Code Planning Manual

UTRAN Mobile Radio Networks

Scrambling Code Planning

V 2.0

July 2003

s

Scrambling Code Planning, UMTS FDD Macro Internal Use Only Version 2.0

Page 2 of 43

ICM N PG NM NE P1M. Koonert

Trademarks:

All designations used in this document can be trademarks, the use of which by third parties for their own purposes could violate the rights of their owners.

Copyright (C) Siemens AG 2003

Issued by the Information and Communication Mobile Group Hofmannstraße 51 D-81359 München

Technical modifications possible. Technical specifications and features are binding only insofar as they are specifically and expressly agreed upon in a written contract.

s


Page 3 of 43


Authors: Michael Koonert SIEMENS AG, ICM N PG NM NE P 1

s


Page 4 of 43


Contents

1 Introduction .............................................................................................................. 5

2 Downlink Scrambling ............................................................................................... 7 2.1 Downlink Physical Channels ................................................................................ 7 2.2 Downlink Scrambling Codes................................................................................. 7

3 Uplink Scrambling.................................................................................................. 10 3.1 Uplink Physical Channels ................................................................................... 10 3.2 Uplink Scrambling Codes ................................................................................... 11

4 Cell Search and Synchronization .......................................................................... 15

5 Scrambling Code Planning Approaches within one Network ................................ 17 5.1 Minimum Code Reuse Distance......................................................................... 17 5.2 Code Group Assignment .................................................................................... 22 5.3 Graph coloring based scrambling code assignment .......................................... 24 5.4 Special cases...................................................................................................... 25 5.4.1 Second Frequency for Macro Cells ................................................................. 26 5.4.2 Micro Cells ....................................................................................................... 26

6 Scrambling Code Planning at Neighboring Country Borders ................................ 27 6.1 Code Planning Principles at Borders and recommended Field-strength Levels 27 6.2 Preferential use of Frequencies.......................................................................... 28 6.3 Preferential Codes for UTRA.............................................................................. 29

7 Workarounds for Network Planning Tools ............................................................. 32 7.1 Scrambling Code Planning Example - Workaround........................................... 32

8 Conclusions ........................................................................................................... 38

8 Annex 1 – Release specific restrictions................................................................. 39

9 Annex 2 – Revision History ................................................................................... 40

10 Annex 3 – References ......................................................................................... 40

11 Annex 4 – Abbreviations...................................................................................... 41

12 Annex 5 – List of figures ...................................................................................... 43

13 Annex 6 – List of tables ....................................................................................... 43

s


Page 5 of 43


1 Introduction This document describes Network Planning related aspects of code planning in the downlink of the UTRA Physical Layer FDD mode and gives a short presentation of the scrambling operation in uplink. Code planning in uplink direction is not necessary due to the large amount of available codes and no restrictions of the available number of codes (as e.g. due to UE cell search time in downlink). Uplink codes are assigned to the UE by higher layers of the system. The downlink scrambling codes are assigned to the Node B during the planning phase and are not dynamically assigned, as it is the case for the mobiles and the uplink scrambling codes.

DATA

Bit Rate Chip Rate Chip Rate

Channelization Code Scrambling Code

DATA

Bit Rate Chip Rate Chip Rate

Channelization Code Scrambling Code

‘

figure 1: schematic view of spreading and scrambling

In the W-CDMA system as implemented for UMTS, spreading and scrambling is done in separate operations. These operations are applied to physical channels. The spreading operation (channelization) transforms every data symbol into a number of chips by applying an OVSF-code (orthogonal variable spreading factor), thus increasing the bandwidth of the signal. The number of chips per data symbol is called the Spreading Factor (SF). The second operation is the scrambling operation, where a scrambling code is applied to the spread signal. The scrambling (randomization) operation is a bit-wise operation onto a data sequence that does not lead to further bandwidth increase (cf. figure 1 and figure 2).

s


Page 6 of 43


figure 2: exemplary scrambling operation

Code planning can be applied to both spreading codes and scrambling codes. Spreading codes have to be planned with regard to efficient usage of the available OVSF code tree within one cell. This will be done dynamically by the system itself in operation. Scrambling codes in downlink have to be planned with regard to interference between different cells. In the remainder of this document only scrambling code planning is addressed.

s


Page 7 of 43


2 Downlink Scrambling

2.1 Downlink Physical Channels Each cell is allocated one and only one primary scrambling code. The primary CCPCH, primary CPICH, PICH, AICH, AP-AICH, CD/CA-ICH, CSICH and S-CCPCH carrying PCH are always transmitted using the primary scrambling code. The other downlink physical channels can be transmitted with either the primary scrambling code or a secondary scrambling code from the set associated with the primary scrambling code of the cell. The mixture of primary scrambling code and secondary scrambling code for one CCTrCH is allowable. However, in the case of the CCTrCH of type DSCH then all the PDSCH channelization codes that a single UE may receive shall be under a single scrambling code (either the primary or a secondary scrambling code).

2.2 Downlink Scrambling Codes The UMTS downlink scrambling code sequences are constructed as a complex combination of two real 10ms (38400 chips) segments of a totally 218-1 chip long Gold sequence. As scrambling does not affect data rates, the 38400 chip scrambling codes are repeated for every 10 ms radio frame. Overall 218-1 = 262,143 scrambling codes can be generated from the above-mentioned Gold sequence by simply applying an offset on the first bit. In practice, only a subset of 8192 codes from all the available codes is selected for the normal mode of operation. This is mainly due to simplification of the cell search procedure. In addition to the 8192 scrambling codes (code number k=0,1,2…8191) two groups containing as well 8192 scrambling codes are selected from the overall available codes as left (code number k+8192) and right (code number k+16384) alternative scrambling codes. Each code of the main group is accompanied by a left alternative scrambling code and a right alternative scrambling code, which can be used for compressed mode operation e.g. during inter-system measurements to ensure that a free channelization code with half of the spreading factor (SF/2) is available. The subset of 8192 scrambling codes used in the downlink is further divided into 512 sets where each set comprises one primary scrambling code and 15 secondary scrambling codes. The partitioning of the codes into primary and secondary codes is shown in figure 3.

s


Page 8 of 43


........1 2 3 40 5

11

512 Downlink Primary Scrambling Codes

1 2 3 40 8192

5 6 7 8 9 10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33 ...

