WO1999003289A1

WO1999003289A1 - A method for flexible capacity dynamic frequency allocation in cellular radio networks

Info

Publication number: WO1999003289A1
Application number: PCT/EP1997/003580
Authority: WO
Inventors: Harri Posti; Petri Jolma; Olli Piirainen; Juha SÄRKIOJA
Original assignee: Nokia Networks Oy
Priority date: 1997-07-07
Filing date: 1997-07-07
Publication date: 1999-01-21
Also published as: AU4009597A

Abstract

The present invention provides a method for flexible capacity dynamic frequency allocation in a cellular radio network, wherein a portion (f1, ..., fm) of total available frequencies (f1, ..., fn) is allocated to the cells of the cellular network in a fixed manner, and remaining frequencies (fm+1, ..., fn) form a common frequency pool from which frequencies are temporarily allocated to a respective cell based on traffic consideration. Moreover, the present invention provides a method for flexible capacity dynamic frequency allocation in a cellular radio network, wherein all allocatable frequencies are divided into frequency groups (f1, ..., f6), each of which frequency groups containing at least one frequency, which frequency groups are assigned to cells of the cellular radio network such that at least two groups are assigned to each single cell, and at least one frequency group (f6) is moved from cells neighboring a respective cell to be temporarily allocated to said respective cell based on traffic consideration.

Description

A method for flexible capacity dynamic frequency allocation in cellular radio networks

FIELD OF THE INVENTION

The present invention relates to a method for flexible capacity dynamic frequency allocation in cellular radio networks, as for example in a cellular radio network according to the GSM standard, operating according to TDMA.

BACKGROUND OF THE INVENTION

In current cellular radio networks, available frequencies within the overall available bandwidth are distributed among cells. That is, frequencies are allocated to respective cells in a fixed manner, which implies a frequency update interval in the order of months. Thus, in terms of available frequencies (channels), each cell must be dimensioned according to the maximum amount of traffic (e.g. number of calls per unit time) expected in the cell. However, during most of the time the amount of traffic is apparently considerably less than the maximum expected traffic amount the cell is rated for.

Hence, as a result, unused capacity is present in some parts (cells) of the network, while it would be needed elsewhere in the network. Consequently, spectrum usage of the available overall bandwidth is not optimal and only less than optimal (or maximum possible) capacity of the network is rendered available for radio communication purposes .

Moreover, as the penetration of mobile phones used in a cellular radio network steadily increases since several years, the usable radio bandwidth is going to be one limiting factor with regard to increasing the capacity of a mobile (cellular radio) network, at least in city areas in which a high traffic amount has to be expected.

In the near future, software radio systems, i.e. software based or controlled radio systems become feasible for cellular radio networks such as GSM. This renders the price of a single traffic channel (radio channel) TRX much cheaper when compared to the systems used today and it also simplifies to control the amount and location of radio channels used in one base station BS . Therefore, for an operator it becomes quit cheap to provide a lot of capacity in a base transceiver station BTS .

However, it can not be well used with a cellular radio network with fixed frequency allocations in a base station BS, because the total amount of usable radio channels by an operator will determine the maximum capacity of the network.

Hitherto, several approaches have been conceived in order to increase the capacity of a cellular radio network or to effectively utilize the available capacity.

A first approach is based on the fact that a base station BS can handle a fixed number of (traffic) radio channels TRX which also determine the capacity of the base station. These radio channels have been divided between different base stations, thus leading to a reuse of certain frequencies and a cellular arrangement with a so-called frequency reuse pattern. Stated in other words, cellular radio systems have conventionally been designed such that the same frequency or channel, respectively, is used in cells located at a sufficient distance from each other, whereby interfering signals resulting from propagation attenuation of the signals remain within acceptable limits. This, as described for example in same applicant's (and same inventor's) former patent application number PCT/FI95/00653 (international publication number WO

96/17485) leads to the above mentioned frequency reuse pattern in the cellular structure. For example, a reuse pattern of n means that 1/n-th of all available frequencies (frequency bands) is allocated to every cell and the same frequencies are reused after a relative reuse distance from each other.

