WO2004077714A1

WO2004077714A1 - Low correlation/ zero correlation spread spectrum code sets coding and the method of its application

Info

Publication number: WO2004077714A1
Application number: PCT/CN2003/000152
Authority: WO
Inventors: Shaojun Xu; Yan Gao; Rakesh Tamrakar
Original assignee: Linkair Communications,Inc.
Priority date: 2003-02-26
Filing date: 2003-02-26
Publication date: 2004-09-10
Also published as: AU2003211674A1; CN1647431A

Abstract

A coding of low correlation/ zero correlation spread spectrum code sets and method of its application, which has NK spread spectrum codes altogether, said NK spread spectrum codes are grouped into N code sets, and each set comprises K spread spectrum codes; wherein said group satisfies following conditions: there are low correlation/ zero correlation characteristics among N grouped code sets. Said method of applying low correlation/ zero correlation spread spectrum code sets can effectively eliminate interference from adjacent cell and adjacent sector, and reduce ACI significantly. Thus the capacity of the system can be improved.

Description

Low correlation / zero correlation spreading code set encoding and application method

Technical field

The present invention belongs to the field of wireless communication technology, and particularly relates to a CDMA wireless communication system, and specifically relates to a low-correlation / zero-correlation spreading code set encoding and application method.

Background technique

With the continuous popularization of personal communication services and the relative scarcity of wireless spectrum resources, people have increasingly higher requirements for spectrum utilization in wireless communications. Traditional access methods, such as FDMA (Frequency Division Multiple Access) and TDMA (Time Division Multiple Access), have become increasingly incapable of meeting the needs of modern wireless communications, due to their inefficient spectrum utilization. The new access method, CDMA (Code Division Multiple Access), has been recognized as the main access method in next-generation wireless communications due to its high spectrum utilization.

The difference between Cili A and other access methods lies in: The CDMA system is a soft capacity. Its capacity depends on the interference level of the system. Any means that can reduce the interference can increase the capacity of the CDMA system. The capacity of other access methods such as FDMA and TDMA systems is a hard capacity, and its capacity has been determined after the system is established.

Because the capacity of a CDMA system depends on the interference level of the system, how to reduce the interference in the system is very important for the CCA system. There are three main types of interference in a CDMA system: one is inter-symbol interference (ISI), the other is multiple access interference (MAI), and the third is adjacent cell interference (ACI). Inter-symbol interference is mainly caused by the non-ideal autocorrelation of spreading code words, and multiple-access interference is mainly caused by

caused. Designing a spreading code group with good auto-correlation and good cross-correlation characteristics is very important for CDMA systems. The ideal autocorrelation means that the autocorrelation function of the spreading codeword is zero at all time offsets except the origin. The ideal crosscorrelation means that the crosscorrelation function between the spreading codewords is offset at all times. Zero everywhere. However, according to the Welch bound, the autocorrelation function and the crosscorrelation function are a contradiction. Decreasing one inevitably causes the other to increase. Therefore, the codeword that the autocorrelation function and the crosscorrelation function are completely ideal at the same time does not exist. In traditional CDMA systems, the spreading sequence uses a common orthogonal sequence. For example, in IS-95, the spreading sequence uses a Wal sh sequence. This codeword remains orthogonal only at the origin, and the correlation value outside the origin is not zero, because the mobile channel is a multipath channel. In the channel, the system of this code will inevitably introduce ISI and MAI, which makes the capacity of the system impossible; L

In addition to the WALSH code, some systems also use the GOLD code. The GOLD code is a non-orthogonal code, but has better secondary peaks (correlation functions outside the origin) than the WALSH code. Limited by the WELCH bound, the secondary peak of the GOLD code is still relatively large. Therefore, the CDMA system using the GOLD code is still an interference-limited system. .

The inventor Li Daoben proposed a new type of spreading code-complementary zero correlation window code, LS 'code in the application patent application number PCT / CN00 / 00028. Assuming the maximum time spread of the channel is Δ, this codeword guarantees the correlation characteristics within [-Δ, Δ] is ideal. The system using this codeword eliminates ISI and MAI and greatly increases the capacity of the system. A system using this codeword as the core address code is called a LAS-CDMA system.

Whether it is a WALSH code, a GOLD code or an LS code, the problem is how to reduce or eliminate the ISI and MAI of the CDMA system, that is, the problem of single-cell operation. In a cellular mobile communication system, we also need to solve the problem of how multiple cells / multi-sectors work together, that is, how to solve the problem of interference from neighboring cells / sectors.

The above-mentioned WALSH code, GOLD code, and code are collectively referred to as ShortCode. In IS-95 and other traditional CDMA systems, do short pseudo-random (PN) sequences be used for short spreading codes? Horse (scrambling). In the same cell, ^ T frequency codes (WALSH code or GOLD code) are used to distinguish different channels, which are called channelization codes. Different cells use different long PN sequences to scramble codes, which are called long codes. In fact, this method of scrambling does not reduce the interference of the system, but only makes the interference suffered by the users more homogeneous, thus ensuring that the users in the system have the same communication quality.

In the LAS-CDMA system, a network of cells is formed by using a LA code + LS code. The LA code is a specially designed multi-pulse long code. The LA code itself has good correlation characteristics, and the correlation characteristic has a large zero correlation window. LS code as short code is inserted every pulse of LA code Here. In this way, the LS code is used to distinguish different channels in the same cell, which is called a channelization code, and different cells use different LA codes to distinguish. Because the LA code has good cross-correlation characteristics, the LAS-C belly-A system can well suppress interference from neighboring cells.

In the LAS-CDMA system, because there is no interference between users in the same cell, in the downlink, users at different locations can perform effective power allocation: users who are far from the base station allocate a larger transmission power, and the distance Users near the base station allocate less transmit power, and ultimately all users in the cell maintain the same signal-to-interference ratio level. This can reduce the base station's transmit power by approximately 3-5dB, and at the same time reduce adjacent cell interference by ^ 氐 3-5dB. In the traditional CDMA system, due to interference between users in the same cell, effective power allocation cannot be implemented in the downlink. Otherwise, high-power users will seriously interfere with low-power users.

In addition to the various interference reduction technologies mentioned above, there is another important technology that is Dynamic Channel Allocation. The idea of dynamic channel allocation is as follows: Use measurement technology to perform interference measurement on the channel where the user is located and other free channels. If a user is in a serious interference state, the interference on the other free channel is relatively Smaller, you can consider switching this user to this idle channel. Dynamic channel allocation can reduce the interference of CDMA system to a certain extent, but its role is limited, and there will be certain problems in implementation. The main limitations of dynamic channel allocation are: (1), dynamic channel allocation is very limited to reduce system interference, especially when the user is almost full, when there are not many free channels left, when a user is affected by In severe interference, a better idle channel can no longer be switched, and dynamic channel allocation has lost its role at this time. (2) The dynamic channel allocation algorithm requires a complex set of measurement algorithms, allocation algorithms, and control algorithms, which will greatly increase the complexity of the system. (3) The dynamic channel allocation algorithm requires users to switch frequently during the communication process, which not only greatly increases the complexity of the system, but also requires a large number of control instructions, which will occupy a lot of system resources and cause huge waste. .

In 1998, K. Hamidian and L Payne co-published a paper "Frequency reuse appl ied to CDMA channels in a cel lular system". In this thesis It mentioned a new type of cellular networking implemented in traditional CDMA systems-F should be A / CDMA. This new type of networking can be briefly described as follows: First, each cell is divided into a ring-shaped area, and we label them as: AA ₂ , ……, k _N. In the same cell, users in different rings use different frequencies. We mark them as:, f ₂ , ..., f _N , and perform power allocation between different rings, for example, in a ring that is closer to the base station. Of users allocate less transmit power, while users in a ring farther from the base station allocate greater transmit power. And between adjacent cells, different frequency arrangements are used to ensure that the outer ring users of adjacent cells have different frequencies, so that users at the border of each cell are not affected by the interference of outer ring users of adjacent cells ( And it transmits with a higher power transmission power), and is only affected by the interference of the ring users in the adjacent cell (while it transmits with a lower transmission power), so that the ACI of the border users can be reduced, and the users closer to the base station, because The attenuation factor of distance is relatively small. Such a system can greatly reduce the ACI of the system. Take a simple example to illustrate this method: Suppose we divide each cell into three concentric rings, labeled A _p A ₂ , A ₃ from the inside to the outside. There are three frequencies in total: f, f ₂ , ί. We only consider the case of three communities. More communities can be deduced by analogy. Suppose that the frequency in the first cell is arranged from the inner ring to the outer ring in the order of / ₂ , ₃ , and the frequency in the second cell is arranged from the inner ring to the outer ring in the order of ₂ , f ₃ , f, third The frequencies in each cell are arranged from the inner ring to the outer ring in the order of / ₃ , f, and f _{2. In} this way, there is no interference between users in the same ring of adjacent cells. Interference from users in the first ring or the second ring of the neighboring cell, and due to the power allocation, the transmit power of the users in the first ring and the second ring is relatively small, so the users in the neighboring cell receive the neighboring cells ^ d ^ In addition, users in the inner ring can also get good performance due to different path attenuation between users in different cells.

The FDMA / CDMA cell networking scheme proposed by K. Hamidian can effectively solve the problem of interference in neighboring cells, but since it was proposed, it has not received enough attention. Mainly because in the traditional CDMA system, in addition to ACI, there are also severe ISI and MAI between users in the same cell. Therefore, even if the problem of ACI is effectively solved, the problem of ISI and MAI cannot be solved. The capacity of the system is still limited by ISI and MAI, so the system's The capacity can only be increased slightly. In the downlink, the capacity of the FDMA / CDMA system is approximately 1/7 higher than that of the conventional CDMA system, and in the uplink, the capacity of the FDMA / CDMA system is approximately 5.8-20% higher than that of the traditional CDMA system. Considering the overhead of the system and the complexity of the system, there is no big advantage to using FDMA / CDMA.

SUMMARY OF THE INVENTION The object of the present invention is to provide a low-correlation / zero-correlation spreading code set encoding and application method, wherein the low-correlation / zero-correlation spreading code set encoding method can be briefly described as follows: 殳 A total of one spreading code Then, according to a certain rule, this spreading code is grouped and divided into different code sets. We label them as SET, SET, SET _N , where each code set contains a spreading code, and we label it as CODE. , ... C0DE _K. This represents the: spreading code in the first code set as CODE. The grouping criteria are as follows: The grouped code sets have good cross-correlation characteristics, that is, the cross-correlation function between any two spreading codes from different code sets is maintained within a relatively large time offset around the origin. Good correlation characteristics (low or zero correlation). There are no excessive requirements for the codes in the same code set. It can be an orthogonal code set, a window code set, or even a non-orthogonal code set.

The application method of the low correlation / zero correlation spreading code set can effectively suppress interference from neighboring cells and interference from neighboring sectors through channel allocation, power allocation, and cell / sector networking, and can greatly reduce ACI has greatly improved the system capacity.