15 Secondary Codes per Primary Code..............

........1 2 3 40 5

11

1 2 3 40 511

512 Downlink Primary Scrambling Codes

1 2 3 40 8192

5 6 7 8 9 10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33 ...

15 Secondary Codes per Primary Code..............

figure 3: primary and secondary scrambling code grouping

From network planning point of view, only the assignment of the primary scrambling code is relevant. Secondary scrambling codes are identified with regard to the already associated primary code and on demand dynamically assigned by the system. The secondary scrambling codes can be assigned to all downlink channels except the common channels that have to be heard in the whole cell. If e.g. a user demands an additional downlink dedicated channel which could not be established with the existing spreading code tree, this channel could be assigned a secondary scrambling code which has not been used until then, thereby allowing usage of a complete available spreading code tree at the cost of increased interference. Secondary scrambling codes also could be assigned e.g. to single beams for smart antennas, whereas the primary scrambling code is assigned to the entire cell. In this case, different beams within one cell are logically separated by different secondary scrambling codes. A secondary scrambling code may be used to introduce a second channelization code tree in case of fragmentation of the previously used spreading code tree. However, introduction of the secondary scrambling code in one cell increases the downlink interference due to decreased orthogonality. Therefore, as many users as possible should be kept under the first scrambling code, only those users not fitting under the primary scrambling code should be assigned to the secondary code.

The set of 512 primary scrambling codes is further divided into 64 scrambling code groups, each consisting of 8 primary scrambling codes (for illustration cf. figure 4). Each cell is allocated one and only one primary scrambling code (this is the motivation of code planning). The primary CCPCH physical channel, which carries the BCH transport channel, and the primary CPICH are always transmitted using the primary scrambling code. The other downlink channels except the synchronization channels

s


Page 9 of 43


(which are not scrambled) can be transmitted with either the primary or one of the secondary codes assigned to the code group of the cell. The detailed construction rules and numbering schemes for the scrambling code groups can be found in [1], the cell search procedure of the mobile is described in more detail in the following chapter.

Group of Primary Scrambling Codes

512 Elements

Group #08 Elements

Group #638 Elements

8192 Downlink Scrambling Codes512 Groups with 16 Elements each

(1 primary, 15 secondary codes)

#0 primary15 secondary


Group 1 Group 512


512 Elements

Group #08 Elements

Group #638 Elements


512 Elements

Group #08 Elements

Group #638 Elements

8192 Downlink Scrambling Codes512 Groups with 16 Elements each

(1 primary, 15 secondary codes)



Group 1 Group 512

figure 4: primary scrambling code grouping scheme

For radio network planning, the assignment of the primary scrambling code group is of major relevance. A careful choice of scrambling codes within a group of adjacent cells can help to decrease the co-channel interference and can improve network performance with regard to code acquisition time from the UE point of view.

s


Page 10 of 43


3 Uplink Scrambling This chapter deals with the general description of the structure of uplink scrambling codes for UMTS FDD as standardized in [1]. Its aim is just to give a more complete overview and general insight into interdependencies between downlink assignments and according uplink scrambling code assignments, thereby emphasizing that there is no need for uplink scrambling code planning once the downlink scrambling code assignment has been done. So if the reader is looking for planning rules, this chapter can easily be skipped.

3.1 Uplink Physical Channels Uplink physical channels can be classified as either dedicated or common channels.

There are two types of uplink dedicated physical channels, the uplink Dedicated Physical Data Channel (uplink DPDCH) and the uplink Dedicated Physical Control Channel (uplink DPCCH). The uplink DPDCH is used to carry the DCH transport channel. There may be zero, one, or several uplink DPDCHs on each radio link. The uplink DPCCH is used to carry control information generated at Layer 1. The Layer 1 control information consists of known pilot bits to support channel estimation for coherent detection, transmit power-control (TPC) commands, feedback information (FBI), and an optional transport-format combination indicator (TFCI). The transport-format combination indicator informs the receiver about the instantaneous transport format combination of the transport channels mapped to the simultaneously transmitted uplink DPDCH radio frame. There is one and only one uplink DPCCH on each radio link. Multi-code operation is possible for the uplink dedicated physical channels. When multi-code transmission is used, several parallel DPDCH are transmitted using different channelization codes, see [1]. However, there is only one DPCCH per radio link.

Uplink common physical channels are the Physical Random Access Channel (PRACH) used to carry the RACH transport channel, and the physical common packet channel (PCPCH), used to carry the CPCH transport channel. The random-access transmission on the RACH is based on a Slotted ALOHA approach with fast acquisition indication (see [3]). The UE can start the random-access transmission at the beginning of a number of well-defined time intervals, denoted access slots. There are 15 access slots per two frames and they are spaced 5120 chips apart. Higher layers give information on what access slots are available for random-access transmission.

The structure of the random-access transmission is shown in figure 4. The random-access transmission consists of one or several preambles of length 4096 chips and a message of length 10 ms or 20 ms.

s


Page 11 of 43


Message partPreamble

4096 chips10 ms (one radio frame)

Preamble Preamble

Message partPreamble

4096 chips 20 ms (two radio frames)

Preamble Preamble

figure 5: structure of the random-access transmission

Each preamble is of length 4096 chips and consists of 256 repetitions of a signature of length 16 chips. There are a maximum of 16 available signatures (see [1] for details).

The CPCH transmission is based on DSMA-CD (Digital Sense Multiple Access with Collision Detection) approach with fast acquisition indication. The UE can start transmission at the beginning of a number of well-defined time-intervals, relative to the frame boundary of the received BCH of the current cell. The access slot timing and structure is identical to RACH as described above. The PCPCH access transmission consists of one or several Access Preambles (A-P) of length 4096 chips, one Collision Detection Preamble (CD-P) of length 4096 chips, a DPCCH Power Control Preamble (PC-P) which is either 0 slots or 8 slots in length, and a message of variable length N*10 ms. The structure is shown in figure 6.

