++ -+one by one+- ++ -+one by one+- ++ -+one by one+- ++ -+- mono-+- ++ -+one by one+- ,++++++++++++++++++++++++++++++++) code word 5

(+++-+++ + ++++-+++ + ++++-+++ + ++++-+++ + +₅

+-+++-++ -++ one by one+- +++-++ -++ one by one+- +++-++ -+H ---+- +++-++ -++ one by one) code word 6

(1 11 h of h+H of+H h+H 111 ~ 111 hH of l h-H-h+H 11 1-,++ ++ one +++++ +++-+++++++-+++++++-++) ί ) Ηi+— + hτ +++++ί++ +++ +++++ Γί〜

(++++++ -+T ++++ +++++ i+ --- Ι 1 --- --- l Ι i) +++-++ -++ ++ Bian+++- +++ Bian+I | --- I --- --- |

( +—+ + + + + + + + + + + + + + ++ + + + + + + + + + + + + +I I

6)++ -+++ ++++++++++++ ++++++ Bian++ 1 |

( + + + +—+++—++—++ + + +++| II II|

+ + + + + + + + + + + +— + + + + Γ— II |

++++++ ± +++ten++!.ιι } ί + ΐί ίί+ + +++++ΐ+—+ ++++ +

(++-+++ +++ +++++++++ ten++++ -+T --- II --- ----II II1. 1 --- Ι }++++ ++ +++Bian ++++Τ ++++-+1 +++-+Bian +++ Ι | --- | ------- Ι Ι --- 1 --- 1 ---

(— +++ +++—+++ ++Ϊ+ ++++ ++ ++++ + +++++

Π ) + + + + + + + + +— + +Τ+Ϊ + + ΐ+ + ί + + +ΐ + + Ι1 - !

(— + +++++—++++ +++†+† +++ + ++†——1 II1 ) ++++— + +++—+++++ + ++++—+| II II||1

( + ++++++++++ + ++4† ++ + ++ +11 O/iiAV 9is3/:/isl£ C99i )+++ ++++++++ ++++ + ++ +++++++ + +++ +++ΙΙ -

(+++++ ++++ +++ ++++++++++ + ++ + +++++ +IΙΙ ίI1 -IΙΙΙ 1-ΙΙΙ IIΙ—— IIΙΙ )+++ Τ++ ++++++++++ _+++++ + + +++ +++ ++ ΓΙΙ——ΙΙΙ——ΙΙΙΙ——ΙΙ——.-

( + + + + + +++ + +++++++ + + ++ ++ + + + + + ++ + ++! Ι

+++++ +++++ +++ + + +++——I——I IIIII-

(++++ ++++ ++ ++++++++ ++++ + + +IIII ϋ ε9υ/:1£ 9ii OooiAV ) + +++++ ++ +++ + ++++++— -

o

0Z (++++++++++++- ++++++++- +++ +++++ ++++-,

+-++-H H+- ++ -+---+- +++-++ -++ ---+- +++-++-H+- ++ -+h one by one+- ++) code word 32

(++ -+one by one++-++ -+one by one+- ++-+++ -+one by one++-++-+++ -+one by one+- ++ -+_ mono-++-++ -+,

+ +++-++++++++H +++-+++++ +++++++-++++) can verify, the stupefied position of window of this group of code word is equal at 4k points in relative skew, and wherein k is a nonzero integer.

To the basic D0C code characters（It is not added with protection period between wherein the 1st and the 2nd subcode part, and is included in overlapping between the 1st and the 2nd subcode part）Scrambler is carried out, scrambler sequence is produced by computer random.Using different scrambler sections（Its length be N=yi, when, correlation between gained code word is usually differed in addition to origin, from the point of view of statistically, when the long scrambler of use is sufficiently random, correlation function value at each relative offset point is by the good Gaussian Profile for being zero from average.Given in Fig. 6 in above-mentioned basic DOC code characters, code word 1 and code word 10 are after scrambler, the mean-square value of correlation function value at each relative offset point（The average value of the square value of i.e. all statistical samples）.Assuming that receiver receives the corresponding target code channel of code word 1, then the mean-square value directly to should be appreciated that expand after the average interference power value that is caused to code word 1 in each deviation post of code word 10.As seen from Figure 6, on the stupefied position of window（4/_CAt point, wherein A is nonzero integer）There is larger correlation function mean-square value between code word 1 and code word 10；And on the stupefied position of non-window, correlation function mean-square value very little.Note the non-zero correlation on the stupefied position of non-window herein, be due to DOC code words the 1st and the 2nd subcode part between it is overlapping caused, but their value is typically small（Particularly near origin）Even if the situation of 31 other interference users is considered, in former associated window（That is,+1/-1 ,+2/-2 ,+31-3 these with respect on deviation posts）, it is total caused by two subcodes partly overlap at every bit to disturb relative very little equal compared with echo signal（Below -20dB, it is assumed that have 31 interference code words）, thus can typically ignore in actual system design.