The technical solution of the present invention is:

A low-correlation / zero-correlation spreading code set encoding method, which is characterized in that: a total of spreading codes are set, and the NK spreading codes are grouped into V code sets, and each code set contains f Spreading codes

The grouping described therein meets the following conditions: The code sets after the grouping have low correlation / zero correlation characteristics. The feature of low correlation / zero correlation between the grouped code sets means that: the cross-correlation function between any two spreading codes from different code sets is within a relatively large time offset around the origin. Keep low correlation / zero correlation characteristics; and the codes in the same code set can be: an orthogonal code set, or a zero correlation window code set, or a non-orthogonal code set.

The low correlation / zero correlation spreading code set encoding method includes the following steps:

Step 1: Suppose there are spreading codewords, where N, can take any positive integer, and grouping will be performed on this basis;

Step 2: Group the spreading codes, divide them into code sets of the same size, and mark them as SET \, ST ₂ SET. Each code set has f code words, and mark them as CODE ^ C0DE ₁ C0DE. _K , so that the first codeword in the second code set can be expressed as, SET _n , C0DE _k

Step 3. The grouping is performed according to the following conditions: grouping makes different code sets have a correlation characteristic of low correlation or zero correlation within a large time offset around the origin, that is, any two codes from different code sets. Have low correlation or zero correlation between them, and the codes in the same code set can be orthogonal sequences, zero correlation window sequences, or non-orthogonal sequences;

Step 4. A code set is obtained from the above steps, and each code set contains a code word. The above description is to obtain a code set of the same size. Of course, it can also be divided into code sets of different sizes, but in general, it is used. All the code sets are of the same size, which is the case in the following description.

Given any of the above-mentioned spreading code words, it is not easy to divide the spreading codes into groups according to the above requirements according to the above grouping requirements. However, we can construct the above-mentioned low-correlation / zero-correlation spreading code set through a specific construction method. The present invention provides a method of constructing a low-correlation / zero-correlation spreading code set, and the construction methods include time-domain construction, frequency-domain construction, code-domain construction, time-frequency joint construction, time-code joint construction, and frequency-code joint construction in time. Similar construction methods such as frequency code joint construction.

Code domain construction: Through special coding methods, different code sets are distinguished from the code domain, so that the code sets have good cross-correlation characteristics. The low-correlation / zero-correlation spreading code sets distinguished in the code domain are as follows. It is called code division low correlation / zero correlation spreading code set. Code domain construction includes the following steps:

(1). Construct a pair of unrelated general complementary codes with complementary parts, and the length of each complementary part is L. (For specific construction methods, please refer to [1, 2, 5]).

Code words S _L5 S ₂ ,…, S, length Z, except for the zero point, if the sum of the autocorrelation of the code words is all zero at all offsets, then the code word (S _L5 S ₂ ,…, S is a generalized complementary code with complementary parts.

Let two general complementary codes with complementary parts: (S s ₂ .. ·, s), (s, s ₂ ² , ···, s), if the complementary parts of the two are correlated with each other The sum is zero at all time offsets, then the two generalized complementary codes are said to be uncorrelated.

For a generalized complementary code with complementary parts, the following conclusions hold: There are at most N pairs of generalized complementary codes with complementary parts that are unrelated. (See references [1,2, 5]) We write this generalized complementary code as: (s, s ₂ \…, s), (S, S ₂ ² ,. ·., S, ...... When N = 2, the above codeword is a complementary code that is often said, and there are at most two mutually unrelated complementary codewords. When N> 2, the above codeword is called a generalized complementary code. There are at most two unrelated general complementary codes.

When the length of each complementary part 1 = \ is a special case, the matrix formed by the generalized complementary codes is actually an orthogonal matrix, that is, matrix A:

When N, Z takes a relatively small value, # generalized complementary codes can be obtained by computer simulation search. When N, takes a relatively large value, it can be obtained through a certain construction method (see the literature

[1,2,5]).

(2) Select an orthogonal matrix H of χΓ

(3) Through the above two steps, we have obtained a pair of complementary parts that are not related in pairs. Generalized complementary codes of points: (s ^, s ₂ …, s), (s, s ₂ ² , s), ..., (s, s, .., ...

S), an orthogonal matrix H ^^ with code length Z and a Γχ. For each generalized complementary code, the orthogonal matrix is used to perform Kronecker expansion to obtain an extended spreading code set. The spreading code set contains spreading code words. The specific spreading method is as follows:

_{SET = μ ΚχΚ 05 /, n} KxK ®s \ - n KxK ®s N l)

_{_{SET 2 = (u KxK ®s Ή}} ΚχΚ ®sl - n KxK ®)

ns h _l2 s… ½s where '(¾' represents the operation of Kronecker, such as: H ^ ®S = 1 ^S ½ ^S "· ^{! l} 2K ^S ^h Kl ^{S h} K2 ^S ··· ^h KK ^S. For each extension The spreading code set SET _n , 1 n N, whose ^ "" line represents an extended codeword, then each spreading code set contains a spreading codeword, each spreading codeword has a complementary part, each The spreading code sets obtained by changing the length of the complementary part to KL have the following properties:

(a) For the K spreading codewords in each spreading code set, there is a zero correlation window of [-+1, Z-1], that is, Γ spreading codes within [-Z + 1,-1] Words have ideal auto-correlation and cross-correlation. If = 1, then the spreading codewords in each code set after the expansion only remain orthogonal.

(b) For different spreading code sets, at all time offsets, the codes in the two code sets have zero correlation characteristics. That is, the cross correlation between any two codes from different code sets in [-+1, -1] is zero.

(4) A spreading code set has been obtained from the above three steps, and each code set has f spreading codes. Among them, the codewords from different code sets have zero correlation characteristics, and the codewords inside the codeset can be zero-correlation window codewords or ordinary orthogonal codewords. Each of the generated spreading codes is a generalized complementary code having a complementary part, and the complementary parts are required to be calculated separately when performing correlation calculation, and cannot overlap. So in practice, the complementary parts should be transmitted in N synchronously fading channels. We can send the complementary parts in time slots, with guard intervals inserted between adjacent time slots. This may not be very efficient. We can also transfer the middle insurance The guard interval is canceled, so although the zero correlation characteristics are lost between the code sets, the low correlation characteristics are still maintained. Time-domain construction: Different code sets are distinguished from the time domain. The generated spreading code set is called a time-division low-correlation / zero-correlation spreading code set. The time domain construction method includes the following steps:

(1) Suppose there are f spreading codewords. This spreading codeword may be a zero correlation window code, or a general orthogonal codeword, or even a non-orthogonal codeword.

(2) A frame length is rammed into different time slots, and the width of each time slot is T, and the guard interval width between the time slots is Δ, as shown in FIG. 1.

(3) The spreading code is divided into different spreading code sets according to time slots. The first time slot represents the first spreading code set. At this time, all the spreading codes are sent in the first time slot, and The other time slots do not send anything. The spreading code sets constructed at this time are shown in Figure 2:

(4) For the spreading code sets divided as shown in FIG. 2, the time offset is within [-Δ, Δ], and the cross-correlation between the two different spreading code sets is zero, even if the time offset Beyond Δ, if the overlapping portion of the two time slots is not very large, the two different spreading code sets still maintain low correlation characteristics.

(5) From the above steps, N low correlation / zero correlation spreading code sets have been obtained, and each spreading code set has Γ spreading codes.

The time domain construction has some advantages over the code domain construction: (a). The codeword constructed by the time domain construction method has higher efficiency than the code domain construction method, because if the time width occupied by each time slot is relatively large, The time width occupied by the middle guard slot is relatively small, so it has high transmission efficiency. For the code domain construction method, a guard interval needs to be inserted between the complementary parts, which requires more resources to be sacrificed. (B) The time domain construction method is simpler and more flexible than the code domain construction method. Frequency code, we can all assign it to different time slots to obtain a 4's correlation / zero correlation spreading code set, which is easy to obtain for any integer, but not so easy for the code domain construction method For example, when the number is odd, it is impossible to find a generalized complementary code with complementary parts that are uncorrelated in the binary field. In this way, a low correlation / zero correlation spreading code set cannot be constructed by using the above method.

(6) In the above-mentioned time domain structure, the same number of spread spectrums are transmitted in each time slot. In fact, in the time-domain construction method, each time slot can send different spreading codes. For example, different time slots can send a transform of the spreading codes, or even completely different spreading codes. The divided spreading code sets still maintain low correlation / zero correlation characteristics.

/ Zero correlation spreading code set, the resulting code set is called frequency division low correlation / zero correlation spreading code set. The frequency domain construction method includes the following steps:

(1) Suppose there is a spreading codeword, which is respectively labeled as COD ₂ , ......, CODE. The Γ spreading codewords can be zero correlation window codes, or ordinary orthogonal codewords, or even non-standard codes. Orthogonal codeword

(2) select unrelated _{carrier: i, f 2, ......,} f N;

(3) Modulate each of the spreading codewords onto the aforementioned uncorrelated carriers to obtain an uncorrelated spreading code set: SET f _x + {C0DE _x , C0D ₂ , ..., C0DE, SET ₂ : F ^ iCODE ,, C0DE ₂ , ......, CODECS,. ·····, SET _h : f _N + {CODE ^ C0DE ₂ , ......, C0DE _k }

(4). An irrelevant spreading code set is obtained through the above steps, and each code set has a spreading code. Frequency-division low-correlation / zero-correlation spreading code sets have the following advantages: (a) Because the spreading code sets are modulated on different carriers, different spreading code sets have low correlation characteristics in all time offsets. (B) It is not necessary to insert a guard slot for a low-frequency / zero-correlation spreading code set, so its efficiency is higher than the time-division spreading code set and the code-division spreading code set. However, since different spreading code sets are modulated on different carriers, and there are a total of carriers, implementation problems will be encountered in practical applications.

In addition to the three construction methods described above, the present invention also includes the joint construction methods of the above three construction methods: including time-frequency joint construction, time-code joint construction, frequency-code joint construction, and time-frequency code joint construction.

Code time structure: It combines the above code domain structure and time domain structure to generate a set of low-correlation / zero-correlation spreading code sets. The code time structure provided by the present invention includes the following steps:

(1) Using the above code domain construction method to construct ^ spreadings with low correlation / zero correlation The code sets are labeled as A _l5 A ₂ ,..., A _NI , where each spreading code set contains a spreading code. (2) Consider the above-mentioned ^ spreading code sets as a whole, refer to the foregoing time domain construction method, and place them in ^ timeslots for transmission, so that = ^>< ₂ low correlations / Zero correlation spreading code set. The spreading code set is:

SET k _x + time slot 1 Γ ₂ : A + time slot 2 ... SET _Ni : Aj + time slot

SET _{N2 + X} : A ₂ + time slot 1 SET _{N2 + 2} : A ₂ + time slot 2 ... SET _2NI : A ₂ + time slot N ₂

SET _{MNI + L} A + slot 1 SET _{MNI + 2} : + slot 2 ... SET _NXN Aw, + slot N ₂

(3) A low correlation / zero correlation spreading code set is obtained from the above steps, where each spreading code set includes Γ spreading codes.