4096 chips

P0P1

Pj Pj

Collision DetectionPreamble

Access Preamble Control Part

Data part

0 or 8 slots N*10 msec

Message Part

figure 6: structure of the CPCH access transmission

3.2 Uplink Scrambling Codes All uplink physical channels are subjected to scrambling with a complex-valued scrambling code. There are 224 long and 224 short uplink scrambling codes. Higher layers of the system assign uplink scrambling codes. Long scrambling code sequences as complex-values sequences are constructed from two real-valued m-sequences (segments of a set of Gold sequences) and have a length of 38400 chips. Short scrambling sequences are of length 256 chips.

s


Page 12 of 43


The DPCCH/DPDCH may be scrambled by either long or short scrambling codes assigned on-demand when the dedicated channel is to be set up. The PRACH message part is scrambled with a long scrambling code. Also the PCPCH message part is scrambled with a long scrambling code. Long and short uplink scrambling codes are defined in detail in [1].

The scrambling code for the PRACH preamble part is constructed from the long scrambling sequences. There are 8192 PRACH preamble scrambling codes in total. The n-th preamble scrambling code, n = 0, 1, …, 8191, is defined as:

Sr-pre,n(i) = Clong,1,n(i), i = 0, 1, …, 4095; where the sequence Clong,1,n can be found in [1]. The 8192 PRACH preamble scrambling codes are divided into 512 groups with 16 codes in each group. There is a one-to-one correspondence between the group of PRACH preamble scrambling codes in a cell and the primary scrambling code used in the downlink of the cell. The k-th PRACH preamble scrambling code within the cell with downlink primary scrambling code m; k = 0, 1, 2, …, 15 and m = 0, 1, 2, …, 511, is Sr-pre,n(i) as defined above with n = 16m+k. This correspondence is illustrated in figure 7.

1 2 3 4m =0 511512 Downlink Primary

Scrambling Codes ........Assignment of one primary scrambling code for each celldefines set of 16 PRACH preamble codes in uplink

65

66

67

68

79

...n =6

4

n=16m+k, k=0..15 (0<=n<=8191)

Set of 16 PRACH preamble scrambling codes

1 2 3 4m =0 511512 Downlink Primary

Scrambling Codes ........Assignment of one primary scrambling code for each celldefines set of 16 PRACH preamble codes in uplink

65

66

67

68

79

...n =6

4

n=16m+k, k=0..15 (0<=n<=8191)

Set of 16 PRACH preamble scrambling codes

figure 7: correspondence between downlink primary scrambling code and uplink PRACH preamble scrambling code group

s


Page 13 of 43


The scrambling code used for the PRACH message part is 10 ms long, and there are 8192 different PRACH scrambling codes defined. The n-th PRACH message part scrambling code, denoted Sr-msg,n, where n = 0, 1, …, 8191, is based on the long scrambling sequence and is defined as: Sr-msg,n(i) = Clong,n(i + 4096), i = 0, 1, …, 38399

where the lowest index corresponds to the chip transmitted first in time and Clong,n is defined in [1]. The message part scrambling code has a one-to-one correspondence to the scrambling code used for the PRACH preamble part. For one PRACH, the same code number is used for both scrambling codes, i.e. if the PRACH preamble scrambling code used is Sr-pre,m then the PRACH message part scrambling code is Sr-msg,m, where the number m is the same for both codes.

The assignment of one primary scrambling code per cell in the downlink direction uniquely defines the set of 16 PRACH preamble scrambling codes. The message part scrambling code has a one-to-one correspondence to the scrambling code used for the preamble part, the code number for the message part of the PRACH is the same as for the preamble part.

The set of scrambling codes used for the PCPCH message part are 10 ms long, cell-specific, and each scrambling code has a one-to-one correspondence to the signature sequence and the access sub-channel used by the access preamble part. Either long or short scrambling codes can be used to scramble the CPCH message part. There are 64 uplink scrambling codes defined per cell and 32768 different PCPCH scrambling codes defined in the system. The n-th PCPCH message part scrambling code, denoted Sc-msg,n, where n = 8192,8193, …,40959 is based on the scrambling sequence and is defined as:

In the case when the long scrambling codes are used: Sc-msg,n(i) = Clong,n(i ), i = 0, 1, …, 38399

where the lowest index corresponds to the chip transmitted first in time and Clong,n is defined in [1]. In the case the short scrambling codes are used:

Sc-msg,n(i) = Cshort,n(i), i = 0, 1, …, 38399 where the sequence Cshort,n can be found in [1].

The 32768 PCPCH scrambling codes are divided into 512 groups with 64 codes in each group. There is a one-to-one correspondence between the group of PCPCH preamble scrambling codes in a cell and the primary scrambling code used in the downlink of the cell. The k-th PCPCH scrambling code within the cell with downlink primary scrambling code m; k = 16, 17, …, 79 and m = 0, 1, 2, …, 511, is Sc-msg,n as defined above with n = 64m+k+8176.

s


Page 14 of 43


As we have seen, for the uplink common channels, there is a one-to-one mapping between the scrambling code assigned in downlink and the associated uplink resources. In combination with the code planning procedure in downlink, where different primary scrambling codes are assigned to neighboring cells, this assignment procedure ensures the assignment of different uplink scrambling codes without further need for planning. The main goal is the usage of different codes in neighboring cells and this target is already reached by the automatic assignment correspondence between downlink and uplink, i.e. during scrambling code planning for downlink codes as described e.g. in chapter 5.

s


Page 15 of 43


4 Cell Search and Synchronization W-CDMA as used for UMTS is an asynchronous system, which makes the cell search procedure more complicated in comparison to a synchronous system. During the cell search, the user equipment determines the downlink scrambling code and the common channel frame synchronization of the desired cell. The cell search procedure itself is not standardized; an informative example is included in [2]. In general the cell search can be divided into three steps, though from standardization there are no requirements, which steps to perform and in which order:

• Slot synchronization

• Frame synchronization and code-group identification

• Scrambling code identification

During the slot synchronization step of the cell search procedure the UE uses the synchronization channel’s (SCH) primary synchronization code, a 256-chip sequence, to acquire slot synchronization to a cell. This is typically done with a single matched filter (or any similar device) adjusted to the primary synchronization code, which is common to all cells and each slot. The general structure of the synchronization channel is depicted in figure 8. The primary SCH code word is sent without modulation, it is constructed as a repetition of shorter 16-chip sequences to reduce complexity of the required hardware at the UE. The slot timing of the cell can be obtained by detecting peaks in the matched filter output. The detected peak corresponds to the slot boundary.