As it was previously stated, can also choose other possible basic cycles comes tectonic transition sequence Q and I, come Generate the code character in other DOC code characters set.If set by the basic cycle（+ 1+1+1-1) and（+ 1-1+1+1) transform sequence for repeating to obtain is respectively Ql, Q2;By the basic cycle（+ 1-1) and（- 1+1) transform sequence for repeating to obtain is respectively II and 12, then can define following 4 code character set and corresponding transform sequence（Q, I) it is right：

Code character set SETA: (Ql, II)

Code character set SET B: (Q2, II)

Code character set SET C: (Ql, 12)

Code character set SET D: (Q2, 12)

Then there is certain related spy to be listed in table 3 between the ^ in this 4 code character closings, and (relevant characteristic can also be referring to Fig. 3, and the unilateral IFW width of the interblock corresponding to the difference basic cycles is marked in shaded box there）：Correlation properties between the code character set of table 3（It is designated as unilateral IFW width）

Example 4:DOC code characters (N=16, R=4, M=2)

The Walsh sequences group containing the sequence of 16 a length of 16 is first chosen in this example as initial orthogonal sequence group, to generate the basic DOC code characters containing the code word of 16 a length of 128.Here Selecting All Parameters are R=4, also will initially each sequence is repeated by turn successively in orthogonal sequence group- 1=1 time.Transform sequence Q basic cycle can be chosen in the 2nd row of its spanning tree, the correspondence different basic cycle, 4 kinds of transform sequences Ql, Q2, Q3 and Q4 can be generated altogether after repeating 8 times, wherein：

Basic cycle（+ 1+1+1-1+1+1-1+1) → transform sequence Q1;Basic cycle (+1+1-1+1+1+1+1-1) → transform sequence Q2;Basic cycle (+1-1+1+1+1-1-1-the 1) → bad ij Q3 of conversion sequence; Basic cycle (+1-1-1 one 1+1-1+1+1) → conversion sequence is bad ' J Q4;

And another transform sequence I has two kinds of selections, if by the basic cycle（+ 1-1) sequence generated after repeating is II, and by the basic cycle（+ 1-1) sequence generated after repeating is 12.Therefore, transform sequence pair（Q, I) 8 kinds of possible selection combinations are had, and 8 code character set SET A-SET H are corresponded to respectively.Wherein, this corresponding transform sequence of 8 code character set is to being respectively：

Code character set SETA: (Ql, 11)

Code character set SET B: (Q2, 11)

Code character set SET C: (Q3, 11)

Code character set SETD: (Q4, 11)

Code character set SET E: (Ql, 12)

Code character set SETF: (Q2, 12)

Code character set SET G: (Q3, 12)

Code character set SETH: (Q4, 12)

Code character in this 8 code character set has certain correlation properties between them in itself, is listed in table 4 (relevant characteristic can also be referring to Fig. 3):

Correlation properties between 48 code character set of table（It is designated as unilateral IFW width）

SETA SETB SETC SETD SETE SETF SETG SETH

IFW = IFW = IFW = IFW =

SETA

7311 it is orthogonal it is orthogonal ' it is orthogonal

IFW = IFW = IFW = IFW =

SETB

3711 is orthogonal

IFW = IFW = IFW = IFW =

SETC

1173 is orthogonal

IFW = IFW = IFW = IFW =

SETD

1137 is orthogonal

IFW = IFW = IFW = IFW =

SETE orthogonal 7311

IFW = IFW = IFW = IFW =

SETF orthogonal 3711

IFW = IFW = IFW = IFW =

SETG orthogonal 1173

IFW = IFW = IFW = IFW =

SETH orthogonal 1137 Two basic DOC code characters in above-mentioned code character set SET A and SET B are provided individually below, wherein each code character contains the code word of 16 a length of 128.