Time-frequency structure: It combines the above-mentioned time-domain structure and frequency-domain structure to jointly generate a set of low-correlation / zero-correlation spreading code sets. The code time structure provided by the present invention includes the following steps:

(1) A spreading code set is included, including spreading codes, denoted as A

(2) Modulate the above spreading code set onto ^^ uncorrelated carriers / ^., To obtain ^ uncorrelated spreading code sets, respectively: Α ^ Α + ΙΙ, Α ₂ = Α + / ₂ , ..., A _N] = A + f _Ni

(3) Consider the above-mentioned spreading code sets as a whole, and refer to the foregoing time-domain construction method, and place them in N ₂ time slots for transmission, so that N = Ν, χN ₂ low correlations can be obtained. / Zero correlation spreading code set, the spreading code set is:

SET k _x + time slot 1 Γ ₂ : + time slot 2 ...... SET _Ni : A _: + time slot

SET _N ^ _X : A ₂ + time slot 1 SET _{N 2} A ₂ + time slot 2 ... ET _2NI : A ₂ + time slot W ₂

^SET (N,-: A _Wi + time slot 1 SET _{MNI + 2} : Aw, + time slot 2 ... ... SET: + time slot N ₂

(4) A low correlation / zero correlation spreading code set is obtained from the above steps, where each spreading code set includes spreading codes.

Code frequency structure: It is a combination of the above code domain structure and frequency domain structure to jointly generate a set of low correlation / zero correlation spreading code sets. The code time construction includes the following steps: (1) Use the above code domain construction method to construct Λ spreading code sets with low correlation / zero correlation, which are respectively labeled as A ₁₃ A ₂ , ..., A ^, where each spreading code set contains f Spreading codes.

(2) Taking the above-mentioned ^ spreading code sets as a whole, referring to the foregoing frequency domain construction method, modulating them into _two unrelated carriers for transmission, so that we can obtain

Low correlation / zero correlation spreading code set, the spreading code set is as follows:

SET ,: k, + f _x SET ,: A, + / ₂ ... SET _N2 k, + f _Ni

SET _N ^ _X A ₂ + _t SET _{Ni + 2} : A ₂ + / ₂ …-… SET _Wi A ₂ + f _Ni

(3) A low-correlation / zero-correlation spreading code set is obtained from the above steps, where each spreading code set includes f spreading codes.

Construction of time-frequency code combination: The specific steps are similar to the foregoing, and the package description is as follows: For a spreading code set containing K spreading codes, it is divided into y from the code domain, time domain, and frequency domain, respectively ,, w ₂ , w ₃ low-correlation / zero-correlation spreading code sets, which can be obtained by combining them; ν =

Low correlation / zero correlation spreading code set.

In addition to the time domain, code domain, and frequency domain mentioned above, any other domain that can divide a spreading code set into multiple low-correlation / zero-correlation spreading code sets can use the method described above to form the present invention. Low correlation / zero correlation spreading code set. The invention also includes an equivalent transformation of any of the aforementioned low-correlation / zero-correlation spreading code sets.

The invention also includes a low-correlation / zero-correlation spreading code set application method, a novel channel allocation, power allocation, and cell / sector networking method. This new channel allocation, power allocation, and cell / sector networking method can effectively suppress interference from neighboring cells and interference from neighboring sectors. This new method of channel allocation, power allocation and cell networking includes:

(1) Obtaining low correlation / zero correlation spreading code sets from the above coding method;

(2) Divide each cell into concentric rings, and the size of the concentric rings can be determined according to the actual situation. (Business distribution or design considerations, etc.);

(3) The above-mentioned division of each cell into concentric rings is only a special case. In addition, other division methods may be used. For example, the partitions may be determined according to the path loss, the size of the interference, or the signal-to-interference ratio, and the partitions may even be irregular. There may also be other divisions;

(4) Assign the obtained low-correlation / zero-correlation spreading code sets to the loops respectively, and from the inner loop to the outer loop, the arrangement of the spreading code sets is denoted as Λ;

(5) Perform power allocation for the users of each ring. The power allocation is different between the downlink and the uplink. In the downlink, power allocation is required for users in each ring in accordance with the power balance rule. The user in the outer ring has a large path loss and allocates a large transmit power, while the user in the inner ring has a small path loss and allocates a small transmit power. . In the uplink, the traditional power allocation is performed in accordance with the principle of power balance, that is, the power of the users arriving at the base station must be equal at the same time. In the present invention, the power required for the users in the inner ring to reach the base station is higher than that in the outer ring Of users reach the base station with high power. If more accurate power allocation is performed, the uplink and downlink can uniformly follow the signal-to-interference ratio (SIR) balancing criterion to perform;

(6) In the neighboring cell, the spreading code set from the inner ring to the outer ring should be arranged differently from the local cell, and is marked as P ₂ , Λ,... In a multi-cell networking environment, the arrangement of spreading code sets in different cells should be performed in accordance with the following guidelines: The outer rings of neighboring cells should be allocated different spreading code sets as much as possible, that is, users in the outer rings of neighboring cells should use different The code words in the code set, so that there can be no interference or little interference between the outer ring users of adjacent cells.

In the downlink, because the above-mentioned new spreading codewords are used, the codewords from different code sets have good correlation characteristics, so the interference between users in the outer ring of adjacent cells is very small or even no interference at all. The users in the outer ring are only affected by the interference from the users in the inner ring of the neighboring cell, and the users in the inner ring have a small transmit power, so for the users in the outer ring, the interference in the neighboring cell is greatly reduced. It is said that it is subject to interference from users in the outer ring of the neighboring cell, but for users in the inner ring, the neighboring base stations are much longer than the base stations in the cell, After the reduction, the interference has been reduced, so users in the inner loop can still work normally. In the uplink, if the traditional power allocation method is adopted, that is, the power balance criterion of the users arriving at the base station is followed, then the users in the outer ring allocate a larger transmit power, and the users in the inner ring allocate a smaller transmit power. When the path attenuation reaches the base station, the power of the outer ring users and the inner ring users are equal. In this case, we analyze the interference experienced by the inner ring users and outer ring users: For the outer ring users, the spreading codes used by the outer ring users in neighboring cells have good correlation characteristics with them, so their interference is very small or even No, it only interferes with the users in the inner ring of the neighboring cell. According to the traditional power allocation criteria, the users in the inner ring have small transmit power, so the users in the outer ring suffer relatively little interference from the neighboring cell. In other words, the interference it receives comes from the outer ring users of neighboring cells. Because the transmit power of the outer ring users is relatively large, the inner ring users experience relatively large interference from neighboring cells.

From the above analysis of the uplink, it can be seen that if the traditional power allocation criterion is used, the new channel allocation mentioned in the present invention will cause an interference imbalance (because it uses the power balance criterion, the interference Imbalance is also equivalent to signal-to-interference ratio imbalance). How to overcome this phenomenon, we introduce the signal-to-interference-ratio balance criterion in the uplink, and it is easy to solve this problem, that is, to appropriately increase the transmit power of the inner-ring users, so that the inner-to-outer ring user's SNR level reaches a balanced status. At this time, the transmit power of the inner ring users is still smaller than that of the outer ring users, but the difference is much smaller than the power allocation criterion. The advantage of this is that the dynamic range of the transmit power of the uplink users is greatly reduced. It greatly reduces the neighboring cell interference of the outer ring users and improves the signal-to-interference ratio level in the uplink.

In fact, the new channel allocation, power allocation, and cell networking methods provided by the present invention have found a method for balancing the communication quality of users in a cell, that is, whether it is an inner ring user or an outer ring user Can communicate normally. There will be no situation where only users close to the base station can work;

(7) The above-mentioned channel allocation and power allocation are not only applied to cell networking, but also applicable to sector networking and similar networking; (8) The above-mentioned channel allocation and power allocation are also applicable to the downlink and the uplink;

(9) The Γ spreading code words in each spreading code set can use orthogonal sequences or zero correlation window sequences. If ordinary orthogonal sequences such as WALSH codes are used, although the interference of neighboring cells is reduced, However, because ISI and MAI are still very large, the system capacity is still unlikely to be greatly improved. If the spreading codewords in each code set use a zero correlation window sequence, the ISI and MAI are eliminated because the sequence is eliminated. The channel allocation, power allocation and cell / sector networking method provided by the invention greatly reduce the ACI, and the system capacity will be greatly improved.

According to the application method of the present invention, the channel allocation and the power allocation are applicable to a cell network, a sector network, and other similar networks, and are applicable in the downlink; it includes the following specific steps:

a) First determine a division, that is, the base station allocates the same power to all users, and then detect the signal-to-interference ratio level in different areas of each cell or sector, and divide each cell or sector according to the level of the signal-to-interference ratio. Divided into different parts, and other criteria can be used to determine N divisions;

b) allocating the spreading code sets to the partitions, and at the same time ensuring that the partitions with relatively low interference to the neighboring cells are assigned different spreading code sets;

c) Allocate larger power to users in areas with low SNR, and allocate less power to users in areas with high SNR, until the SNR levels of users in all areas are consistent.

According to the application method of the present invention, the channel allocation and the power allocation method are applicable to cell networking, sector networking, and other similar networking, and are applicable in the uplink; it includes the following specific steps:

a) First determine the division, the method is as follows: All mobile stations are allocated the same power, and then the user's SNR level in different areas of each cell or sector is examined, and each cell or sector is compared according to the level of the SNR level. The area is divided into N different parts, and other criteria can also be used to determine the N divisions;

b) Allocating spreading code sets to the partitions, while ensuring that the interference of adjacent cells is relatively low The allocation of different spreading code sets;

The effects of the present invention are as follows: (1) A new type of spreading codeword-low correlation / zero correlation spreading code set is provided. This codeword can be divided into different spreading code sets. The frequency code set contains different spreading codewords, and the codewords of different codesets have a low correlation / zero correlation characteristic; (2), a grouping method and a coding method for this new type of spreading codeset are provided, including time Domain construction method, frequency domain construction method, code domain construction method, time-frequency joint construction method, time code joint construction method

(3) A new channel allocation, power allocation, and cell / sector networking method based on a low correlation / zero correlation spreading code set is provided; (4) A new type of channel allocation, power allocation, and cell / fan In the area networking method, the group spreading code set is fully used in each cell / sector. It can be seen that the system using this new type of cell networking has a high capacity; (5) The present invention provides A special channel allocation, power allocation, and cell / sector networking method, which greatly reduce the interference of adjacent cells and interference between adjacent sectors of the cellular mobile system, and also make the entire system reach a balanced state, that is, the internal Both users and users at the cell boundary communicate with the same quality (or the communication quality of each user is specified according to different Qos) without the imbalance of the traditional CDMA system. Users within the cell can communicate. However, users at the cell boundary cannot communicate. This balanced communication system can provide the maximum system capacity. (6) The above-mentioned channel allocation and Power allocation is not only applied to cell networking, but also applicable to sector networking and similar networking;

(7) The above-mentioned channel allocation and power allocation are not only applied to the downlink, but also applicable to the uplink; (8) Each spreading code word in each spreading code set can use an orthogonal sequence or zero. For the correlation window sequence, if ordinary orthogonal sequences such as WALSH codes are used, although the neighboring cell interference is reduced, since ISI and MAI are still very high, the system capacity cannot be greatly improved. The frequency codeword uses a zero correlation window sequence. Because the sequence elimination has eliminated ISI and MAI, the channel allocation, power allocation and cell / fan provided by the present invention The area networking method greatly reduces the ACI, and the system capacity will be greatly improved. (9) In the traditional dynamic channel allocation algorithm, if the full user's working state, the dynamic channel allocation algorithm has failed, and the channel allocation and power allocation algorithm provided by the present invention is still effective in the case of full users, so It can provide larger capacity. In addition, the traditional dynamic channel allocation has a very limited interference suppression effect. The channel allocation and power allocation algorithm provided by the present invention can greatly reduce the interference of neighboring cells in the system. In addition, the traditional dynamic channel allocation algorithm It is quite complicated and needs dynamic time slot switching, code channel switching and channel switching. However, the channel allocation and power allocation algorithm provided by the present invention is very simple. It only switches based on the geographical location of the user. It is actually equivalent to static .