PrimarySCH

SecondarySCH

256 chips

2560 chips

One 10 ms SCH radio frame

acsi,0

acp

acsi,1

acp

acsi,14

acp

Slot #0 Slot #1 Slot #14

figure 8: structure of the synchronization channel (SCH)

During the frame synchronization and code-group identification step of the cell search, the UE uses the SCH’s secondary synchronization code to find frame synchronization and identify the code group of the cell detected in the first step. There are 64 different

s


Page 16 of 43


secondary SCH sequences (denoted by index i in figure 8), each consisting of 15 codes of each 256-chip length as defined by a look-up table in [3]. Each possible primary scrambling code group is assigned to one unique secondary SCH code sequence. Over the duration of one frame, the UE detects the secondary SCH code in 15 consecutive slots. The code repetition pattern is unique per code group even after a cyclic shift, so that after the detection of the code sequence, the code group as well as the frame synchronization is determined.

During the scrambling code identification step of the cell search procedure, the UE determines the exact primary scrambling code used by the selected cell. The primary scrambling code is typically identified through symbol-by-symbol correlation over the common pilot channel (CPICH) with all codes within the code group identified in the second step of the cell search. After the primary scrambling code has been identified, the primary common control physical channel (P-CCPCH) can be detected and the system- and cell specific broadcast channel (BCH) information can be read.

If the UE has received information about which scrambling codes to search for, within a certain tradeoff steps 2 and 3 above could be simplified. If e.g. only the first primary scrambling code of each group of eight codes is assigned during the planning (and the UE is aware of that by operator specific settings and capable of handling this information!) step 3 contains only one correlation operation. Another possibility would be to assign only codes from one group of primary scrambling codes. In this case (again, UE-knowledge assumed!), step 2 would be manageable within one search step. However, there will always be a tradeoff between optimization of the one search step at the expense of the other search step.

After cell search, the radio frame timing of all common physical channels can be determined. Good performance of a direct sequence system, such as DS-CDMA, is correlated to UEs with the ability of fast synchronization. This ability is dependent on two issues, firstly the implementation of the code acquisition strategy within the UE and second, the scrambling code planning in the network carried out by radio network planning. The above-mentioned topics are of relevance for the initial cell search, when the UE is switched on, and for handover and cell-selection / -reselection procedures.

s


Page 17 of 43


5 Scrambling Code Planning Approaches within one Network Code planning (more exact: downlink scrambling code group allocation) in W-CDMA resembles frequency planning in GSM, but it is not such a key performance factor as frequency planning is in frequency division multiple access systems. The number of available downlink scrambling codes is high. In contrast to frequency planning, in scrambling code planning it is not crucial which scrambling codes are assigned to neighbors as long as they are not the same codes. Cross-correlation properties (and thereby interference properties) of mutually different codes from a Gold-sequence are comparatively low and approximately the same for all possible combinations of different codes.

Scrambling codes are assigned per sector, i.e. in a configuration with three sectors per site, three primary scrambling codes per site are needed. In general the set of available codes is large enough, the initial code planning procedure can be automated and included in a radio network planning tool.

For a mobile, all available cells should have different scrambling codes. A simple method would be the usage of different scrambling code groups in neighboring cells, the reuse could for example be 64, equal to the number of code groups.

5.1 Minimum Code Reuse Distance This approach emanates from the requirement of a certain allowable ‘carrier to co-channel interference ratio’. The major task is to determine the minimum distance of two cells that can have an identical code set for downlink scrambling codes. A tolerable co-channel interference threshold is assumed, the calculations lead to a minimum code reuse distance for a homogenous hexagonal cell layout. In all cells at least this distance apart from each other, the same code can be used again. The task is comparable to frequency planning in GSM and results in a kind of ‘code reuse factor’. A uniform cluster distribution is assumed and from geometric considerations a minimum cluster size corresponding to a minimum number of used scrambling code groups can be calculated. A typical cellular layout is shown in figure 9.

The number of cell tiers K surrounding the center cell of a cluster can be calculated from the number of cells in one cluster N (cf. equation 1). E.g. 7 cells per cluster result in one surrounding layer, 19 cells per cluster result in two tiers and so forth.

��

��

��

��

� +⋅=6

5log)2log(

1 NceilK (1)

s


Page 18 of 43


reuse distance D

Rreuse distance D

R

‘

figure 9: cell grouping for code reuse clustering with a 7 cell cluster

An estimation of the observed co-channel carrier-to-interference ratio in an hexagonal cell layout as shown in figure 9 can be derived as follows. The worst-case distance of a user to the serving cell is the cell radius R. All interfering co-channel cells of the first surrounding tier are approximately a distance D apart from the observed user. All interferers more than one tier apart from the cell under consideration are not taken into account, as the pathloss is approximately proportional to the 4th power of the distance1. Let M denote the number of surrounding co-channel cells in the first tier, for a hexagonal layout with omni-directional antennas M equals 6. Assuming an interference level well above the noise level, the co-channel carrier-to-interference ratio (in linear scale) for the user under investigation can be approximated by

MN

DMR

IC 2

4

4 9=⋅

≈ −

−

(2)

1 Therefore, for interferers in the second surrounding tier let us assume ≈10dB additional pathloss in comparison to interferers in the first surrounding tier. There are six interferers in the first surrounding tier and 12 additional interferers in the second surrounding tier. The amount of interference generated by the second tier is approximately 7dB lower than the interference generated by the first tier ( ( ) dBdBPP ii 710126 ≈− ).

s


Page 19 of 43


where NRD 3/ = denotes the ratio of reuse-distance to cell radius for the hexagonal layout as a function of the number of cells per cluster. Using the two equations above, a lower limit for the number of tiers needed to fulfil a certain co-channel C/I requirement in the center cell can be approximated as follows

( )��

�

�

��

�

�

��

�

�

��

�

�+�

�

��

�⋅≥

18

15log

2log1 linearI

CMceilK (3)

For M=6 surrounding co-channel interferers in the first tier the simplified calculation can be carried out as given in the following formula:

��

�

�

��

�

�−

��

�

�

��

�

�+�

�

��

�≥ 263.01163.0log322.3linear

ionalomnidirect ICceilK (4)

where the ‘ceil’ operation rounds the result up to the next integer. The total number of primary codes in this cluster is equal to the number of cells N and as a function of the number of tiers K given by equation 4

( )1261 −+= KN (5)

So an exemplary co-channel interference requirement below or equal to 40dB would result into 4 tiers with 91 cells and therefore 91 different scrambling code groups. For an exemplary C/I of 27dB (for an omnidirectional scenario, sectorization increases the co-channel attenuation to a value higher than 27dB) 2 tiers with only 19 scrambling code groups would be sufficient to ensure the required co-channel requirements. This example is illustrated in figure 10.

s


Page 20 of 43


54

16

72

316

1718

198

9

10

111214

13

15

’

figure 10: exemplary cluster reuse of 19 code groups

If the scrambling codes are assigned with the same pattern in each cluster (cluster-reuse), sectorization ensures that all cells with the same scrambling code have the same look direction in the different clusters. This leads to a further decrease in the co-channel C/I so that there is an additional attenuation because of the sectorization and the corresponding antenna pattern.