Basic DOC code characters in DOC code character set SET A（N=16, R=4, M=2), transform sequence Q the cycle be（+ 1+1+1-1+1+1-1+1), become ^^ row ^ the cycle into（+1—1 ).Code word 1

(+++—++—++++—++— ++++_++—++++_++— ++++—++— ++++—++— ++++—++— ++++—++—+,

+-+++ -+- ++++- ++++- +++ -+- ++++- ++++- +++ -+- +++) code word 2

(++++- ++ H H-++ H+- ++++- ++++- ++++- ++++- ++++- ,+- ++ _ ++++- ++-++++- ++-++++- ++-++++- ++-++++- ++-++++- ++-++++_ ++-+++) code word 3

(+++ -+-+- mono- H " one +++-++ -++ ---+- +++ -+_++-mono- H "-+++-++ -++-mono-+-,

+-++++++++- +++++++- ++++++++- +++++++) code word 4

(+++++++-+++++ +++-+++++++-+++++++ -+,

+-++-+++ -+---++-++-+++ -+---++-++-+++ -+- mono-++-++-+++ ~+---+) code word 5

(+++-++-++++- ++ -++ one by one++ _ mono-+- +++-++-++++- ++ -++ one by one++ one by one

+-+++ -+- +++++++ -+++++- ++++- +++++++ -++++) code word 6

(++++- +++++++-+++++-+++ h +++ -++++ -+,

+-++-++++- ++-+++ -+one by one H+_ mono-++-++-++++- ++-+++ -+one by one++ -+) code word 7

(+++-++ -++ one by one++-mono-+- +++-++-++++- ++ -++ one by one++ one by one+- +++-++ -+₅H " one " ++ 1 h+H " one " I mono- " (-+H 1 " " h+H-of a h+H 1 " I h+++-++H) /:/:/ O ε99000ίοοίζ>1£ 9SS0S0iAV ) + + ++ ++ + + + ++ ++ ++ +++++++ ++ + + + + ++++ΙΙΙΙ Ι

(+++ +++ Bian +++ +++ +++ ^++Τ +++++ Τ +++ +++| Ι II Ι Ι Ι Ι Ι --- Ι Ι

6

)+ + + + + +++ + _++++++ ++++++++++++++++ -I -I————.---||---

(+++++ ++ +++++ ++++ ++ + ++ +++++++ ++++ΙΙΙΙΙΙΙ ) ++ +++++++ + +++++++ ++ +IIII

(+++++T++++++++++++++++++++++++ --- II II)++ ++++++ ++++++++Bian+--- --- IIIII--

(+T++ ++ +++ ++ _+Bian++++, I ++ +++ +++++++ +++ +++ +++++ Τ+Ι Ι Ι

(++++++—1 +++++++++++++ + +ΙΙΙΙΙ——— -

( ++ +++++++++ ++ +++++ ++++++ ++ +++ ++ +ΙΙΙΙΙΙΙ

$ cuns

}+ ++++ +++++++ ++++++ +++ ++++ +++ ++++ ONΙΙΙΙ-Ι—-■

^+ί++++++++++ + ++++++ + + +++++++ΙΙΙΙ——ΙΙ II 1-ΙΙΙ }++++++++ ++++++ + ++++++ + + +++ ++ ++++IIII -II—II—— 1-I-- (+++++++ -++++-+++++ +++-+++++-+++++++ -+,

+ _ ++ _ +++-H++ one by one++-++-+++ -+one by one++-++-++++- ++-+++ -+- mono-+)

Basic D0C code characters in D0C code character set SET B（N=16, R=4 ,=2), transform sequence Q basic cycle is (+1+1-1+1+1+1+1-1), transform sequence I_tBasic cycle be（+1 -1 )

Code word 1

(++-++++- ++-++++- ++-++++- ++-++++- ++-++++- ++-++++- ++-++++- ++-++++- ,++-++++- ++++- ++++- ++++- ++++- ++++- +++ h-++) code word 2

(++- +— +++- +— +++-+— +++- +— +++-+— +++- +— +++- +— +++- +— +₅

Code word 3

(++-++++— + +++-++++— + +++-++++ l· +++-++++— +

+ -+- ++-+++ -+one by one+-+- ++-+++ -+one by one+-+- ++-+++ -+one by one+-+- ++-+++ -+one by one) code word 4

(++ -++ one by one+- +++-++ -++ one by one+- +++-++ -++ one by one+- +++-++ -++ one by one+- +++-,

++h+++-++++++++- ++++++++- ++++++++- ++) code word 5

(++-++++- ++-++++μ+---++++-++++- ++-++++-++-mono-++, the H h+ of 1-h+H of H 1-I h+H --- 1 l^ of h+H of I 1 1-1 h+H of l+--I) code word 6

(++ -+- +++ -+-+one by one+- +++ -+- +++-++ -+- +++ -+-+one by one+- +++ -+- +++-₅

- 111 h++H HH --- 111 H-+H of I-+H-H-I h+H-h+) code word 7

(++-++++h+one by one++++-++++- ++-++++++- mono-++++-++++- )+++ T++ +++++ +++ ++++--- an I. iIII III

( ++ +++ + ++++++++++++ ++++ + ++++++++—II 1-II

) ί+— + + + ++τΐ + +++—++ !——-1

" ++ 4+++++ Bian ++ Bian+++ +++ ++ τ τ+--- 1 --- I ι!