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic diagram of the division of N time slots;

FIG. 2 is a schematic diagram of time-domain division of N low-correlation / zero-correlation spreading code sets;

Figure 3 is a schematic diagram of the division of three uncorrelated LS code sets;

4 is a schematic diagram of an arrangement of three spreading code sets in three cells;

FIG. 5 is a comparison of interference between neighboring cells of three schemes in a specific embodiment; FIG.

FIG. 6 is a schematic diagram of an arrangement of four spreading code sets in three cells;

Figure 7 is a schematic diagram of the arrangement of three spreading code sets in 19 cells;

Figure 8 is a schematic diagram of two sectors.

detailed description

(1) Low correlation / zero correlation spreading code set, that is, the present invention provides a new type of low correlation / zero correlation spreading code set and a grouping method thereof:

(1) Suppose there are spreading codewords, where N, can take any positive integer, and grouping will be performed on this basis;

(2) Group the spreading codes, divide them into code sets of the same size, and mark them as SET and SET ₂ SBT _P. Each code set has a code word, and mark them as

C0DE, CODE, C0DE _K. This is the first codeword in the first codeset. Shown as, SET _n , CODE,);

(3) The grouping is performed according to the following criteria: The grouping tries to make the different code sets have good correlation characteristics (low correlation or zero correlation within a large time offset around the origin), that is, any two come from different The codes in the code set have good cross-correlation characteristics, but there is no special requirement for the correlation characteristics between the codes in the same code set. It can be an orthogonal sequence, a zero correlation window sequence, or even non-orthogonal. Sequence

(4) According to the above steps, we have obtained a code set, each code set contains f code words, among which the code words from different code sets have good correlation characteristics, and the internal code of the code set There are no special requirements for related characteristics between words.

An example of the above-mentioned low-correlation / zero-correlation spreading code set and grouping method is described below. The following eight complementary spreading codewords with a code length of 8 (with two complementary parts) are:

c ₁ = (++++++ ~, +-+-+-+) c ₂ = (+ "+-++++, +-++-+-)

c ₃ = (++++ ~ ++, +-+-++-) c ₄ = (++, +-+-+-+)

c ₅ = (+-+-+ ~ + ₅ ++++++-) c ₆ = (+ ._ ++ _ + _-5 ++ __ ++++)

c ₇ = (+-+ — ++-

c ₈ = (+-+-+-+, ++)

We divide the above codewords into two groups: ^ Ji ^ ^^), SET ₂ = {c ₅ , c ₆ , c ₇ , c _s ) _{0 The} codewords in the two spreading code sets after grouping have the following properties: The codewords of the same code set have the characteristics of zero correlation window. The zero correlation window is (-1,1), and the cross-correlation between codewords in different code sets is zero at all time offsets, that is, two code sets. Has ideal cross-correlation characteristics.

Table 1 lists the correlation functions of the first code and other codes

Table 1: Autocorrelation and cross-correlation functions of code 1.

IS

As can be seen from Table 1, the codewords in code 1 and code set 1 (code 1, code 2, code 3, code 4) have zero correlation window characteristics, and the zero correlation window is (-1, 1), code 1 It has ideal cross-correlation characteristics with code set 2 (code 5, code 6, code 7, code 8). Table 1 only lists the auto-correlation and cross-correlation functions of code 1. Other codes have similar properties with code 1. The above grouping yields N = l code sets, each code set has K = 4 spreading codes, and the two code sets have ideal cross-correlation characteristics.

Given any of the above spreading code words, it is not easy to divide the spreading code into N groups according to the above grouping requirements. However, we can construct the above-mentioned low-correlation / zero-correlation spreading code set through a specific construction method. The present invention also provides a method for constructing a low-correlation / zero-correlation spreading code set. The construction method includes a time-domain construction method, a frequency-domain construction method, a code-domain construction method, a time-frequency joint construction method, and a time-code joint construction method. A method of joint construction of frequency and code and a method of joint construction of frequency and code.

The code domain construction method uses a special construction method to distinguish different code sets from the code domain, so that the code sets have good correlation characteristics. The low-correlation / zero-correlation spreading code sets distinguished in the code domain are as follows. It is called code division low correlation / zero correlation spreading code set.

The code domain construction method includes the following steps:

(1), construct a pair of unrelated generalized complementary codes with complementary parts, and the length of each complementary part is L. (For specific construction methods, please refer to the literature [1, 2, 5])

Codewords of length _{_{S 1; S 2, ...,}} S ^, except zero, if the self-correlation of codewords and all all-zero offset, called code words (S _l5 S _2, ... , S is a generalized complementary code with complementary parts.

Suppose two general complementary codes with complementary parts: (S,…, S), (Sj ² , S ₂ ² ,…, s;), if the sum of the cross-correlation of the complementary parts between the two is in all Time offset If it is zero, the two general complementary codes are said to be uncorrelated.

For a generalized complementary code with complementary parts, the following conclusions hold: There are at most N pairs of generalized complementary codes with complementary parts that are unrelated. (See references [1,2, 5]) We write this generalized complementary code as: (S, S ₂ ·. ·, S), (S, S ₂ ² , S), ..., (S, S, · .., S).

When = 2, the above codewords are often called complementary codes. At this time, there are at most two mutually unrelated complementary codewords. When> 2, the above codewords are called generalized complementary codes. At this time, there are at most two generalized complementary codes that are irrelevant. .

When the length Z-1 of each complementary part is a special case, then a general complementary code structure si s ^1. '

The resulting matrix is actually an orthogonal matrix, that is, matrix A = is an orthogonal matrix _c

^¾ 1 ^¾ 2 · " ^b N

When N, L takes a relatively small value, a generalized complementary code can be obtained by computer simulation search. When N, takes a relatively large value, it can be obtained through a certain construction method (see references [1,2, 5] ).

(2) Select an orthogonal matrix H κχΚ of ΓχΓ

(3) Through the above two steps, a pair of general complementary codes with complementary parts that are unrelated in pairs have been obtained: (S, S ₂ S, (s, s ₂ ² , ···, S), ... , (S, s, · ..,

S), an orthogonal matrix with code length and an x. For each generalized complementary code, the orthogonal matrix 11 ^^ is used to perform Kronecker extension to obtain an extended spreading code set. The spreading code set contains a spreading code. Word, the specific extension is as follows:

_{_{_{^{SET N = μ ΚχΚ ®s x N}}}} , u KxK ®S 2 w ... H KXK Where '®, stands for Kronecker operation, such as:

For each extended spreading code set SET _n , 1 n N, whose ^ "" line represents an extended codeword, each spreading code set contains a spreading codeword, and each spreading codeword has a complement Part, the length of each complementary part becomes the obtained spreading code set has the following properties:

(a) For each spreading codeword in the spreading code set, there is a zero correlation window of [-Z + 1, Z-1], that is, Γ spreading codewords in [--1] have ideal Autocorrelation and cross-correlation. If = 1, then the f spreading codewords in each code set after expansion remain orthogonal only.

(b) For different spreading code sets, at all time offsets, the codes in the two code sets have a zero correlation characteristic. That is, in [-1, NL-1], the cross-correlation between any two codes from different code sets is zero.

(4) Spreading code sets have been obtained from the above three steps, and each code set has f spreading codes. Among them, the codewords from different code sets have zero correlation characteristics, and the codewords inside the codeset can be zero-correlation window codewords or ordinary orthogonal codewords. The generated spreading codes are generalized complementary codes with complementary parts. When performing correlation calculation, N complementary parts are required to be calculated separately, and they cannot overlap. Therefore, in practice, the complementary parts should be transmitted in the synchronously fading channels. We can send the complementary parts in timeslots with guard intervals inserted between adjacent timeslots. This may not be very efficient. We can even change the # ^ 户间 in the middle, so that although ^^ 马 ^) Ί loses the objective, it still has the characteristic of the objective.

Let's take an example to explain the horse method. Please choose 4, L = 2, K = 4.

(1) First, construct four generalized complementary codes with four complementary parts that are unrelated in pairs and have a code length of 2. The codeword given here is as follows (for the specific structure, see references 1, 2, 5):

^{_{(Si l, S 2 ',}} S 3,, S4 l) = (++, ++, + -, + -) (Si 2, S 2 2, S 3 2, S 4 2) = (++, -, +-,-+)

(S, ³ , S ₂ ³ , S3 ³ , S ₄ ³ ) = (+-, +-, ++, ++) (S! ⁴ , S ₂ ⁴ , S3 ⁴ , S ₄ ⁴ ) = (+- ,-+, ++,-)

Table 2 lists the auto-correlation and cross-correlation functions of the codes (SA S, S3 ¹ , S) Autocorrelation and cross-correlation functions of codes (S, S ₂ S 3 ¹ , s)

Table 2 only lists the auto-correlation and cross-correlation of the first code. The correlation of other codes is similar to that of code 1, that is, it has ideal auto-correlation and cross-correlation functions. (2) Select a 4x4 orthogonal matrix H _4X4 = lll

+

+ + + +

(3) One or four spreading code sets can be obtained through the above Kronecker extension:

SET:

code 1: ++++++ ―, ++++++-, +-+-+-+, +-+ _ + --- +

code 2: ++ — ++++, ++ — ++++, + — ++ _ +-, + — ++-+ _

code 3: ++++-++, ++++-++, +-+-++-, +-+-++ _

code 4: ++, ++, + — +-+-+, + — +-+-+

SET ₂

code 5: ++++++ ―, ++, +-+-+-+,-+-+-++-

code 6: ++ — ++++, one ++ —―, + — ++-+-,-++ — +-+

code 7: ++++-++, ---- ++-, +-+-++-,. +-++-+

code 8: ++, — ++++++, + — +-+-+,-++-+-+-

SET ₃ :

code 9: +-+-+ — +, +-+-+ — +, ++++++ —, ++++++ —

code 10: +-++-+-, +-++-+-, ++ ++++,

code 11: +-+ — ++-, +-+ — ++-, ++++ — ++, ++++ — ++

code 12: + — +-+-+, + — +-+-+, ++, ++

SET ,:

code 13: +-+-+ — +,-+-+-++-, +++ · Η · + ―, ++

code 14: + — ++-+-,-++ +-+, ++ — ++++, — ++ —―

code 15: +-+ — ++-,-+-++ — +, ——— ++-code 16: +-+-+-+,-++-+-+-, ++-+ +++++ The autocorrelation and cross-correlation functions of code 1 are listed in Table 3. Table 3: Autocorrelation and cross-correlation functions of code 1.

Note: In the above related calculations, the four complementary parts cannot overlap each other. In practice, each complementary part can be sent on non-overlapping channels with synchronous fading. For example, a guard interval can be inserted between the four simple parts and then sent. However, even without a guard interval, different code sets are still transmitted. Has low correlation characteristics.