Assuming sectorization with three sectors per cell, there is a reduction of the number of visible co-channel interferers in the first tier of interferers with regard to the observed cell. This is due to the reduction in antenna horizontal beamwidth and in this case the factor M in the equations above can be chosen equal to 2 surrounding co-channel interferers in the first tier (for illustration cf. figure 11).

s


Page 21 of 43


figure 11: reduced number of interferers for sectorization

This leads to an additional decrease of approximately 4.77dB in co-channel interference, which is included in the following formula derived by including M=2 into the detailed expression (3) as shown above:

��

�

�

��

�

�−

��

�

�

��

�

�+�

�

��

�≥ 263.010943.0log322.3seclinear

torized ICceilK (6)

An example how a number of necessary code groups can be derived from a requirement on co-channel attenuation in a sectorized cell layout is given below. However, the value for the requirement should be aligned with the operator in each specific project.

s


Page 22 of 43


5.2 Code Group Assignment The assignment of code groups e.g. with a reuse factor of 64, equal to the number of groups of primary scrambling codes, can lead to an advantage in terms of code acquisition time for the UE (cf. chapter 3). The speed of the code acquisition depends on the match between scrambling code allocation in the network and the acquisition strategy applied in the UE. However, the implementation in the mobile is manufacturer specific and the mobile shall perform well for any scrambling code allocation. For release specific restrictions see Annex 1.

With certain manufacturer specific implementations, a reuse within the codes of a certain limited number of code groups would simplify step 2 of the cell search procedure, frame synchronization and code-group identification. Step 3, the scrambling code identification becomes more complicated in this case. Another possibility is the reuse of e.g. one code from each code group, thereby simplifying step 3 at the cost of a more complicated step 2 of the cell search. Reduced cell search time is beneficial for the initial synchronization when the mobile enters a network for the first time as well as for soft handover and cell-reselection procedures.

An example (for simplification of illustration with reuse 7) is illustrated in the following figures. A sub- optimal configuration is shown in figure 12. All cells in all clusters have been assigned scrambling codes from code group 1 (Notation x,y for scrambling code group, scrambling code number). In this constellation, the neighboring cells of the actual cell in which the UE is located have the same code group as the actual cell, therefore step 3 of the cell search procedure (scrambling code identification) has to be carried out.

Example: 32dB required co-channel attenuation → 6 Equation2 Layers

surrounding center cell needed → 5 Equation 19 hexagons per cluster.

As we have 3 sectors per hexagon, we would need 3 times 19 = 57 different scrambling codes to fulfill the requirement of 32dB co-channel attenuation. If the cell radius is known, the reuse distance can be calculated as

RNRD 55.73 ≈= for this example.

s


Page 23 of 43


1,21,4

1,51,7

1,1 1,61,3

1,21,4

1,51,7

1,1 1,61,3

1,21,4

1,51,7

1,1 1,61,3

UEUE

’

figure 12: sub-optimal scrambling code group allocation

An alternative, optimal code group allocation is shown in figure 13. All neighboring cells for the UE have the first primary code of different code groups assigned. In this example, the identification of the code group (cell search procedure step 2) is sufficient, if the UE ‘knows’ that the first code of each group is assigned in all cells. Cell search procedure step 3 can be omitted in this case. The information for the UE can be included in the neighbor cell list. This list is not available in case of initial cell search but only for cell reselection and handover. In case of initial cell search, all three search steps have to be carried out.

2,14,1

5,17,1

1,1 6,13,1

2,14,1

5,17,1

1,1 6,13,1

2,14,1

5,17,1

1,1 6,13,1

UEUE

’ figure 13: optimal scrambling code group allocation

As mentioned before, this example is only valid, if the cell search procedure, which is not standardized, is implemented in the mobile exactly in the assumed way. All other mobiles would not gain from the code assignment in the example, for those mobiles, a different code assignment strategy could be beneficial.

s


Page 24 of 43


5.3 Graph coloring based scrambling code assignment In [4] the benefit of another assignment method, with increased complexity compared to the cluster reuse based code assignment described in the previous sections, has been investigated. In the paper it is shown that for a set of examples the graph coloring based method yielded more efficient results than the cluster reuse based method. This especially holds for non-uniform network layouts.

In general, graph coloring methods deal with the task of assigning one color out of a minimized set of available colors (assume colors with indices {1,2,3,…}) to each node of a graph. This is done under the constraint, that a minimum difference between color indices for all possible combinations of two nodes in the graph is guaranteed. This task can be mapped on the task of assigning a scrambling code to each cell of a network.

However, the graph coloring problem is an NP-complete problem (i.e. feasibility is ‘checkable’ by an efficient algorithm, but an optimal solution can only be found by very time-consuming search or with a lot of luck in the first assumption). Therefore, efficient algorithms resulting in suboptimal solutions for the graph coloring have been introduced. The algorithm proposed in [4] has already been introduced for channel assignment in FDMA systems and can therefore be reused in scrambling code assignment for W-CDMA.

In a first step, all cells in the network are sorted with regard to the number of neighboring cells within the reuse distance. In inhomogeneous layouts with small cells surrounded by larger cells, the small cell will have fewer neighbors within the reuse distance as in a homogeneous layout (cf. figure 14).

s


Page 25 of 43


figure 14: number of neighboring cells with regard to cell layout

The graph coloring based scrambling code assignment procedure starts with assigning a primary scrambling code to the cell with the lowest number of neighboring cells within the reuse distance. Now, step-by-step, for each cell in the list with ascending number of neighbor cells, one primary scrambling code index as low as possible is assigned. This happens under the constraint that the index has to be greater than all indices previously assigned to all cells with a neighbor relation to the current cell. This approach will result in a minimum total number of assigned primary scrambling codes, still fulfilling the requirement with regard to the reuse distance.