)—+++†+ +—4 +—+++Ϊ +++ + +.. II———Ι Code word 16

(++ -+-+-+- +++ -+- +++-++ -+-+one by one+- +++-++ -+- +++ -+-+one by one+- +++-, ++- ++++- ++-++++- ++++- ++++- ++++--- h+- ++++- ++) can verify, the stupefied position of window of correlation function in both the above code character is equal at k points in relative skew, and wherein k is a nonzero integer；And the stupefied position of window of the correlation function of two interblocks is then equal at 8+4 points in relative skew, wherein being a nonzero integer.

To the two basic DOC code characters（It is not added with protection period between wherein the 1st and the 2nd subcode part, and is included in overlapping between the 1st and the 2nd subcode part）Different long PN sequences are respectively adopted and carry out scrambler（The length of wherein scrambler section is N=16), and scrambler sequence is produced by computer random.In the basic DOC code characters that code character set SET A are provided in Fig. 7, code word 1 and code word 10 are after scrambler, the mean-square value of correlation function value at each relative offset point.From Fig. 7, on the stupefied position of window（At ^ points, wherein ^：For nonzero integer）There is larger correlation function mean-square value between code word 1 and code word 10；And on the stupefied position of non-window, correlation function mean-square value is typically small（Particularly near origin）, can typically ignore in actual system design.It is noted that compared with previous example, although window is stupefied just to be occurred once at interval of 8 points, i.e. the stupefied number of window is than few half in previous example, but the mean-square value of each stupefied place's correlation function value of window is also about one times or so in previous example at respective point.This also reflects the real shield of DOC codes from a side --- and DOC codes are not reduced total jamming power at be possible to offset point, but jamming power has been focused on into the stupefied position of corresponding window up. '

The code word 1 in code character set SET A basic DOC code characters is given in Fig. 8, and the code word 1 in Ma Group set SET B basic DOC code characters, after different PN sequences scramblers, in each mean-square value with respect to correlation function value at offset point between them.From Fig. 8, on the stupefied position of window at point, wherein k is nonzero integer）There is larger correlation function mean-square value between the two code words；And on the stupefied position of non-window, correlation function mean-square value is typically small（Particularly near origin）, can typically ignore in actual system design. (4) the relevant consideration of DOC code characters is designed in real system：

The spread spectrum coefficient of system, unilateral IFW width（Or window it is stupefied between interval）Mutually restricted between available DOC codes number of codewords.If code word overall length is NRM, unilateral IFW width（It is also the interval of the stupefied position of window simultaneously）For RM -1, then it is N that the method described by the present invention, which can construct number of codewords contained in a DOC code character, the code character,.During design system, it is considered as above-mentioned each factor to choose suitable DOC code characters.If for example, spread spectrum coefficient be 128, it is desirable to unilateral IFW width be 3, then can design DOC code characters, its parameter be (N, R, M)=(32,2,2), that is, contained number of codewords be 32.For some communication situations, for example：Honeycomb（Microcell in), multidiameter delay is generally smaller, as long as so DOC code characters are designed, when making its unilateral IFW width more than maximum multipath time delay, it is possible to multi-access inference is completely eliminated.

On the other hand, in some communication situations（For example in great Qu cellular communication systems）And under some severe propagation conditions, multidiameter delay may exceed the unilateral IFW width of spread spectrum code character.For example, it is assumed that the section of CDMA systems（Chip) speed is 1.28Mcps, then the width each cut into slices is 0.78125 microsecond, and 3 slice widths are about 2.344 microseconds；And in some severe wireless propagation environments（Such as mountain area undulating topography）, maximum multipath time delay is up to 10 ~ 20 microseconds.In this case, it is less feasible that interference is eliminated by increasing IFW width, because so # zhangs can greatly reduce available code word number, so as to cause the decline of power system capacity.Now, when designing DOC spread spectrum multiple access code, it code word number is reached the basic capacity demand of system first, then be made between certain spread spectrum coefficient and IFW width compromise.As previously described, because IFW width can not possibly be very big, so the multipath component just positioned at the stupefied place of window is likely to result in larger interference.At this point it is possible to which eliminated by some other means in these interference, such as balancing technique, Pre-RAKE technologies or Multiuser Detection thousand disturb cancellation technology etc..Due to need to only eliminate these multi-path jammings on the stupefied position of window, without eliminating all multi-path jammings, so relevant interference cancellation techniques can obtain simplification to a certain extent.