Table 3 only lists the auto-correlation and cross-correlation functions of Code 1. Other code words have similar properties to Code 1. It can be seen from Table 3 that the correlation function between code 1 and each code of the first spreading code set has a zero correlation window characteristic, that is, among the codes in the first code set within (-1, 1) Have ideal correlation characteristics between them, while code 1 and the other three code sets have zero correlation at all time offsets.

The four spreading code sets generated above have the following properties: Code words within the same code set are at (-1, 1) It has zero correlation characteristics, and the codes of different code sets have zero correlation characteristics at all time offsets.

In the above example, the code sequence in the same code set is a zero correlation window sequence. In addition, the code sequence in the same code set may also be an orthogonal sequence. In the following example, the code sequence in the same code set is an orthogonal sequence, and different code sets have zero correlation characteristics.

Take the special case of Zn = 1, = 4.

(1). First, construct four generalized complementary codes with four complementary parts that are irrelevant and have a complementary code length of 1. The codeword given here is as follows (for the specific structure, see reference 5):

(S, ¹ , S ₂ ^] , Ss ¹ , S ₄ ') = (+, +, +,-) (S, ² , S ₂ ² , S ₃ ² , S ₄ ² ) = (+,-, + , +)

(S, ³ , S ₂ ³ , S3 ³ , S ₄ ³ ) = (+, +,-, +) (β S ₂ ⁴ , S3 ⁴ , S ₄ ⁴ ) = (+,-,-,-)

Obviously, the above four codewords are mutually orthogonal.

(2) Select a 4x4 orthogonal matrix H _4X4

(3) Four spreading code sets can be obtained through the above Kronecker extension:

SET:

code 1: +++-, +++-, +++-, — +

code 2: +-++, +-++, +-++,-+-code 3: ++-+, ++-+, ++-+,-+-code 4: +--, 4 -—-, + —-,-+++

SET ₂ :

code 5: +++-, — +, +++-, +++-code 6: +-++,-+-, +-++, +-++

code 7: ++-+,-+-, ++-+, ++-+

code 8: + —,-+++, + ---, + —

SET ₃ :

code 9: +++-, +++-, — +, +++-code 10: +-++, +-++,-+-, +-++

code 11: ++-+, ++-+,-+-, ++-+

code 12: + —, + ---,-+++, + —

SET ^. '

code 13: +++-, --- +, — +, — +

code 14: +-++,-+--,-+-,-+- code 15: ++-+,-+-,-+-,-+-code 16: + —,-+++,-+-H-,-+++ Correlation and cross-correlation functions Table 4: Autocorrelation and cross-correlation functions of code 1

It can be seen from Table 4 that Code 1 and other codes in the first spreading code set remain orthogonal only at the origin, and have ideal cross-correlation characteristics with codes in other spreading code sets.

The two examples above show how to use the code domain construction method to obtain a low correlation / zero correlation spreading code set. The invention also includes any equivalent transformation of the above-mentioned code set.

In the time domain construction method, different code sets are distinguished from the time domain. The generated spreading code set is called a time-division low correlation / zero correlation spreading code set. The time domain construction method includes the following steps: (1) There are f spreading codewords in history. This spreading codeword may be a zero correlation window code, or Therefore, it is a general orthogonal codeword, and even a non-orthogonal codeword.

(2) Divide a frame length into different time slots, and the width of each time slot is T, and the guard interval width between the time slots is Δ, as shown in FIG. 1.

(3) The spreading codes are divided into different spreading code sets according to time slots. The first time slot represents the first spreading code set. At this time, all the spreading codes are sent in the first time slot, and Nothing is sent in the other time slots. The spreading code sets constructed at this time are shown in Figure 2:

(5) From the above steps, a low correlation / zero correlation spreading code set has been obtained, and each spreading code set has spreading codes.

The time-domain construction method has some advantages over the code-domain construction method: (a), the codeword constructed by the time-domain construction method has higher efficiency than the code-domain construction method, because if the interval of each time slot is compared, the middle The protection time slot is smaller than the interval for sending information. For the code domain construction method, a guard interval needs to be inserted between the two complementary parts, which requires more resources to be sacrificed. (B) The time domain construction method is simpler and more flexible than the code domain construction method. Frequency code, we can assign it to different time slots to get a low-correlation / zero-correlation spreading code set, which is easy to obtain for any integer, but not so easy for the code domain construction method, such as When the number is odd, no general complementary code with complementary parts can be found in the binary field. In this way, N low correlation / zero correlation spreading code sets cannot be constructed by using the above method. '

(6) In the time domain construction method described above, the same spreading codes are sent in each time slot. In fact, in the time domain construction method, each time slot can send different f spreading codes. For example, if different timeslots can be used to transform f spreading codes, or even completely different spreading codes, the divided spreading code sets still maintain low correlation / zero correlation characteristics.

The invention also includes a frequency domain construction method for a low-correlation / zero-correlation spreading code set. The idea of the method is to divide the service spreading code into a low-correlation / zero-correlation spreading code set from the frequency domain. The resulting code set is called a frequency-respective low-correlation / zero-correlation spreading code set. The frequency domain construction method includes the following steps:

(1) Suppose there are Γ spreading codewords, which are respectively labeled as CODE ,, ..., C0DE. This spreading codeword can be a zero correlation window code, or a normal orthogonal codeword, or even non-positive. Cross codeword

(2) Select an unrelated carrier: i, f ₂ , ...,

(3) Modulate the if spreading codewords onto the aforementioned uncorrelated carriers respectively to obtain N uncorrelated spreading code sets: SET f. + ICODE ^ C0DE ₂ , ..., CODE ^, SBT ₂ : f ^ iCODE ^ C0DE ₂ , ..., CODE ^, ..., SET. F _N + {C0DE _x , C0D £ ₂ , ..., C0DE

(4) According to the above steps, there are obtained irrelevant spreading code sets, and each code set has spreading codes. The frequency division low correlation / zero correlation spreading code set has the following advantages: (a) Since the spreading code sets are modulated on different carriers, different spreading code sets have low correlation characteristics in all time offsets. ; (B) no need to insert a guard slot for a low-frequency / zero-correlation spreading code set, so its effect is higher than the time-division spreading code set and the code-division spreading code set. However, since different spreading code sets are modulated on different carriers, and there are a total of carriers, implementation problems will be encountered in practical applications.

In addition to the above three construction methods, the present invention also provides a joint construction method of the above three construction methods: including a time-frequency joint construction method, a time-code joint construction method, a frequency-code joint construction method, and a time-frequency code joint construction method. . Combined to generate N sets of low correlation / zero correlation spreading code sets. The code time construction method provided by the present invention includes the following steps:.

(1) Using the above code domain construction method, construct ^ spreading code sets with low correlation / zero correlation, respectively labeled as A ₁₅ A ₂ , ...,,, where each spreading code set contains Γ Spreading code.

(2) Considering the above-mentioned ^ spreading code sets as a whole, referring to the foregoing time domain construction method, Send them in ₂ time slots respectively, so you can get =>< ₂ low-correlation / zero-correlation spreading code sets. The spreading code set is:

SET [time slot 1 SET time slot 2 ... SET _NI time slot ₂

SET _{NI + L} : A ₂ + time slot 1 SET _N : A ₂ + time slot 2 ...... SET _1NI : A ₂ + time slot N ₂

SET _{NI _ _{L) N2 + L} : + time slot 1 SET _{MNI + 1} : Aw, + time slot 2 ... SET: + time slot N ₂

(3) A low correlation / zero correlation spreading code set is obtained from the above steps, where each spreading code set includes spreading codes. Combined together to generate a set of low correlation / zero correlation spreading code sets. The code time construction method provided by the present invention includes the following steps:

(1) A spreading code set is included, including spreading codes, denoted as A

(2) Modulate the above spreading code set onto ^ irrelevant carriers / i, / ₂ , ..., ^ to obtain

^ Irrelevant spreading code sets, respectively: Α ^ Α + ,,, Α ₂ = Α + / ₂ , ..., Α _{Ν <} = A + f _N]

(3), the above-mentioned spreading codes _Ν χ set as a whole, with reference to the time domain method of construction, which are transmitted in _two time slots, which can obtain = ^ _<2 low correlation / Zero correlation spreading code set. The spreading code set is:

SET ,: Aj + time slot 1 + time slot 2 ... SET _NI : k _x + time slot _N _τ

SET N, ₂ +] A ₂ + time slot 1 SETN ^: A ₂ + time slot 2 ... SET _2N2 : A ₂ + time slot N ..

SET _{N _ _{L) NI +]} : Aw, + time slot 1 SEH ₊₁ : Aw, + time slot 2 ... SET: + time slot N ₂

(4) A low correlation / zero correlation spreading code set is obtained from the above steps, where each spreading code set includes a spreading code. Then, together generate a set of low correlation / zero correlation spreading code sets. The code time construction method provided by the present invention includes the following steps: (1) Use the above code domain construction method to construct ^ spreading code sets with low correlation / zero correlation, which are respectively labeled as A ₁₃ A ₂ , ..., A ^, where each spreading code set contains a Spreading code.

(2) Taking the above-mentioned ^ spreading code sets as a whole, referring to the foregoing frequency domain construction method, modulating them into uncorrelated carriers and sending them, so that = ^ χ ₂ low correlations / Zero correlation spreading code set. The spreading code set is:

SET. A, + /, SET ,: Aj + / ₂ …… SET _Ni + f _Ni

^SET N ₂ + \ '■ A ₂ + /】 SET _{N2 + 2} : A ₂ + / ₂ SET _2f f ₂ : A ₂ + / w ₂

(3) A low correlation / zero correlation spreading code set is obtained from the above steps, where each spreading code set includes spreading codes.

The present invention also provides a method for constructing a joint time-frequency code. The specific steps are similar to the foregoing. The package description is as follows: For a spreading code set containing K spreading codes, the code-domain, time-domain, and will be divided into the frequency domain, ^, Λ ^ low-Related / zero correlation set of spreading codes, which can be obtained together _Ν = Ν Χ Ν ₂ Ν ^ f s two related / associated spreading code set to zero.

(2) A new channel allocation, power allocation and cell / sector networking method

The invention also includes a channel allocation, power allocation and cell / sector networking method based on the above-mentioned low-correlation / zero-correlation spreading code set. This new channel allocation, power allocation, and cell / sector networking method can effectively suppress interference from neighboring cells and interference from neighboring sectors. This new method of channel allocation, power allocation, and cell networking includes the following steps:

(1) A low correlation / zero correlation spreading code set is obtained by the above coding method; (2) Divide each cell into concentric rings, and the size of the concentric rings can be determined according to the actual situation (business distribution or convenient design considerations, etc.);

(3) The above-mentioned division of each cell into concentric rings is only a special case. In addition, other division methods may be used. For example, the partitions can be determined based on path loss, interference conditions, and signal-to-interference ratio levels. There may also be other divisions;

(4) Allocate the obtained low-correlation / zero-correlation spreading code sets to the loops respectively. From the inner loop to the outer loop, the arrangement of the spreading code sets is denoted as

(5) Perform power allocation for the users of each ring. The power allocation is different between the downlink and the uplink. In the downlink, power allocation is required for users in each ring in accordance with the power balance rule. The user in the outer ring has a large path loss and allocates a large amount of power, while the user in the inner ring has a small path loss and allocates a small amount of power. In the uplink, the traditional power allocation is performed in accordance with the principle of power balance, that is, the power of the users arriving at the base station must be equal at the same time. In the present invention, the power required for the users in the inner ring to reach the base station is higher than that in the outer ring. Of users reach the base station with high power. If you can make sure your skills ^ "S and X Bin can each control the shield shield (SIR) level ^ then OK;

(6) In the neighboring cell, the spreading code set from the inner ring to the outer ring should be arranged in a different manner from the local cell, and is marked as corpse ₂ , Λ, .... In a multi-cell networking environment, the arrangement of spreading code sets in different cells should be performed in accordance with the following guidelines: The outer ring of the neighboring cell should be allocated with different spreading code sets as much as possible, that is, the users of the neighboring d and outer ring should Use code words from different code sets.