Then, according to such an approach, system capacity has conventionally been increased, e.g. by shortening the reuse distance, while specific measures for preventing a decrease of transmission quality (for example, reduction of transmission power level in order to reduce co-channel interference as an interference phenomenon due to signals transmitted in near-by cells on the same frequency) have to be provided for.

However, this approach as briefly outlined above is limited by co-channel interference requirements when shortening the reuse distance.

A second conventional approach is based on so-called soft capacity which is applicable in cellular radio networks offering a possibility for frequency hopping, as for example in GSM networks. That is, in frequency hopping networks, it is possible to use some quality/traffic measure to decide on a per-call-basis whether to allocate a channel or not. This will effectively provide soft capacity based on the actual traffic amount in different parts of the network.

However, such a soft capacity functionality is, according to the understanding of the present inventors, currently not supported by any manufacturers base station system BSS or network planning tool under GSM, while other cellular radio network standards do not provide the possibility of frequency hopping.

In particular, such a soft capacity feature is difficult to implement in non-hopping networks because of the vastly different interference levels on individual channels.

Moreover, such an approach would also require radio network synchronization in order to perform in an optimal way.

In addition, a third approach known as the so-called hard capacity approach is conceivable. According to this approach, capacity has traditionally been increased by mere addition of more channel units, i.e. more channels or available frequencies.

However, due to the limitations of available bandwidth, the potentially added channel units or frequencies are limited. Moreover, this method is not optimal in the sense of efficient capacity use, since a part of the available capacity would be mostly unused.

SUMMARY OF THE INVENTION

It is therefore an object of the present invention to provide a method for flexible capacity dynamic frequency allocation in cellular radio networks which is free of the above drawbacks associated with conventional approaches and which allows changes in local capacity, to thereby optimize the efficient use of the available frequency spectrum or bandwidth, respectively.

According to a first aspect of the present invention, this object is achieved by a method for flexible capacity dynamic frequency allocation in a cellular radio network wherein a portion of total available frequencies is allocated to the cells of the cellular network in a fixed manner, and remaining frequencies form a common frequency pool from which frequencies are temporarily allocated to a respective cell based on traffic considerations.

Alternatively, according to a second aspect of the present invention, this object is achieved by a method for flexible capacity dynamic frequency allocation in a cellular radio network, wherein all allocatable frequencies are divided into frequency groups, each of which frequency groups containing at least one frequency, which frequency groups are assigned to cells of the cellular radio network such that at least two groups are assigned to each single cell, and at least one frequency group is moved from cells neighboring a respective cell to be temporarily allocated to said respective cell based on traffic considerations.

Stated in other words, according to the first and second aspects of the present invention, due to flexibly allocatable frequencies, there exists the advantage that changes in local capacity within the cellular radio network are allowable and an optimized usage of the available frequency spectrum is possible.

Further, according to the first aspect of the present invention, an arbitrary frequency not currently allocated to any of the interfering cells may be assigned from a common frequency pool to a cell requiring additional capacity .

Moreover, according to the second aspect of the present invention, while still providing an optimized usage of the available frequency spectrum, further advantageously all of the usable frequency band can be used at the same time for highest possible capacity in network, with no reserve banks or frequencies have to be reserved for heavy load situations, i.e. a high amount of traffic, due to arranging the frequencies into movable groups to be moved ("loaned") to a respective neighboring cell.

BRIEF DESCRIPTION OF THE DRAWINGS

In the following, the present invention will be described in greater detail with reference to the accompanying drawings, in which:

Fig. 1 shows a network architecture of a network based on a hexagonal grid used as an example for explanatory purposes;

Fig. 2 shows an example of interfering cells within a network as shown in Fig. 1, the cells being arranged with a reuse distance D;

Fig. 3 shows an example of a standard reuse pattern with reuse factor k of four;

Fig. 4 schematically depicts, according to the first aspect of the invention, the frequency partition of the total allocatable frequencies; Fig. 5 is a table for illustrating allocation of frequencies from the common frequency pool to a specified target cell according to the first aspect of the invention;

Fig. 6 is a diagram illustrating the number of allocated channels as a function of time according to the first aspect of the invention; and

Fig. 7 illustrates the method for dynamic frequency allocation according to the second aspect of the present invention within an example cell structure adopting a frequency reuse factor k of three.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

Subsequently, examples and embodiments of the present invention will be described in more detail with reference to the accompanying drawings .