(5) during system group network DOC code characters the method for salary distribution： Because the quantity of DOC code characters is very more, thus system group network be each cell/section distribute code character when, have more flexible mode.Two kinds of possible code character distribution methods are described below.

General, can be distributed for each cell/section has certain IFW characteristic and the stupefied position characteristic of relevant window in same basic DOC code characters, the Ma Group, can be used to define different code channels.Then some long PN sequence is distributed as the scrambler sequence of the cell/section for each cell/section.With reference to Fig. 9, wherein total all cells use identical code character set（Marked with SET A）, and the specific long PN scramblers of basic DOC code characters and the cell in the code character set（Marked with PNx）The multiple spread spectrum structure of composition, to determine the spread spectrum multiple access code used to each symbol.Distinguished between each cell using long PN scramblers.Here long PN codes can be some PN sequence（Such as m-sequence）Different skews or the good sequence of different correlation properties, such as Gold sequence.The effect of PN sequences is in addition to distinguishing the code character in each cell, also by the various interference positioned at the stupefied place of window（Including ISI, MAI and ACI) component albefaction.

Another code character method of salary distribution is as shown in Figure 10, wherein employing three different code character set altogether（SET A, SET B, SET C flags are used respectively）, code character set here be it is mutually orthogonal, neighbor cell uses different code character set；And each cell specific long PN sequences of cell（Marked with PNx）Carry out scrambler.Wherein, certain IFW relation (for example, having the IFW that One-sided Widths are 1 between code character set A and B) is there is likely to be between some code character set.The orthogonal property of code character even IFW characteristics between neighbor cell, part ACI interference can be eliminated to a certain extent.

It is a kind of spread spectrum multiple access code encoding device of the present invention as shown in Figure 11 a and figure lib, including：PN sequence generators, Walsh code generators, transform sequence generator, frequency divider, correlator；Wherein：

Output data after the output data of PN sequence generators is related to the output data of Walsh code generators again with transform sequence generator is related, obtains output data.

Described device, it is characterised in that described PN sequence generators refer to：There can be N number of PN sequence generators, described Walsh code generators refer to：There can be N number of Walsh codes generator, Described transform sequence generator refers to：There can be N number of transform sequence generator；Wherein：

Output data after the output data of 0th PN sequence generator is related to the output data of the 0th Walsh code generator again with the 0th transform sequence generator is related, obtains the 0th output data；Output data after the output data of n-th PN sequence generators is related to the output data of n-th Walsh code generators again with n-th transform sequence generator is related, obtains n-th output data；The 0th described output data is related to described n-th output data to obtain output data.

Device of the present invention can generate DOC (Distinctive Orthogonal Code) spread spectrum multiple access code, and the code character inner code word has zero interference window（Interference Free Window, IFW) characteristic, and only on some specific positions（Referred to as " window is stupefied "）There may be larger non-zero correlation.

So far, the coding method to a kind of spread spectrum multiple access code that interference in CDMA systems can be concentrated on to some ad-hoc locations and its equipment are had been completed.One of ordinary skill in the art should also be appreciated that, here it is used to describe various logic unit, module, circuit and algorithm steps of the present invention etc., electronic hardware (electronic hardware) can be used, computer software (computer software) or combinations thereof are put into effect.Here generally all it is to be described according to their function to various elements, unit, module, circuit and step, uses hardware or software when realizing actually, be concrete application and design constraint by whole system to determine.One of ordinary skill in the art should also be appreciated that the interchangeability of hardware and software under specific circumstances, and a class coding method described in the invention can be realized using best mode for concrete application.