The input of additional knowledge into the assignment procedure (knowledge about neighbor relations between different cells) seems to be especially promising in case of hierarchical cell structures in multi-layer networks. In this case significant benefits in terms of total number of assigned scrambling codes in comparison to cluster reuse based methods can be expected.

5.4 Special cases In general, code planning should mainly take into account that the same downlink scrambling code is not used twice in neighboring cells. Clustering is not mandatory but may be easier to handle for code planning approaches as introduced above. For some special configurations, additional suggestions are applicable.

D

b) inhomogeneous cell size

D

a) homogeneous cell size

D

b) inhomogeneous cell size

D

a) homogeneous cell size

s


Page 26 of 43


5.4.1 Second Frequency for Macro Cells If an operator introduces a second carrier with macro cells, the complete set of scrambling codes can be used as well on the first as on the second frequency. For the multi-layer network, planning should take into account that the same code does not occur again in all cells in the neighbor list of one cell, i.e. all possible neighbors have different codes than the code in the actual cell. If the first layer was planned with clusters as described above, an offset between the code clusters of the first and the second layer could solve the problem.

5.4.2 Micro Cells

If an operator introduces a hierarchical cell structure with micro cells on an additional carrier, the same rule as above applies. All possible neighbors of a cell should be assigned to different scrambling codes.

In case of micro cells on the same frequency, the requirement on neighbors applies as well. In addition, the available code groups could be split into a subgroup of codes only used in micro cells and a second subgroup of codes only used in macro cells.

s


Page 27 of 43


6 Scrambling Code Planning at Neighboring Country Borders In [5] the European Radiocommunications Committee (ERC) within the European Conference of Postal and Telecommunications Administrations (CEPT) recommends that co-ordination between UMTS/IMT-2000 systems in border areas shall be based on bilateral or multilateral agreements between Administrations. A summary of the recommendation is included in the following.

6.1 Code Planning Principles at Borders and recommended Field-strength Levels The coordination in border areas is based on the following concept:

1. Preferential code groups or preferential code group blocks are selected as a subset of the overall available number of scrambling codes. Preferential code groups or code group blocks (assigned to each country involved) shall be agreed between Administrations concerned (see chapter 6.3 for guidance). This case (1) is applicable between countries where center frequencies for the frequency blocks defined during spectrum assignment are aligned.

2. Frequencies in the band 2110-2170 MHz may be used without coordination with a

neighboring country if the predicted mean field strength of each carrier produced by the base station does not exceed a value of 45 dBµV/m/5MHz. This is only valid for systems using preferential codes, or where center frequencies are not aligned, or for systems not using a CDMA IMT-2000 radio interface. The field strength is measured at a height of 3 m above ground at and beyond the borderline between two countries2. Administrations may agree by bilateral and/or multilateral coordination agreement a reference line at some distance beyond the border.

3. In the bands 1900-1980 MHz and 2010-2025 MHz TDD systems may be used

without coordination with a neighboring country if the predicted mean field strength of each carrier produced by the base station does not exceed a value of 36 dBµV/m/5MHz3. This is valid for TDD systems using preferential codes, or where center frequencies are not aligned. The field strength is measured at a height of 3 m above ground at and beyond the borderline between two countries2. Administrations may agree by bilateral and/or multilateral coordination agreement a reference line at some distance beyond the border.

2 Depending on the propagation model, the area beyond the border, which is relevant, may be agreed by the

concerned Administrations. 3 The value will be reconsidered when the recommendation is reviewed within 2 years of its original adoption.

s


Page 28 of 43


4. Frequencies used at the border may be used without coordination with a

neighboring country if the predicted mean field strength of each carrier produced by the base station does not exceed a value of 21 dBµV/m/5MHz3. This is valid for systems using non-preferential codes and with center frequencies aligned. The field strength is measured at a height of 3 m above ground at and beyond the borderline between two countries2.

Note: In bilateral and / or multilateral coordination agreements levels up to 15-20dB higher might typically be agreed between the Administrations concerned.

In practice, the essence of these possibilities is that in neighboring countries applying UMTS-FDD, in a first step, the total available scrambling codes should be split among the neighbors. If the predicted field strength level of one countries network is then below a given threshold at a certain point in the other country, no further coordination is required.

6.2 Preferential use of Frequencies In Annex 3 to document [5] the preferential frequencies for use at borders are recommended. For some cross-border situations, in addition to code co-ordination, it may be possible to agree on a frequency co-ordination based on preferential frequencies, while ensuring a fair treatment of different operators within a country. The set of totally available frequencies is split into preferential and neutral frequencies as exemplary depicted in figure 15. Frequencies preferential in one country are the non-preferential frequencies in the neighboring country and vice versa, thereby ensuring certain interference levels as long as no non-preferential frequencies are assigned. This could be implemented in the case when UMTS-FDD is used as described in the following (also illustrated in figure 16).

figure 15: exemplary assignment of preferential and neutral frequencies

preferential neutral non preferential Country A

Country Bpreferentialneutralnon preferential

Spectrum Allocation

preferential neutral non preferential Country A

Country Bpreferentialneutralnon preferential

Spectrum Allocation

s


Page 29 of 43


1. Preferential frequencies (or preferential frequency bands) and neutral frequencies

(or neutral frequency bands) shall be agreed between Administrations concerned.

2. Preferential frequencies may be used without co-ordination with a neighboring country if the predicted mean field strength of each carrier produced by the base station does not exceed a value of 65 dBµV/m/5 MHz at a height of 3m above ground at and beyond the borderline between two countries.

3. Neutral frequencies may be used without co-ordination with a neighboring country

if the predicted mean field strength of each carrier produced by the base station does not exceed a value of 45 dBµV/m/5MHz at a height of 3m above ground at and beyond the borderline between two countries.

4. Non-preferential frequencies may be used without co-ordination with a

neighboring country if the predicted mean field strength of each carrier produced by the base station does not exceed a value of 45 dBµV/m/5MHz at a height of 3m above ground at and beyond the borderline between two countries.