Shown in Fig. 1 is a network architecture of a cellular radio network based on a hexagonal grid used as an example for explanatory purposes. That is, for the following explanations it is assumed that the cellular radio network is divided in cells of hexagonal shape, each cell representing the area in which, for example according to GSM standard, a base station BS (not shown) establishes communication links between said base station and numerous mobile stations MS (not shown) present within the cell. However, the invention applies to cells of any shape and size. Hence, it is also relevant for so-called multi-layer cellular radio networks.

With regard to Fig. 2, an example of interfering cells (cells numbers 1 to 7) within a network as shown in Fig. 1 is depicted. As previously stated above, the cells of the network are arranged at a reuse distance D at which the frequencies allocated to respective cells are allowed to be reused. The reuse distance D, in turn is dependent on the required ratio C/I. Fig. 2 shows an example of interfering cells with a reuse factor k of four, which means a reuse distance D=3.46*r (r being the radius of the cell and/or the range of the base station) . With a path loss exponent of 3.5 this is equivalent to a C/I requirement of 14dB (10*log (2.46^Λ3.5) ) . That is, a reuse factor k of four means that only 1/4 of available frequencies is allocated to a respective cell. Now, assuming that only four frequencies are available, a single respective frequency only is allocated to each of the cells, the respective frequency being reused or again allocated at (or after) the reuse distance D. Practically spoken, assuming that frequencies 1 to 4 are to be allocated in the cellular network depicted in Fig. 2, the frequency allocated to cell number 1 in Fig. 2 is, for example, reused in the cell indicated by D, while frequencies 2 to 4 are, for example, allocated to cells numbers (2 + 5), (3 + 6), and (4 + 7), respectively, with the reuse distance being maintained for each frequency.

Fig. 3 of the accompanying drawings exemplifies a cellular radio network in which frequency allocation to the individual cells may be effected according to the method according to the first aspect of the present invention.

That is, a portion of the available frequencies is allocated to the cells in a fixed manner, thus establishing a standard frequency reuse pattern as previously described. As a minimum, the pilot or control channel frequency (the BCCH channel according to GSM) in each cell is permanently allocated. Fig. 3 depicts a situation with a reuse factor k of four. The rest of the available frequencies which are not permanently allocated form a so-called common frequency pool. From this common frequency pool, frequencies are temporarily allocated to cells based on traffic considerations, i.e. depending on the amount of traffic within a respective one of the cells.

Fig. 4 illustrates the frequency partition of the total allocatable frequencies n into the common frequency pool n-m and those frequencies with fixed allocation m, according to the first aspect of the invention.

Fig. 5 is a table for illustrating allocation of frequencies from the common frequency pool to a specified target cell according to the first aspect of the invention. This table is based on the assumption that the number of frequencies for total frequency allocation is fourteen, and that between the cells a reuse pattern of four is established. That is, four frequencies (frequencies 1 to 4) are fixedly allocated as a respective control channel as explained herein above with reference to Fig. 3, while the remaining frequencies (frequencies 5 to 14) form the common frequency pool. However, the above number of frequencies serves as an example only. Generally, the number of total available frequencies is assumed to be n (fl, ..., fn) , with a portion of m frequencies (fl, ..., fm) thereof being allocated to the cells of the cellular network in a fixed manner, while the remaining n-m frequencies (fm+1, ..., fn) form the common frequency pool.

Then, in order to dynamically allocate frequencies to respective cells in order to provide for flexible capacity, according to the first aspect of the present invention, an interfering cell list is defined for each cell. This happens as a part of the manual or automatic frequency planning process. In principle, this list includes those cells that are closer than the allowed reuse distance.

In the given example, when referring to Fig. 2 (and/or 3), an interfering cell list defined for cell number 1 contains the list of the potentially interfering (six) neighbors, i.e. cells numbers 2 to 7.

The list is contained in the base station controller BSC, the switch or other network element responsible for radio resource management RRM. Moreover, in case of a multi-layer network, the list will contain cells of all layers.

The table shown in Fig. 5 of the accompanying drawings is an arbitrarily chosen example of such an interfering cell list based on the previously made assumptions for the chosen example: fourteen different frequencies (1 to 14) totally allocatable, reuse factor of four with frequencies 1 to 4 fixedly allocated within the reuse pattern, and frequencies 5 to 14 thus forming the common frequency pool.