For example, be used for various logic unit, module, circuit and the algorithm steps etc. for describing the present invention here, can in the following ways or their group is realized, including：Digital signal processor（DSP), specific use integrated circuit（ASIC), field programmable gate array（FPGA) or other PLDs, separation（Discrete) gate（) or transistor gate（Transistor) logic, the hardware component of separation（Such as register and FIFO), perform a series of firmwares（Firmware) processor of instruction, traditional programming software（Programmable software) and relevant processor (rocessor) etc..Wherein, processor can be microprocessor（), microprocessor can also It is traditional processor, controller（), controller microcontroller（) or state machine microcontroller（State machine) etc.；Software module may be present in RAM memory, flash memory（Flash memory), in ROM memory, eprom memory, eeprom memory, register, hard disk, moveable magnetic disc, CD-ROM, or any existing known storage medium.

One of ordinary skill in the art is obviously clear and understands, the preferred embodiment that the present invention is lifted is not intended to limit the invention only to illustrate the present invention, and the present invention is for the technical characteristic in each embodiment, can in any combination, without departing from the thought of the present invention.According to a kind of coding method of spread spectrum multiple access code that interference in CDMA systems can be concentrated on to some ad-hoc locations disclosed by the invention and its generation equipment, modification disclosed invention can be there is many ways in which, and in addition to the above-mentioned preferred embodiment specifically given, the present invention can also have other many embodiments.Therefore, all category are according to the present inventive concept getable method of institute or improvement, within the interest field that should be included in the present invention.

Claims (1)

1. Claim

   1. a kind of spread spectrum multiple access code encoding method, it is characterised in that comprise the following steps：

   Any orthogonal sequence group is chosen for initial orthogonal sequence group；

   Element of each sequence in described initial orthogonal sequence group on everybody is repeated in, another sequence group is obtained；

   Generate transform sequence Q;

   By each sequence in described another sequence group respectively with described transform sequence Q wise multiplications, obtain a subcode part of DOC codes；

   Generate transform sequence I；

   By each subcode in a described subcode part respectively with described transform sequence I wise multiplications, obtain another subcode part of DOC codes；

   Described subcode part is combined, DOC Ma Group are obtained.

   2. the method according to claim 1, it is characterised in that any Zheng of described selection hands over Xu Lie Group to be that initial orthogonal sequence group refers to：The sequence group W that one group of orthogonal sequence containing N number of a length of N can arbitrarily be chosen is initial orthogonal sequence group；Wherein： W = {Wp W₂、 …、 D。

   3. the method according to claim 1, it is characterised in that the element for founding each sequence in described initial orthogonal sequence group is repeated in, and is obtained another sequence group and is referred to：

   When described initial orthogonal sequence group is W={ W W₂..., when, can by each sequence w a=i in W, 2, N) element on everybody repeats R-i times respectively successively, comprising

   N number of a length of NR another sequence group A={ A_P A₂、 …、 A ;

   Wherein, it can set：

   

   And：

   Α_;· = (a a_!)2 … Ά_ίΜ) (i=1,2, N) then have： (z'=l, 2 ", N; j = 0,l,"',N- 1 ) .

   4. according to the method described in claim 1, it is characterised in that transform sequence Q is resulting after the length is repeated NI2 times for 1R basic cycle, and resulting transform sequence Q length is NR.

   5. according to the method described in claim 1, it is characterised in that each sequence by described another sequence group is respectively with described transform sequence Q wise multiplications, and the subcode part for obtaining DOC codes refers to：The sequence group W of one group of orthogonal sequence containing N number of a length of N can arbitrarily be chosen for initial orthogonal sequence group, and W={ W_n W₂、 …、 W^} _;And element of each sequence (i=1,2, N) in W on everybody is repeated R-1 time respectively successively, obtain one group of another group of sequence group for including N number of a length of NR=, A₂、 …、

   Each sequence in described another sequence group A is entered, obtains the 1st subcode part C of DOC codes with transform sequence Q wise multiplications respectively¹ = {C¹;, i=1,2 ,-^i }, that is,：

   C\ = A_; « Q G' = 1, ",N)

   Wherein, symbol " " represents the wise multiplication of two sequences.

   6. the method according to claim 1, it is characterised in that include the step of described generation transform sequence I：Described transform sequence I can be generated after repetition by some basic cycle, and described transform sequence I may include multiple sequences： ι_Ρ 1₂、 …。

   7. the method according to claim 1, it is characterised in that each subcode by a described subcode part is respectively with described transform sequence I wise multiplications, and another subcode part for obtaining DOC codes refers to：A described subcode part and a series of transform sequence I successively wise multiplication can be obtained other multiple subcode parts of DOC codes, described transform sequence I may include multiple sequences： I

   1₂、 …；

   Wherein：When the transform sequence wise multiplication of each subcode in a described subcode part respectively with an a length of NR, when obtaining described another subcode part, it can set：