In the downlink, because the above-mentioned new spreading codewords are used, the codewords from different code sets have good correlation characteristics, so the interference between users in the outer ring of adjacent cells is very small or no interference at all. The users in the outer ring are only affected by the interference from the users in the inner ring of the neighboring cell, and the users in the inner ring themselves have low transmission power. Therefore, for the users in the outer ring, the interference in the neighboring cell is greatly reduced. It is said that it is affected by the users from the outer ring of the neighboring cell, and its neighboring cell interference will increase, but for the users of the inner ring, the distance between the neighboring base station and the base station of the cell is large, and After attenuation, the interference is already very small, so users in the inner loop can still work normally. In the uplink, if the traditional power allocation method is adopted, that is, the power balance guidelines for users arriving at the base station are followed. At this time, users in the outer ring are allocated higher power, while users in the inner ring are allocated less power and reach the base station through path attenuation. The power of the outer ring users and the inner ring users is equal. In this case, we analyze the interference experienced by the inner ring users and outer ring users: For the outer ring users, the spreading codes used by the outer ring users in neighboring cells have good correlation characteristics, so their interference is small or even small. No, it only interferes with the users in the inner ring of neighboring cells. According to the traditional power allocation guidelines, the users in the inner ring send less power, so the users in the outer ring suffer relatively less interference from the neighboring cells. In other words, the interference it receives comes from the outer ring users of neighboring cells. Due to the transmit power of the outer ring users ^^, the neighbors received by the inner ring users are affected.

From the above analysis of the uplink, it can be seen that if the traditional power allocation criteria are adopted, for the new channel allocation mentioned in the present invention, an interference imbalance phenomenon will occur (because it uses the power balance criteria, the interference Imbalance is also equivalent to signal-to-interference ratio imbalance). If this phenomenon is overcome, we will introduce a signal-to-interference-ratio balance criterion in the uplink, which can easily solve this problem, that is, to appropriately increase the transmit power of the users in the inner ring, so that the signal-to-interference ratio level of the users in the inner and outer ring reaches a balanced level. status. At this time, the transmit power of the inner ring users is still smaller than that of the outer ring users, but the difference is much smaller than the power allocation criterion. The advantage of this is that the dynamic range of the transmit power of the uplink users is greatly reduced. It greatly reduces the neighboring cell interference of the outer ring users and improves the signal-to-interference ratio level in the uplink.

(7) The above-mentioned channel allocation and power allocation are not only applied to cell networking, but also applicable to sector networking and similar networking.

For any type of networking (including cell networking, sector networking, and mixed cell / networking networking, etc.), there is a unified method for determining partitions and power allocation strategies. The detailed description of the downlink The description is as follows: (a) First, a division is determined as follows: The base station allocates the same power to all users, and then examines the signal-to-interference ratio levels in different areas of each cell (or sector). Divide each cell (or sector) into different parts in high and low levels; (b) Allocate a spreading code set to each partition, and at the same time ensure that the lower-interference partitions of adjacent cells are assigned different spreading Code set; (c) allocating larger power to users in areas with relatively low SNR, and allocating less power to users in areas with high SNR, until the SNR levels of users in all areas are consistent.

The specific description in the uplink is as follows: (a), first determine a division, the method is as follows: all mobile stations are allocated the same power, and then the user signal-to-interference ratio level in different areas of each cell (or sector) is examined, Divide each cell (or sector) into N different parts according to the level of the signal-to-interference ratio; (b) Allocate the spreading code sets to the partitions, while ensuring that the signal-to-interference of neighboring cells is relatively low Divide and allocate different spreading code sets; (c), allocate larger power to users in areas with relatively low interference, and allocate less power to users in areas with relatively high interference, until users in all areas are disturbed. The ratio is consistent.

(8) The above-mentioned channel allocation and power allocation are also applicable to the downlink and the uplink;

(9) Each spreading code word in each spreading code set can use an orthogonal sequence or a zero correlation window sequence. If an ordinary orthogonal sequence such as a WALSH code is used, although the interference in the neighboring cell is reduced, ISI and MAI are still very large, so it is still impossible to greatly increase the system capacity. If f spreading codewords in each code set use a zero correlation window sequence, ISI and MAI have been eliminated because the sequence is eliminated, and the present invention The provided channel allocation, power allocation, and cell / sector networking methods greatly reduce the ACI, and the system capacity will be greatly improved.

In the following, we will analyze the significance of the aforementioned channel allocation, power allocation, and cell networking to reduce neighboring cell interference based on the actual model:

First let's talk about the downlink situation.

We take the LAS-CDMA system as an example to discuss the great benefits brought by the new channel allocation, power allocation, and cell network interference reduction to neighboring cells. LAS-CDMA traditional networking is U + LS mode, That is, users in the cell are distinguished by the LS code, and different cells are distinguished by the LA code. The new networking method is based on LA + LS plus the aforementioned ring splitting technology. By comparing these two networking methods, it can be seen that the new networking method can greatly reduce the interference of neighboring cells.

We take the three-cell model as an example, and take = 3. Each cell is divided into three rings of equal distance: inner ring, middle ring, and outer ring. We use the time-domain construction method to obtain three mutually independent spreading code sets. Each spreading code set contains 32 LS codes, and each LS code has a length of 128. The 32 LS codes are allocated to different Sent on a time slot to obtain three mutually unrelated spreading code sets SET SET, SET. In simulation

The division of the three spreading code sets is shown in Figure 3:

The allocation of the three spreading code sets in the three cells is shown in Figure 4:

Due to the good correlation between different spreading code sets, for users in a code set, the interference it receives mainly comes from users in the same code set in other cells. For the outer ring users in the first cell, the interference it receives comes from the inner ring users in the second cell and the middle ring users in the third cell; The interference comes from the outer ring users in the second cell and the inner ring users in the third cell. For the inner ring users in the first cell, the interference is from the middle ring users in the second cell. And outer ring users in the third cell. Obviously, the interference to the users of the outer ring is very small. This is also true for the other two communities.

Figure 5 shows a comparison of the three different ACI schemes. Option 1: LA + LS without power allocation; Option 2: LA + LS, power allocation; Option 3: LA + LS, new networking solution (three rings) + Power allocation). It can be seen that the third scheme is far superior to the other two schemes, and the specific comparison is shown in Table 6.

Table 5 lists some of the simulation conditions. Table 5: Some simulation parameters

Table 6: Comparison of adjacent cell interference between the three schemes

It can be seen from Table 6 that in the outer ring, the ACI level of the third option is about 17dB lower than that of the second option, and about 22dB lower than that of the first option. In the middle ring, the ACI level of the third option is about 3dB lower than that of the second option. It is about ⁸ dB lower than the first scheme. In the inner loop, the ACI level of the third scheme is about MB lower than that of the second scheme and about 7dB lower than that of the first scheme.

It can be seen from the analysis of Figure 5 and Table 6 that the new channel allocation, power allocation, and cell networking provided by the present invention greatly reduce the ACI level of the system, especially for users at the cell boundary, which is exactly what other current solutions provide. lack of.

It can also be seen from Figure 5 that the new channel allocation, power allocation, and cell networking schemes lead to inconsistent levels of interference for users at different locations in the cell, and ultimately lead to inconsistent SNR levels, that is, the SNR levels for users in the outer ring highest. In practice, if the SNR criterion is adopted in the power allocation criterion, this problem can be solved.

The above example only enumerates the special case where three cells are divided into three rings. In practice, the present invention is applicable to the case where any cell has any number of divisions. Figure 6 lists three cells and each cell has 4 partitions; Figure 7 lists cell 19 and each cell has 3 partitions. In the figure, A represents S T B represents SET C represents SET ^

The above are just a few typical examples. In practical applications, the novel channel allocation, power allocation, and cell networking provided by the present invention can be flexibly applied, which are mainly reflected in: (1.), Low correlation / zero correlation spreading code sets can be obtained in a variety of ways, including time domain, code domain, frequency domain, time frequency, code frequency, code time and time code frequency, etc .; (2), each cell can be based on different The criteria are divided into different areas, such as distance, path loss, and interference situation. The ultimate goal is to maximize the signal-to-interference ratio, that is, minimize the interference of neighboring cells; (3), the neighboring cells from the inner loop The spreading code sets to the outer loop have different arrangements. This arrangement is flexible. The principle to be followed is to prevent users at the cell boundary from interfering with each other. (4) The power allocation method is flexible and can use power. (5) The new channel allocation, power allocation, and cell networking methods are applicable to both the downlink and the uplink. ; (6), the new channel allocation and power allocation small area networking method applies to both cell networks, and is also applicable to other similar networking sector networking.

The novel channel allocation, power allocation and cell networking method provided by the present invention are not only applicable to cells Networking is also applicable to sector networking to solve the interference problem of neighboring sectors. First, the sector is divided into several non-overlapping parts. The division method is mainly determined according to the interference. The sector boundary receives the most interference from the neighboring sector, and the middle part of the sector receives the smallest interference from the neighboring sector. Therefore, the following divisions can be made: A sector is divided into N different parts from the sector boundary to the sector center, and then the above-mentioned channel allocation, power allocation, and cell networking methods can be applied to this multi-sector networking.

The following is a simple example to illustrate the situation of multi-sector networking, as shown in the case of two sectors shown in Figure 8. Each sector is divided into two parts according to the situation of sector interference, sector boundary part And the central part of the sector, and then select two low correlation spreading code sets SET, SET. In sector 1, the sector boundary part uses SET. The sector center part uses SET. In sector 1, the opposite is true. The power allocation strategy is selected in such a way that users at the sector boundaries allocate larger power, while users at the sector center allocate less power, so that users outside the two sector boundaries can not interfere with each other. In this way, the interference of users at the sector boundary is greatly reduced, and the users at the center of the sector will not receive much interference due to the antenna directional angle. This solves the problem of networking of two sectors. This new method is far superior to the traditional networking method.

The above example only gives a special case of two sectors. If more sectors are needed for networking, consider dividing each sector from the border to the center into more parts, and then perform networking according to the following guidelines. : Users on borders of adjacent sectors should try to use different spreading code sets.