Then the entries in the cell 1 interfering cell list relate to those frequencies of the common frequency pool currently used by cells 2 to 7 neighboring cell 1 and those frequencies of the common frequency pool currently still available to be allocated to cell 1.

If reuse splitting like IOU or a similar one is used, there is an independent interfering cell list for each reuse factor in use (regular, super, ... ) . When temporarily allocating a frequency to a cell, it must be decided which reuse factor the frequency will use and select it according to the respective interfering cell list. The actual dynamic frequency allocation from the common pool, based on such an interfering cell list, is performed as described in the following with reference to Fig. 6 of the drawings .

Within the procedure, thresholds for either increasing or decreasing the number of channels (traffic channels) allocated to a specific cell are defined, dependent on the current traffic amount at a given time. That is, a threshold "increase channels" and a threshold "decrease channels" is defined.

When the amount of traffic in the target cell exceeds the "increase channels" threshold, the network element for radio resource management RRM selects from the common frequency pool a frequency which is not currently allocated to any of the interfering cells and assigns this frequency to the target cell. Thereby, the number of channels in the target cell is increased by one, thus providing increased transmission capacity within the cell.

It is also possible to prevent allocation of a channel adjacent to those in use in the interfering cell. This will advantageously prevent excessive adjacent channel interference (co-channel interference) .

If no such frequency is available, it may be possible to arrange one by intra-cell handover and re-allocation of common pool frequencies.

Further, if no such frequency can readily be assigned to the target cell requiring additional capacity due to an increased amount of traffic within the cell, the cell remains in a queue for an additional frequency. While in the queue, there is a possibility of hard blocking in that cell.

The maximum number of frequencies possible in a cell is limited by the hardware HW installed in the respective base station BS .

Then, when the amount of traffic in the target cell (cell of interest) falls below the "decrease channels" threshold, one of the common pool frequencies assigned to the cell is de-assigned or released, respectively, by the network element handling the radio resource management RRM, and returned to the common frequency pool, thus being available to be assigned to another cell upon need. In particular, the element handling RRM keeps a table of interfering cells for each target cell and frequencies from the common frequency pool currently allocated to those cells.

If this decrease with regard to the traffic amount happens while the respective cell is in the queue, the cell is removed from the queue. Furthermore, it is also possible that the "decrease channel" threshold is set to a different value when a cell is in the queue.

Moreover, the "increase channel" threshold may be different from the "decrease channel" threshold. Preferably, these different thresholds, i.e. the "increase channel" threshold and the "decrease channel" threshold are chosen such that hysteresis occurs. Such hysteresis will advantageously prevent continuous allocation-release of channels.

An example for the increase and decrease of the capacity (number of channels) available in a target cell as function of traffic amount varying as a function of time together with respective thresholds is shown in Fig. 6 of the drawings, which illustrates the method according to the first aspect of the present invention.

Thus, according the flexibly allocatable frequencies according to the above described method according to the first aspect of the invention, changes m local capacity of the cellular radio network are allowed, thus leading to an optimized usage of the available frequency spectrum.

In a cellular radio network operating on the basis of a

TDMA system, the above described process can take place on a per timeslot basis, assuming that the network is synchronized. In such a case, if N timeslots are provided per frame, then N tables similar to the one depicted m and explained with reference to Fig. 5 are provided for.

In the following, the method for flexible capacity dynamic frequency allocation according to the second aspect of the present invention is described. To this end, in the subsequent explanations, reference is made to Fig. 7 which depicts, also as an example only, a cellular radio network divided m hexagonal cells . Between the cells illustrated in the given example, a frequency reuse factor k of three is established.

According to the second aspect of the invention, generally spoken, allocatable frequencies grouped to frequency groups are not allocated to cells in a fixed static manner, but instead, a semi-dynamic allocation scheme, based on traffic considerations, i.e. based on load, namely based on the currently occurring amount of traffic, is proposed.

To be precise, as shown in Fig. 7 of the accompanying drawings, a cellular raαio network is shown m which six different frequency groups fl, f2, ..., f6 are assumed to be available for radio communication. Each of these frequency groups fl to f6 consists of a plurality of different individual frequencies, and consists of at least one frequency.