   A described subcode part is the 1st subcode part C¹, described another subcode part is the 2nd Individual subcode part C²,

   

   Then have：By the described the 1st, 2 transform sequences 1 of subcode part respectively with another a length of NR₂Wise multiplication, and respectively obtain the 3rd, 4 subcode part C³And C⁴, that is,：

   C = C *l₂ ϋ=1,2 ··,Α0

   

   Aforesaid operations can constantly go on, general, when having obtained individual subcode part：

   C¹, C²、 …、 C^L, wherein：L=ll, J' are nonnegative integer,

   Then can be by C C²、 …、 （The transform sequences 1 of ^ respectively with an a length of NR_£Wise multiplication, can obtain another subcode part (^⁺¹、 C^L+\ …、 C^2L, that is,： i=Y,2 ",N、 j=l,2,-,L)

   When aforesaid operations have been carried out time altogether, and transform sequence 1 is used successively₂..., 1, then can generate M=2 altogether^MIndividual subcode part C C²、 …、 C^M, wherein N number of a length of NR subcode code word is contained in each subcode part.

   8. according to the method described in claim 1, it is characterised in that described that described subcode part is combined, obtain D0C code characters and refer to：M subcode part can be obtained to D0C code characters { the DOCp DOC containing N overall length for NR code word by combination₂..., DOC, wherein：

   DOQ=(C C, C^.) G=1,2 ", N)

   Thus generate a D0C code character.

   9. method according to claim 4, it is characterised in that：Transform sequence Q basic cycle is obtained based on a kind of mode of spanning tree；Wherein：

   Contain one " basic cycle " on each node in the spanning tree, corresponding transform sequence Q obtains " basic cycle " after repetition；=0,1,2 is arranged in the r of spanning tree ...）Containing a length of

   

   " basic cycle ", and corresponded to each sequence in initial orthogonal sequence group Every repetition R -1=2''- 1 time when situation；If initial orthogonal sequence group contains N number of a length of N orthogonal sequence altogether, then obtained by transform sequence Q is after the length is repeated M2 times for 2R " basic cycle ", resulting transform sequence Q length is NR

   A kind of producing method based on spanning tree for producing transform sequence Q " basic cycle " can be described as follows：In the spanning tree, the basic cycle contained by root node on the 0th row is（+1+1);Spanning tree can be constructed in the following manner：If some node in r row contains a length of 2 2R " basic cycle " P^Q, it is by subsequence a respectively for 2=^ two sections long^QAnd b^QComposition, i.e.,：

   P°= (a⁰, b°)

   Then in spanning tree, by the node it is derivative, positioned at r+1 arrange two nodes, respectively containing a length of 2²=4 two " basic cycle " P¹And P², Fen Do are：

   (a¹, b¹)

   = (a⁰, b°, a⁰, -b°)

   P²= (a², b²)

   = (a⁰, -b°, a⁰, b°)

   According to the recurrence relation, it is possible to obtain a kind of generation tree construction of transform sequence Q basic cycle.

   10. method according to claim 4, it is characterised in that：

   Described transform sequence Q can be obtained by basic cycle sequence P after repetition；

   Wherein, each sequence repeats the situation of R-1 times by turn in the initial orthogonal sequence group described in correspondence, can set Ρ_2Ώ?For the basic cycle that length is IfiTR, wherein be some positive integer, it be by it is front and rear two sections long be KR subsequence P, P²What ^ was constituted, and P and it is each be made up of segment length for R small tract, that is,：

   P2O = {P】AS, ^KR] and

   pijo? = {^C c²/...,^c

   P²¾ = {^{S S2}R,^S Wherein, and s (i=l, 2 ,-', K is a length of R small tract, and meets between s following orthogonality relation：

   (c^)(s^)^T = 0 (ί = 1,2 -,Κ)

   Wherein, Τ represents the transposition carried out to row vector；It is complementary series with the sequence described in the structure；Described transform sequence Q basic cycle must be a complementary series.

   11. method according to claim 6, it is characterised in that：

   Each a length of NR transform sequence I can by an a length of 2L basic cycle X_2iObtained after repeating NR/2 times, here=1,2 ...；Described basic cycle X_2iCan by continuous "+1 ", immediately continuous L " -1 " constitute, or can immediately continuous L be individual by continuous L " -1 " "+constitute.

   12. according to the method described in claim 1, it is characterised in that described DOC code characters refer to：Overall length is the DOC code characters of NRM code word；In DOC code character of the overall length for NRM code word, parameter used is N, R, M；A series of DOC code characters comprising different number of codewords, code word size and pertinence shield can be generated using different parameter combinations.