The channel allocation, power allocation, and cell networking method provided by the present invention can be extended to the case of a mixed cell / sector networking, which is specifically described as follows:

(1) First, it is assumed that all base stations do not use power allocation, and all users use the same transmission power, and then the cell or sector is divided into different parts according to the size of the signal-to-interference ratio;

(2) Construct a low correlation / zero correlation spreading code set SET and ST ₂ SET _N according to the foregoing method.

(3) In each cell or sector, the spreading code sets are allocated to the above-mentioned parts according to different arrangements, and the allocation should be performed according to the following criteria: In adjacent cells or sectors, each other The most disturbing part uses different spreading code sets as much as possible.

(4) The power allocation is performed for different parts, and the power allocation criteria are as follows: Without the power allocation condition, the part with the smaller signal interference allocates a larger transmission power, and the part with a larger signal interference allocates a smaller transmission power. Finally, the signal-to-interference ratio of all parts is balanced.

The channel allocation, power allocation, and cell / sector networking method provided by the present invention are not only applicable to the downlink, but also applicable to the uplink. The specific implementation method of the method in the uplink is as follows:

(1) First, it is assumed that all mobile stations do not use power allocation, and all users use the same transmission power, and then the cell or sector is divided into different parts according to the size of the signal-to-interference ratio;

(2) Construct a low correlation / zero correlation spreading code set SET, SET ₂ and ET according to the foregoing method.

(3) In each cell or sector, the spreading code sets are allocated to the above N parts according to different arrangement methods, and the allocation should be performed according to the following criteria: Partially try to use different spreading code sets.

(4) Perform power allocation for different parts, and the power allocation criteria are as follows: in the case of no power allocation, the part with the smaller signal interference allocates a larger transmission power, and the part with a larger signal interference allocates a smaller transmission power. Finally, the signal-to-interference ratio of all parts is balanced.

The main contribution of the present invention is to provide a new type of spreading codeword, a low correlation / zero correlation spreading code set, and a new channel allocation, power allocation, and cell / sector networking method based on the codeword. Reduced interference from neighboring cells.

(1) A new type of spreading codeword is provided with a low correlation / zero correlation spreading code set. This codeword can be divided into different spreading code sets, and each spreading code set contains f Different spreading codewords, with low correlation / zero correlation between codewords of different code sets;

(2) The above-mentioned low-correlation / zero-correlation spreading code sets may also have different sizes, but generally the same spread-spectrum code sets are used;

(3) A grouping method and a coding method for such a new type of spreading code set are provided, including a time domain structure Method, frequency domain construction method, code domain construction method, time-frequency joint construction method, time code joint construction

(4) The low-correlation / zero-correlation spreading code set provided by the present invention further includes an equivalent transformation of any code set generated above;

(5) A new channel allocation, power allocation and cell / sector networking method based on a low correlation / zero correlation spreading code set is provided;

(6) In the new method of channel allocation, power allocation, and cell networking, the group spreading code set is fully used in each cell. It can be seen that the system using this new type of cell networking has a high capacity. ;

(7) The present invention provides a special channel allocation and power allocation and cell / sector networking method, which greatly reduces adjacent cell interference and interference between adjacent sectors of a cellular mobile system, and also makes the entire system reach a Balanced state, that is, users within the cell and users at the cell boundary communicate with the same quality (or the communication quality of each user is specified according to different QoS), without the imbalance of the traditional CDMA system. Users within the cell can communicate, but users at the cell boundary cannot communicate. This balanced communication system can provide the maximum system capacity;

(8) The above-mentioned channel allocation and power allocation are not only applied to cell networking, but also applicable to sector networking and similar networking;

(9) The above-mentioned channel allocation and power allocation are not only applied to the downlink, but also applicable to the uplink;

(10) Each spreading code word in each spreading code set can use an orthogonal sequence or a zero correlation window sequence. If a common orthogonal sequence such as a WALSH code is used, although the interference in the neighboring cell is reduced, ISI and MAI are still large, so the system capacity is still unlikely to be greatly improved. If the spreading codewords in each code set use a zero correlation window sequence, ISI and MAI have been eliminated because the sequence is eliminated, and the present invention provides The channel allocation, power allocation, and cell / sector networking method are greatly reduced by ACI, and the system capacity will be greatly improved. (11) In the traditional dynamic channel allocation algorithm, if the full user working state is used, the dynamic channel allocation algorithm has failed, and the channel allocation and power allocation algorithm provided by the present invention is still effective in the case of full users, so it can be Provides larger capacity, and in addition, the traditional dynamic channel allocation has a very limited interference suppression effect, and the channel allocation and power allocation algorithm provided by the present invention can greatly reduce the interference of neighboring cells of the system. In addition, the traditional dynamic channel allocation algorithm is quite equivalent It is complicated and needs dynamic time slot switching, code channel switching, and channel switching. The channel allocation and power allocation algorithm provided by the present invention is very simple, and it is only switched according to the geographical location of the user. It is actually equivalent to static . references:

[1] S. J. Xu, Y. Gao and D. B. Li, "A coding method to create general spread spectrum sequence with zero correlation window." PCT / CN02 / 00193, March 22, 2002.

[2] S. J. Xu, Y. Gao and D. B. Li, "A coding method to create mismatched spread spectrum sequence with zero correlation window." PCT / CN02 / 00194, March 22, 2002.

[3] S. J. Xu, R. Tamrakar, Y. Gao and D. B. Li, "A coding method to create spread spectrum sequence with group-wise zero correlation window", PCT / CN02 / 00238, April 5, 2002.

[4] D.B. Li, "A spread spectrum multiple access coding method with zero correlation window", application no. PCT / CN00 / 00028. 2000.

[5] P.Z. Fan and M. Darnell, Sequence Design for Communications Applications, John Wiley, RSP, 1996.

'

Claims

Rights request

1. A low-correlation / zero-correlation spreading code set encoding method, characterized in that: a total of NK spreading codes are set, and the NK spreading codes are grouped and divided into code sets, each code set Contains a spreading code;

The grouping described therein meets the following conditions: The code sets after the grouping have low correlation / zero correlation characteristics.

2. The method according to claim 1, wherein the low-correlation / zero-correlation between the grouped code sets refers to: between any two spreading codes from different code sets. The cross-correlation function maintains low correlation / zero correlation characteristics within a relatively large time offset around the origin; and the codes in the same code set can be: an orthogonal code set, or a zero correlation window code set, or a Non-orthogonal code set.

3. The method according to claim 1, wherein the grouping the spreading codes comprises: distinguishing different code sets from a code domain and dividing into code division low-correlation / zero-correlation spreading. 4. Code sets, so that the code sets have low correlation / zero correlation correlation characteristics.

The method according to claim 1, wherein the grouping the spreading codes comprises: distinguishing different code sets from the time domain, and dividing into time-division low-correlation / zero-correlation spreading code sets. , So that the N code sets have low correlation / zero correlation correlation characteristics.

The method according to claim 1, wherein the grouping the spreading codes comprises: dividing the spreading codes into a low-correlation / zero-correlation spreading code set from the frequency domain, that is: A frequency division low-correlation / zero-correlation spreading code set, so that the code sets have low-correlation / zero-correlation correlation characteristics.

The method according to claim 1, wherein the grouping the y½r spreading codes comprises: jointly generating a set of low-correlation / zero-correlation spreading code sets from a code domain and a time domain, that is: dividing into The time-divided low-correlation / zero-correlation spreading code sets of each code set have low-correlation / zero-correlation correlation characteristics between the code sets.

The method according to claim 1, wherein the pair of spreading codes is The line grouping includes: generating ^ groups ^^ correlation / zero correlation spreading code sets from the time domain and frequency domain together, that is, dividing into time frequency division low correlation / zero correlation spreading code sets, so that the code sets have Low correlation / zero correlation.

The method according to claim 1, wherein the grouping the spreading codes comprises: generating a set of low-correlation / zero-correlation spreading code sets from a code domain and a frequency domain together, that is, dividing into The code frequency division has low correlation / zero correlation spreading code sets, so that the code sets have low correlation / zero correlation correlation characteristics.

The method according to claim 1, wherein the grouping the spreading codes comprises: generating a group of low-correlation / zero-correlation spreading code sets from a time domain, a frequency domain, and a code domain together, that is, : Divided into time-frequency code division low-correlation / zero-correlation spreading code sets, so that the code sets have low-correlation / zero-correlation correlation characteristics.

10. The method according to claim 1, comprising the following steps:

Step 1. Assume that there are spreading codewords, where 'J can take any positive integer, and grouping will be performed on this basis;

Step 2: Group the spreading codes, divide them into two code sets of the same size, and mark them as SET, SET, SET _N. Each code set has K code words, and mark them as CODE, respectively. C0DE _l C0DE _K , so the / th code word in the first code set can be expressed as (SET _n , C0DE _k );

Step 4. A code set is obtained from the above steps, and each code set contains code words.

The method according to claim 10, wherein the grouping the spreading codes comprises: distinguishing different code sets from a code domain, and dividing into code division low correlation / zero correlation spreading codes. Set, so that the code sets have low correlation / zero correlation correlation characteristics; its specific steps The steps include:

Construct a general complementary code (Sj ¹ , S ^,…,

S, (S, S ₂ ² ,…, S), ......, (S, S, .. ·, S), and the length of each complementary part is L;

h ^h \ u \ ½½ ... hh \\ KK

Choose an orthogonal matrix x _λ ^ =…;

_ ^hh KK \ l ^llll KK22 ······ ^hh KKKK_

For each generalized complementary code, an orthogonal matrix is used to perform Kronecker expansion to obtain an extended spreading code set, and the spreading code set contains spreading code words. A total of N spreading code sets can be generated.

The method according to claim 11, wherein the performing Kronecker extension comprises the following steps:

_{_{SET X = μ ΚχΚ ®s \,}} n KxK ®s ... ΊΪ ΚχΚ ®s N l)

SET ₂ = {ί1χ _χΚ ®5] ² , Ή. _ΚχΚ ®5 ₂ ··· Ή-ΚχΚ ® ^S N)

h _u S h _l2 S… h _1K S where ^s ® 'stands for Kronecker operation, such as: ½S k ₂₂ S ■■■ h _2K S

H ^ ®S:

½1 ^{S h} K2 ^S ■■■ ^h KK ^S. For each extended spreading code set SET _n , 1 n N, whose ^ "line represents an extended codeword, then each spreading code set contains f spreading codes Codeword, each spreading codeword has a complementary part, and the length of each complementary part becomes 2, the obtained spreading code sets satisfy the following conditions: for each spreading codeword in each spreading code set has [ -Z + 1, Z-1] zero correlation window, that is, the spreading codewords within 1] have reasonable auto-correlation and cross-correlation; if Z = 1, then f in each code set after expansion Spreading codewords remain orthogonal only;

For different spreading code sets, at all time offsets, the codes in the two code sets have zero correlation characteristics, that is, the mutual relationship between any two codes from different code sets in [- Off are zero. -

13. The method according to claim 11, wherein the generated spreading codes are generalized complementary codes with complementary parts, and the complementary parts are required to be calculated separately to avoid overlap when performing correlation calculations.

The method according to claim 13, wherein the complementary parts are transmitted in N synchronously fading channels; for example, # complementary parts can be sent in time slots, and when adjacent A guard interval is inserted between the gaps.