However, due to co-channel interference, the same radio channel (frequency or frequency group) can not be allocated to neighboring cells.

Generally, radio channels can be divided into different groups, each of these groups containing several radio channels with the same radio frequency being located only in one group. There can be any number of these groups and each of them can contain any amount of radio channels. One cell (a base station BS) may have one or more of these groups allocated to it. This means that any network can be described using this approach and this method does not restrict the frequency planning.

According to the specific example as shown in Fig. 7 for the purpose of explanations only, the available individual frequencies are divided into several frequency groups fl to fβ each containing at least one frequency. These groups are assigned to the cells of the cellular network, so that at least two frequency groups, i.e. frequency groups fl&f2, f3&f4, f5&f6, respectively, are allocated to each respective single cell. As may be gathered therefrom, each cell uses only 1/3 of the overall available number of frequency groups, thus leading to a reuse factor k of 3 (together with a corresponding reuse pattern and/or reuse distance) .

A base station controller BSC controlling the plurality of individual base stations BS (not shown) forming the cellular radio network has a knowledge of a load (traffic amount) of each base station it controls. The base station controller BSC has also a knowledge of allocated radio channels for each base station BS and for its neighboring cells. Therefore, a base station controller BSC must also know the neighboring cells of each cell (as is the case according to GSM standard, where a base station controller monitors at least the six neighboring cells of a specific cell of interest or target cell, respectively) .

As stated before, these frequencies (channels) are organized to frequency groups. Hence, the base station controller BSC knows the respective frequency groups allocated to each cell as well as those allocated to its neighboring cells.

Then, when the base station controller BSC notices a heavy load in one cell, it can move (or loan, respectively) a radio frequency or frequency group, respectively, from its neighboring cells in order to increase the capacity of the cell (base station BS or base transceiver station BTS) under "heavy load" conditions, i.e. the cell in which a large amount of traffic currently occurs. This added (or moved) frequency group must not be present in neighboring cells or it must be removed from there, because of interference problems.

Depicted in Fig. 7 is a situation in which the frequency group fβ is moved from three cells adjacent to the cell under "heavy load" conditions to the cell under "heavy load" conditions. In this cell, frequency groups fl&f2&fβ are thus present during the period of heavy load or large amount of radio communication traffic.

Consequently, with the described method according to the second aspect of the present invention, the capacity of the "heavy load" cell is increased in expense of the cells where the loaned frequency groups are removed.

After the "heavy load" situation is over, the frequency groups can be returned to the original state in order to keep the frequency planning in control.

The load conditions and/or traffic amount occurring in a target cell may be judged as a "heavy load" situation depending on an appropriately set first threshold value. Similarly, a condition of ""heavy load" situation being over" may be judged depending on an appropriately set second threshold value. As already described further herein above, such threshold values may differ from each other to provide for hysteresis. Thus, also in this case it may be preferable to provide for hysteresis, thus preventing continuous moving of frequency groups .

Accordingly, with the above described method according to the second aspect of the present invention, the possible capacity of a cellular radio network like a GSM network can be increased. Moreover, all of the usable frequency band can be used at the same time for highest possible capacity in the network. Hence, no "reserve" banks have to be reserved for situations of heavy traffic load.

In addition, the base station controller BSC software as a means for radio resource management RRM must know the neighboring cells of each cell and the radio frequencies allocated thereto. Organizing the frequencies to movable frequency groups makes it easier for the base station controller BSC (software) to handle the frequencies compared to a situation with single radio frequency handling. Furthermore, a group can still contain only one radio channel, such that this does not restrict the frequency planning.

Moreover, the network frequency allocation can advantageously be designed with a higher reuse factor. This reduces the amount of frequencies usable by one base transceiver station BTS, but improves co-channel interference situations within the cellular radio network. The disadvantage of lower capacity in each cell can be handled by moving groups from neighboring cells to a cell under "heavy traffic load" conditions, without degrading the co-channel interference properties of a network.

A still further advantage resides in the fact that with a higher reuse factor the moved (or loaned) frequency group has to be removed from less neighboring cells than in low reuse factor networks, while there is also a lot more frequency groups to be loaned. This renders the cellular radio networks configured with a high reuse factor more usable and leads to less co-channel interference in network, hence, to better transmission quality.