Systems operating on non-preferential frequencies must accept interference from services in the neighboring country using preferential frequencies.

Country A(Neutral)

Country B(Neutral)

45 dBµµµµV/m/5MHz 45 dBµµµµV/m/5MHz

Equal field strength limits at border

Country A(Preferential)

Country B(Non-preferential)

65 dBµµµµV/m/5MHz 45 dBµµµµV/m/5MHz

Interference to Rx accepted(potential capacity loss)

figure 16: illustration of neutral and preferential frequency scenarios

6.3 Preferential Codes for UTRA In Annex 4 to document [5] the preferential codes for use at borders are recommended. The code groups defined for the FDD and TDD modes have no particular correlation properties and no particular organization of the repartition is required. Administrations should agree on a repartition of this code groups on an equitable basis. In any case, each country can use all code groups in the most important part of its own territory.

s


Page 30 of 43


In border areas, the codes will be divided into 6 "code sets" containing each one sixth of the available code groups. Each country is allocated three code sets (half of the codes) in a bilateral case, and two code sets (one third of the codes) in a trilateral case.

Four types of countries are defined in a way such that no country will use the same code set as any one of its neighbors (cf. figure 17). The following lists describe the distribution of European countries:

Type country 1: BEL, CVA, CYP, CZE, DNK, E, FIN, GRC, IRL, ISL, LTU, MCO,SMR, SUI, SVN, UKR, YUG Type country 2: AND, BIH, BLR, BUL, D, EST, G, HNG, I, MDA, RUS (Exclave) Type country 3: AUT, F, HOL, HRV, MKD, POL, POR, ROU, RUS, S, TUR Type country 4: ALB, LIE, LUX, LVA, NOR, SVK.

For each type of country, the following tables and figure describe the sharing of the codes with its neighboring countries, with the following conventions of writing:

Preferential code non-preferential code

table 1: codes at neighboring countries, conventions of writing For the FDD mode the 64 scrambling code groups as defined in [1], numbered {0..63}, hereafter will be called code groups. In the area between three countries (called Zone x-y-z in the following table) each country is assigned two subsets of all available code groups, e.g. for country type 2, 3 and 4 country 2 is assigned code group 21-42, country 3 code group 43-63 and country 4 code group 0-20. For the borderline between only two countries, each country obtains three subsets as preferential codes.

Set A Set B Set C Set D Set E Set F Set A Set B Set C Set D Set E Set F

Country 1 0..10 11..20 21..31 32..42 43..52 53..63 Country 2 0..10 11..20 21..31 32..42 43..52 53..63

Border 1-2 Border 2-1

Zone 1-2-3 Zone 2-3-1


Zone 1-2-4 Zone 2-1-4


Zone 1-3-4 Zone 2-3-4

Set A Set B Set C Set D Set E Set F Set A Set B Set C Set D Set E Set F

Country 3 0..10 11..20 21..31 32..42 43..52 53..63 Country 4 0..10 11..20 21..31 32..42 43..52 53..63


Zone 3-1-2 Zone 4-1-2


Zone 3-1-4 Zone 4-2-3


Zone 3-2-4 Zone 4-3-1

table 2: code sharing between neighboring countries

s


Page 31 of 43


Notes

1. All codes are available in areas away from the border where the field strengths into the neighboring country are below the relevant trigger levels.

2. For the other IMT-2000 CDMA radio interface (IMT-MC, or cdma2000),

preferential code allocation schemes are still to be developed.

3. A two countries code sharing should be applied or used by base stations that exceed the relevant trigger level (cf. chapter 6.1) of only one neighboring country. A three countries code sharing should be applied or used by base stations that exceed the relevant trigger level (cf. chapter 6.1) of two neighboring countries.

4. In certain specific cases (e.g. AUT/HRV) where the distance between two

countries of the same Type number is very small (< few 10s km), it may be necessary to address the situation in bi-/multilateral co-ordination agreements as necessary, and may include further subdivision of the allocated codes in certain areas.

Country 1 = Country 2 = Country 3 = Country 4 =

-

Vatican CVA = Country 1

-

Monaco MCO = Country 1

San Marino SMR = Country 1

Andorra AND = Country 2

-

Liechtenstein LIE = Country 4

HOL

AUT

I

CYP

POR

POL

MDA

IRC

S

RUS

UKR

E ALB

LUX

D

SUI

BEL

CZE SVK

F

ROUHNG

SVN

BIHYUG

EST

HRV

G

DNK

MKDBUL

GRC

LTU

BLR

ISL

NOR

RUS(E)

FIN

LVA

TUR

figure 17: classification of country types in Europe

s


Page 32 of 43


7 Workarounds for Network Planning Tools As long as no downlink scrambling code modules are available in the network planning tools currently in use, scrambling code assignment can be carried out with the aid of different workarounds. Possible candidates in DMS3.1 and earlier releases are the frequency-planning tool as known from GSM, the color-code assignment module and the CDMA (IS-95) module. A code planning module will be available in DMS3.2 as well as in Aircom Enterprise 4.1. Input should be the number of available primary scrambling codes and the minimum reuse distance for identical codes or the equivalent pathloss. In a first step, each cell should be assigned to one of the 64 available groups of primary scrambling code groups. If this approach does not satisfy the requirements on co-channel interference, the number of available codes can be extended to 512 and all primary scrambling codes can be assigned.