The method according to claim 14, wherein the protection interval in the middle can be canceled, so that although the zero correlation characteristic is lost between the code sets, the low correlation characteristic is still maintained.

The method according to claim 10, wherein the grouping the spreading codes comprises: distinguishing different code sets from the time domain, and dividing into time-division low-correlation / zero-correlation spreading codes. Sets, so that the code sets have low correlation / zero correlation correlation characteristics; the specific steps include:

Assuming there are f spreading codewords, this spreading codeword can be a zero correlation window code, or a normal orthogonal codeword, or even a non-orthogonal codeword;

Divide a frame length into different time slots, the width of each time slot is T, and the guard interval width between the time slots is Δ;

The spreading code is divided into different spreading code sets according to time slots. The first time slot represents the? Spreading code sets, all spreading codes are sent in the first time slot, and nothing is sent in other time slots;

In the divided spreading code sets, the time offset is within [-Δ, Δ], and the cross-correlation between two different spreading code sets is zero, even if the time offset exceeds Δ. The overlapping part of the gap is not large, and the two different spreading code sets still maintain the correlation characteristics;

There are obtained low correlation / zero correlation spreading code sets, and each spreading code set has spreading codes.

The method according to claim 16, characterized in that: the same number of spreading codes are transmitted in each time slot.

The method according to claim 16, characterized in that each time slot of the plurality of time slots can send different f spreading codes.

The method according to claim 10, wherein the grouping the spreading codes comprises: dividing the spreading codes from a frequency domain into a low-correlation / zero-correlation spreading code set, that is, dividing A frequency division low correlation / zero correlation spreading code set, so that the code sets have low correlation / zero correlation correlation characteristics; the specific steps include:

Suppose there are Γ spreading code words, which are respectively labeled as CODE ,, ..., C0DE. The K spreading code words can be zero correlation window codes, or ordinary orthogonal code words, or even non-orthogonal codes. Word

Select an uncorrelated carrier: i, f ₂ , ..., f _N ;

The spreading code words are respectively modulated onto the aforementioned uncorrelated carriers to obtain an uncorrelated spreading code set: SET f ^ {C0DE _x , CODE ......, C0DE, SET ₂ : f ₂ + {C0DE _x , C0DE ₂ , ……, CODE},…, SETfj. F _N + {C0DE _x , C0DE ₂ , ……, CODE; get an uncorrelated spreading code set, each code set has Γ spreading Frequency code.

The method according to claim 10, wherein the grouping the spreading codes comprises: jointly generating a set of low-correlation / zero-correlation spreading code sets from a code domain and a time domain, that is, dividing into groups Code time-division low-correlation / zero-correlation spreading code sets, so that the code sets have low-correlation / zero-correlation correlation characteristics. The specific steps include:

Construct ^ code-score low-correlation / zero-correlation spreading code sets, respectively labeled as ..., ...,

A _Nl , where each spreading code set contains a spreading code;

Considering the ^ code division spreading code sets as a whole, and sending them in ₂ time slots respectively, so that = ^ x ^ low correlation / zero correlation spreading code sets can be obtained, and this spreading The code set is:

SET + time slot 1, SETi. Time slot 2, · .., SET _Ni : + time slot,

SET _{N + l} : A ₂ + time slot 1, SET ^ ₊₂ : + time slot 2, ..., SET _2Nl : A ₂ + time slot N ₂ , SET _N _, _{) N2} Α _Νι + time slot 1, SET _{MN2 + 2} : + time slot 2, ..., SET ^: A ^, + time slot W ₂ Frequency code set, where each spreading code set includes a spreading code.

The method according to claim 10, wherein the grouping the spreading codes comprises: jointly generating from a time domain and a frequency domain. A set of low correlation / zero correlation spreading code sets, that is, divided into Time-frequency division low-correlation / zero-correlation spreading code sets, so that the code sets have low-correlation / zero-correlation correlation characteristics. The specific steps include:

Set a spreading code set, including f spreading codes, denoted as A

The spreading code set is modulated onto ^ uncorrelated carriers to obtain an uncorrelated spreading code set, which are: AA + yi, A ₂ = A + / ₂ , ..., Α ^ = Α + / ^;

Considering the spreading code sets as a whole, and sending them in time slots respectively, so that N = N ^ N ₂ low correlation / zero correlation spreading code sets can be obtained. This spreading code set Do not:

SET ,: time slot 1, SET ,: time slot 2, ..., SET _Ni : time slot

SET _{Ni + l} A ₂ + time slot 1, SET _{Ni + 2} : A ₂ + time slot 2, ... "SET _2Nl : A ₂ + time slot V ₂ ,

SET _{MN2 + l} : Aw, + time slot 1, SET _{MN2 + 2} : A _N] + time slot 2, _··· , SET _NxN2 : Α + time slot N ₂ ; thus a low correlation / zero correlation spread spectrum is obtained A code set, where each spreading code set includes a spreading code.

22. The method according to claim 10, wherein the grouping the spreading codes comprises: jointly generating a set of low-correlation / zero-correlation spreading code sets from a code domain and a frequency domain, that is, dividing into Code frequency division low correlation / zero correlation spreading code sets, so that the code sets have low correlation / zero correlation correlation characteristics. The specific steps include:

Construct ^ code-score low-correlation / zero-correlation spreading code sets, respectively labeled A ₁ A ₂ , ......,

A _N , where each spreading code set contains a spreading code; Considering the ^ spreading code sets as a whole, and modulating them into uncorrelated carriers to send, so that = χ ₂ low-correlation / zero-correlation spreading code sets can be obtained, and this spreading code set is Do not:

SET _X : A, + _A , SET A, + / ₂ , ..., SET _Ni : A, + f _Ni ,

SET _{N2 +} : A ₂ + _Λ , SET _{N2 + 2} : A ₂ + / ₂ '…, SET _2N2 : A ₂ + f _Ni ,

^SET (N \) N ₂ + '· ^A W, ⁺ / l' ^SET {N \) N ₂ +2 · A. ^ + f ₂ ,. · ', SET _{N [N2} : ⁺ / w ₂ ; A low correlation / zero correlation spreading code set is obtained, where each spreading code set includes a spreading code.

The method according to claim 10, wherein the grouping the spreading codes comprises: generating a set of low-correlation / zero-correlation spreading codes from a time domain, a frequency domain, and a code domain together, That is, it is divided into time-frequency code division low correlation / zero correlation spreading code sets, so that there are low correlation / zero correlation correlation characteristics between ^ code sets; the specific steps include:

For a spreading code set containing a plurality of spreading codes, they are divided into NN ₂ , N ₃ low-correlation / zero-correlation spreading code sets from the code domain, time domain, and frequency domain, respectively, and combined to obtain N = N _} N ₂ N ₃ low correlation / zero correlation spreading code sets.

24. The method according to claim 1 or 10, wherein the grouping the spreading codes comprises: from any spreading code set that can be divided into multiple low-correlation / zero-correlation spreading The domain of the code set generates a set of low correlation / zero correlation spreading code sets.

25. The method according to claim 1 or 10, wherein the low correlation / zero correlation spreading code set further comprises: an equivalent transformation of any of the low correlation / zero correlation spreading code sets.

26. A method for applying a low-correlation / zero-correlation spreading code set, which comprises the following steps: Step 1. Obtain a low-correlation / zero-correlation guard frequency code set;

Step 2: Divide each cell into N non-overlapping regions;

Step 3: Allocate the obtained low-correlation / zero-correlation spreading code sets to the rings, and from the inner ring to the outer ring, the arrangement of the spreading code sets is described as: Step 4: Allocate power to users of each ring;

Step 5. In the neighboring cell, the spreading code set is arranged from the inner ring to the outer ring in a different arrangement manner from the local cell, and is marked as P _τ , P …….

The method according to claim 26, wherein the dividing each cell comprises: dividing each cell into concentric rings, and the size of the concentric rings can be based on service distribution or convenient design, etc. Depending on the conditions.

28. The method according to claim 26, wherein the dividing each cell comprises: determining a division according to a path loss, a size of interference, and a size of a signal-to-interference ratio; the division Can be irregular.

The method according to claim 28, wherein the determining N divisions according to the size of the signal-to-interference ratio comprises: in the downlink, all users are first allocated with the same transmit power, and the intra-cell measurement is performed. The levels of SNR in different areas, each cell is divided into N non-overlapping areas according to the different levels of SNR; in the uplink, all mobile stations are first allocated the same transmit power, and the user of each base station is measured The signal-to-interference ratio level, each cell is divided into N non-overlapping areas according to the different signal-to-interference ratio levels.

30. The method according to claim 26, wherein the performing power allocation to the users of each ring comprises: power allocation in a downlink and power allocation in an uplink; wherein: in the downlink In order to allocate power to users in each ring according to the power balance criterion, users in the outer ring have a large path loss and allocate a large amount of power, while users in the inner ring have a small path loss and allocate a small amount of power;

In the uplink, users in the inner ring reach the base station with more power than users in the outer ring reach the base station.

31. The method according to claim 26, wherein the performing more accurate power allocation for the users of each ring comprises: uplink and downlink uniformly complying with a signal-to-interference ratio balance criterion, that is, uplink In the base station, all users at the base station maintain the same signal-to-interference ratio level. In the downlink, all mobile stations maintain the same signal-to-interference ratio level.

32. The method according to claim 26, wherein, in a multi-cell networking environment, the arrangement of spreading code sets of different cells meets the following conditions: the outer ring of the neighboring cell allocates different spreading code sets as much as possible, That is, users in the outer ring of neighboring cells try to use code words in different code sets.

33. The method according to claim 26, characterized in that the channel allocation and the power allocation are not only applied to a cell network, but also applicable to a sector network and the like.

34. The method according to claim 26, wherein the channel allocation and the power allocation are also applicable to downlink and uplink.

The method according to claim 26, wherein the K-cross sequence in each spreading code set.

36. The method according to claim 26 or 33 or 34, wherein the channel allocation and the power allocation are applicable to a cell network, a sector network, and other similar networks, and are used in the downlink Applicable; it includes the following specific steps:

First determine a division, that is, the base station allocates the same power to all users, and then detect the signal-to-interference ratio level of different areas of each cell or sector, and divide each cell or sector into individual cells according to the level of the signal-to-interference ratio. Different parts, and other criteria can be used to determine N divisions;.

Allocating spreading code sets to the partitions, and at the same time ensuring that the partitions with relatively low interference to adjacent cells are assigned different spreading code sets;

Allocate larger power to users in areas with low SNR, and allocate less power to users in areas with high SNR, until the SNR levels of users in all areas are consistent.

37. The method according to claim 26 or 33 or 34, characterized in that the channel allocation and the power allocation method are applicable to a cell network, a sector network, and other similar networks, and are in the uplink Applicable; it includes the following specific steps:

First determine the division, the method is as follows: all mobile stations are allocated the same power, and then the user's SNR level in different areas of each cell or sector is examined, according to the level of the SNR level Divide each cell or sector into N different parts, and at the same time, other criteria can be used to determine the N divisions;

Allocating the spreading code sets to the partitions, and at the same time ensuring that the partitions with relatively low interference to the neighboring cells allocate different spreading code sets;

Allocate more power to users in areas with low SNR, and allocate less power to users in areas with high SNR, until the SNR levels of users in all areas are consistent.