So, in conclusion, the method according to the second aspect of the present invention can be advantageously employed in order to improve the capacity of a cellular radio network and also it can be used to improve the transmission quality within the network.

Furthermore, it is to be noted that both of the methods as described herein above with regard to the first and second aspects of the present invention may also be combined, if desired. That is, a radio frequency allocation method for use in a cellular radio network is conceivable, according to which radio frequencies organized in movable groups, according to the second aspect of the invention, may be supplemented by a common frequency pool of further available radio frequencies, according to the first aspect of the invention.

Such a synthesis of the individual methods according to the first and second aspects of the present invention would, in general, show the same advantages as the individual methods .

It should, however, be understood that the above description and accompanying figures are only intended to illustrate the present invention by way of example only. Thus, the method according to the invention may also be used in systems other than the described. The preferred embodiments of the method may thus vary within the scope of the attached claims.

Claims

1. A method for flexible capacity dynamic frequency allocation m a cellular radio network, wherein a portion (fl, ..., fm) of total available frequencies (fl, ..., fn) is allocated to the cells of the cellular network in a fixed manner, and remaining frequencies (fm+1, ..., fn) form a common frequency pool from which frequencies are temporarily allocated to a respective cell based on traffic considerations.

2. A method according to claim 1, wherein said portion (fl, ..., fm) of total available frequencies (fl, ..., fn) allocated to the cells of the cellular network in a fixed manner is the control channel frequency in each cell .

3. A method according to claim 1, wherein said portion (fl, ..., fm) of total available frequencies (fl, ..., fn) is allocated to the cells of the cellular network in a fixed manner to thereby build a standard frequency reuse pattern.

4. A method according to claim 3, wherein said reuse pattern has a reuse factor of k.

5. A method according to claim 1, wherein a frequency from the common frequency pool is additionally allocated to a respective cell when traffic m said cell exceeds a first predetermined threshold value.

6. A method according to claim 5, wherein a frequency additionally allocated to said respective cell is de-assigned from the cell and returned to the common frequency pool, when traffic m said cell falls below a second predetermined threshold value.

7. A method according to claim 6, wherein said first and said second predetermined threshold values are different from each other such that hysteresis is provided.

8. A method according to claim 5, wherein a cell remains in a queue for an additional frequency if no frequency can be readily assigned from the common frequency pool.

9. A method according to claim 5, wherein a frequency additionally allocated to said respective cell is a frequency currently not allocated to any of interfering cells.

10. A method according to claim 1, wherein said temporary allocation of frequencies from the common frequency pool is performed under control of a network element responsible for radio resource management which keeps a table of interfering cells for each target cell and frequencies from the common frequency pool currently allocated to those cells.

11. A method according to any of the preceding claims, wherein allocating of frequencies in a cellular radio network operating on the basis of a TDMA system is performed on per timeslot basis in a synchronized TDMA network .

12. A method for flexible capacity dynamic frequency allocation in a cellular radio network, wherein all allocatable frequencies are divided into frequency groups (fl, -••/ 6) , each of which frequency groups containing at least one frequency, which frequency groups are assigned to cells of the cellular radio network such that at least two groups are assigned to each single cell, and at least one frequency group (f6) is moved from cells neighboring a respective cell to be temporarily allocated to said respective cell based on traffic considerations.

13. A method according to claim 12, wherein the at least one moved frequency group (f6) is removed from said neighboring cells from which it is moved.

14. A method according to claim 12, wherein said at least one frequency group (fβ) is moved to said respective cell when traffic in said respective cell exceeds a first threshold value and the presence of increased traffic in said respective cell is determined.

15. A method according to claim 14, wherein the original state of allocated frequencies is reestablished when traffic in said respective cell falls below a second threshold value and the absence of increased traffic in said respective cell is determined.

16. A method according to claim 15, wherein said first and said second predetermined threshold values are different from each other such that hysteresis is provided.

17. A method according to claim 12, wherein the frequencies are moved under control of a network element responsible for radio resource management.

18. A method according to claim 12, wherein said groups of frequencies are allocated to cells of the cellular radio network to thereby build a standard frequency reuse pattern.