7.1 Scrambling Code Planning Example - Workaround In analogy to GSM frequency planning, downlink scrambling codes can be assigned comparable to the way frequencies have been assigned in GSM. In Tornado DMS3.1, the first step is to define as many frequencies as downlink scrambling codes are available for reuse and to assign them to the carrier database. In our example 64 carriers were defined (by editing the project-specific file ‘carrier_types.names’) and afterwards assigned as BCCH frequencies to the carrier type database.

figure 18: carrier type database

s


Page 33 of 43


Next step is the assignment of all available carriers (or scrambling codes) to one carrier group with a 64 reuse pattern in the Carrier Group Database.

figure 19: carrier group database

s


Page 34 of 43


In the global settings of the Carrier Database, the number of required carriers has now to be set equal to one for each cell (equivalent to the assignment of one primary scrambling code per cell) and the change has to be applied.

figure 20: carrier database

Info: Before the next steps (automatic frequency planning, interference table, etc.) are carried out, power settings have to be adjusted according to the W-CDMA settings and a best server array has to be created in the GSM module.

s


Page 35 of 43


Now, after creation of interference table and neighbor list, the Automatic Frequency Planner can be started in the Frequency Planning Tool and the frequency (scrambling code) plan can be generated and applied.

figure 21: frequency planning tool

figure 22: automatic frequency planning

s


Page 36 of 43


The result of the exemplary assignment and the network plot with associated scrambling code indices is shown in the following.

figure 23: output of automatic frequency planning tool

s


Page 37 of 43


figure 24: area plot of sites with assigned DL scrambling code indices

s


Page 38 of 43


8 Conclusions From the Network Planning point of view, scrambling code planning within one operator network seems not to be as crucial as e.g. frequency planning in GSM. The total number of available scrambling codes in the downlink direction seems to be sufficient to solve any problems with regard to co-channel interference. However, if the code assignment in the network matches the code acquisition strategy implemented in the mobile, benefits in terms of increased cell-search performance are likely to be expected. An additional topic is scrambling code planning at neighboring country borders. If certain levels of interference are exceeded, actions should be taken into consideration in discussion with at least one of the concerned operators.

s


Page 39 of 43


8 Annex 1 – Release specific restrictions In addition to UE compatibility, HW limitations inside the UTRAN may have influence on applicable code planning strategies. For UMR2.0 currently up to 16 cells can be assigned to each primary scrambling code. For the application of e.g. reuse of 64 downlink scrambling codes, this would lead to a maximum of 1024 supported cells, for more supported cells a higher reuse than 64 should be selected.

s


Page 40 of 43


9 Annex 2 – Revision History Version Date Revised by Changes

0.1 18.03.2002 First initial draft

0.5 31.07.2002 Draft version released for review

1.0 29.10.2002 Inclusion of review comments; additional explanations, examples and illustrations; new chapter on graph coloring based approach

1.1 15.07.2003 Inclusion of a description for uplink scrambling codes (chapter 3); additional illustration and explanation for downlink codes (chapter 2); minor editorial changes

10 Annex 3 – References

Type of Documentation

Name Release Status

[1] 3GPP TS 25.213 Spreading and Modulation (FDD) 3.8.0

[2] 3GPP TS 25.214 Physical Layer Procedures (FDD) 3.10.0

[3] 3GPP TS 25.211 Physical channels and mapping of transport channels onto physical channels (FDD)

3.11.0

[4] Young-Ho Jung, Yong H. Lee

Scrambling Code Planning for 3GPP W-CDMA Systems, IEEE VTC 2001, May 2001

[5] European Radiocommunications Committee (ERC)

Recommendation (01)01: Border Coordination of UMTS/IMT-2000 Systems

2001

s


Page 41 of 43


11 Annex 4 – Abbreviations 3GPP 3rd Generation Partnership Project

AICH Acquisition Indication Channel BCH Broadcast Channel

CCPCH Common Control Physical Channel

CCTrCH Coded Composite Transport Channel CDMA Code Division Multiple Access

CEPT European Conference of Postal and Telecommunications Administrations

CPCH Common Packet Channel

CPICH Common Pilot Channel DPCCH Dedicated Physical Control Channel

DPDCH Dedicated Physical Data Channel

DS-CDMA Direct Sequence CDMA DSMA-CD Digital Sense Multiple Access with Collision Detection

DSCH Downlink Shared Channel

ERC European Radiocommunications Committee FBI Feedback Information

FDD Frequency Division Duplex

FDMA Frequency Division Multiple Access GSM Global System for Mobile Communication

IMT-2000 International Mobile Telecommunication System 2000

OVSF Orthogonal Variable Spreading Factor P-CCPCH Primary Common Control Physical Channel

PCH Paging Channel

PDSCH Physical Downlink Shared Channel PICH Paging Indication Channel

PRACH Physical Random Access Channel

RACH Random Access Channel SCH Synchronization Channel

S-CCPCH Secondary Common Control Physical Channel

SF Spreading Factor TDD Time Division Duplex

TFCI Transport-Format Combination Indication

s


Page 42 of 43


TPC Transmit-Power Control UE User Equipment

UMTS Universal Mobile Telecommunications System

UTRA UMTS Terrestrial Radio Access UTRAN UMTS Terrestrial Radio Access Network

W-CDMA Wideband Code Division Multiple Access

s


Page 43 of 43


12 Annex 5 – List of figures figure 1: schematic view of spreading and scrambling .................................................. 5 figure 2: exemplary scrambling operation...................................................................... 6 figure 3: primary and secondary scrambling code grouping.......................................... 8 figure 4: primary scrambling code grouping scheme..................................................... 9 figure 5: structure of the random-access transmission................................................ 11 figure 6: structure of the CPCH access transmission .................................................. 11 figure 7: correspondence between downlink primary scrambling code and uplink

PRACH preamble scrambling code group ................................................... 12 figure 8: structure of the synchronization channel (SCH) ............................................ 15 figure 9: cell grouping for code reuse clustering with a 7 cell cluster .......................... 18 figure 10: exemplary cluster reuse of 19 code groups................................................. 20 figure 11: reduced number of interferers for sectorization........................................... 21 figure 12: sub-optimal scrambling code group allocation ............................................ 23 figure 13: optimal scrambling code group allocation ................................................... 23 figure 14: number of neighboring cells with regard to cell layout................................. 25 figure 15: exemplary assignment of preferential and neutral frequencies ................... 28 figure 16: illustration of neutral and preferential frequency scenarios ......................... 29 figure 17: classification of country types in Europe...................................................... 31 figure 18: carrier type database ................................................................................... 32 figure 19: carrier group database................................................................................. 33 figure 20: carrier database ........................................................................................... 34 figure 21: frequency planning tool................................................................................ 35 figure 22: automatic frequency planning...................................................................... 35 figure 23: output of automatic frequency planning tool ................................................ 36 figure 24: area plot of sites with assigned DL scrambling code indices ...................... 37

13 Annex 6 – List of tables table 1: codes at neighboring countries, conventions of writing .................................. 30 table 2: code sharing between neighboring countries ................................................. 30

Documents

Code Planning Manual