   13. according to the method described in claim 1, it is characterized in that, described transform sequence Q and described transform sequence I can carry out a variety of conversion using multiple combinations, so that the DOC code words generated are met only non-zero correlation on the stupefied position of window, and in the presence of a zero correlation window near origin；Said process at least needs some transform sequence Q or transform sequence I to carry out linear transformation.

   14. the method according to claim 1, characterized in that, using the initial orthogonal sequence group described in different, and transform sequence Q and transform sequence I described in identical, a collection of DOC code characters can be constructed, this batch of code character can be claimed to constitute a code character set；Each described code character set is corresponding with one group of transform sequence Q and I.

   15. according to the method described in claim 1, it is characterised in that M described subcode part should be in different mutually orthogonal synchronous fading channels；I.e.：

   M described subcode part can be sequentially allocated to M continuous time slots；In order to avoid each son It is overlapping between code part, the protection period that Ke Fen Do are made up of in their end addition full neutral element.

   16. a kind of application process of the DOC codes based on claim 1 methods described, it is characterised in that scrambler is carried out to the code word in DOC code characters, i.e.,：By the sequence wise multiplication of each code word in DOC code characters respectively with an equal length.

   17. method , No described in 16 are just sought again in paste the sublingual row of horse ^ and walk the smiling scrambler that can intercept a length of N successively from long PN sequences section；

   It will be repeated in R-1 times per bit element in each described scrambler section, obtain an a length of NR sequence；

   Ran integrally repeats a described a length of NR sequence M -1 times, obtains a sequence of a length of N abdomens, i.e. scrambler sequence；Described scrambler sequence is that described scrambler section is constituted by M sections of identical sequences；

   By the M sections scrambler section in described scrambler sequence, wise multiplication is distinguished with M subcode partial codeword in DOC code words respectively, so as to complete scrambler process.

   18. method according to claim 17, it is characterized in that, in the construction process of described DOC codes, Walsh orthogonal sequences group can be chosen as initial orthogonal sequence group, the DOC code characters produced by this orthogonal sequence group based on Walsh can be described as basic DOC code characters.

   19. method according to claim 18, it is characterised in that in a code character set, all code characters can be regarded as what is obtained by a basic DOC code character after different scrambler section scramblers, i.e.,：Basic DOC code characters after scrambler, remain in that the relevant IFW characteristics and the stupefied position characteristic of window of original code character.

   20. method according to claim 18, it is characterised in that when application DOC codes and long scrambler carry out spread spectrum operation, multiple spread spectrum mode can be used.

   21. method according to claim 20, it is characterised in that described multiple spread spectrum refers to：Two-stage can be divided into be spread：The first order is to carry out ■ frequencies using basic DOC code characters, wherein each basic DOC codes define a code channel；The second level is using by disturbing obtained by the scrambler step The basic DOC codes of code sequence pair carry out scrambler.

   22. a kind of method that D0C codes based on claim 1 methods described carry out D0C code character distribution in cellular communication system network, it is characterised in that：It is that a DOC code character is distributed in each cell or sector.

   23. method according to claim 22, it is characterised in that：For each cell or the sector distribution basic D0C code characters of identical one；Meanwhile, each cell or sector are used as scrambler using different long PN codes.

   24. method according to claim 22, it is characterised in that：It is that several different basic DOC code characters are distributed in adjacent cell or sector；Wherein, these basic DOC code characters belong to different code character set；Meanwhile, each cell or sector are used as scrambler using different long PN codes.

   25. a kind of spread spectrum multiple access code encoding device, including：PN sequence generators, Walsh code generators, transform sequence generator, frequency divider, correlator；Wherein：

   Output data after the output data of PN sequence generators is related to the output data of Walsh code generators again with transform sequence generator is related, obtains output data.

   26. device according to claim 25, it is characterised in that described PN sequence generators refer to：There can be N number of PN sequence generators, described Walsh code generators refer to：There can be N number of Walsh codes generator, described transform sequence generator refers to：There can be N number of transform sequence generator；Wherein：

   Output data after the output data of 0th PN sequence generator is related to the output data of the 0th Walsh code generator again with the 0th transform sequence generator is related, obtains the 0th output data；Output data after the output data of the N number of PN sequence generators of the mat woven of fine bamboo strips is related to the output data of n-th Walsh code generators again with n-th transform sequence generator is related, obtains n-th output data；The 0th described output data is related to described n-th output data to obtain output data.