Method for detecting and reducing ping-pong handover effect of cellular network

Abstract

A method for detecting and reducing ping-pong handover effect of a cellular network is disclosed. The method provides a systematic way to determine which cell pairs of the network have high probability of suffering from the ping-pong handover effect. The method comprises collecting network traffic statistic data of multiple items, wherein each of the items corresponds to a state change of multiple mobile units within cells of the network; determining whether a plurality of criteria are met according to the network traffic statistic data; and detecting the ping-pong effect between the cells according to the result of the determining step. The method also comprises changing a handover parameter for the cells with the ping-pong handover effect.

Description

BACKGROUND OF INVENTION

1. Field of the Invention

The present invention relates in general to mobility management in a wireless communication network, and more particularly to a method for detecting and reducing ping-pong handover effect of a cellular network.

2. Description of the Prior Art

In a cellular network (e.g. the GSM system), the ping-pong handover is a very common phenomenon to degrade the network performance. The ping-pong handover means handover to and fro between a cell pair frequently. The ping-pong effect occurs due to the frequent movement of mobile units between the cell pair, or high signal fluctuation at the common boundary of the cell pair. Since the ping-pong handover increases the times of handover and thus the loading of the network, it is necessary for network operators to reduce this undesirable effect. However, the conventional technology does not provide a systematic and objective solution for the operators to find the cell pair suffering from the ping-pong effect, so it depends on the subjective experience of the operators or the try-and-error way to handle this problem. Usually, this approach would cause inefficiency and failure.

SUMMARY OF INVENTION

It is therefore an object of the present invention to provide a method for detecting ping-pong handover effect of a cellular network, thereby detecting the ping-pong handover effect in a systematic and efficient way.

Another object of the present invention is to provide a method for reducing ping-pong handover effect of a cellular network, thereby reducing the ping-pong handover effect in a systematic and efficient way

According to the objects of the present invention, the method for detecting the ping-pong handover effect of the cellular network comprises collecting network traffic statistic data of a plurality of items, wherein each of the items corresponds to a state change of a plurality of mobile units within cells of the cellular network; determining whether a plurality of criteria are met according to the network traffic statistic data; and detecting the ping-pong handover effect between the cells according to the result of the determining step.

In another aspect, the method for reducing the ping-pong handover effect of the cellular network comprises collecting network traffic statistic data of a plurality of items, wherein each of the items corresponds to a state change of a plurality of mobile units within cells of the cellular network; determining whether a plurality of criteria are met according to the network traffic statistic data; detecting the ping-pong handover effect between the cells according to the result of the determining step; and changing a handover parameter for the cells with the ping-pong handover effect.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a flow chart of a preferred embodiment of the method for detecting ping-pong handover effect within a cellular network according to the present invention.

FIG. 2 is a flow chart of a preferred embodiment of the method for reducing the ping-pong handover effect within the cellular network according to the present invention.

DETAILED DESCRIPTION

In all embodiments of the present invention, the cellular network can refer to any communication system with a cellular structure, such as GSM, GPRS, UMTS, CDMA, CDMA2000, WiMax, etc. Besides, the concept of handover may be described with different expressions for different types of cellular networks. For examples, handoff is used in CDMA and CDMA2000, while cell update is used in GPRS and UMTS.

FIG. 1 is a flow chart of a preferred embodiment of the method for detecting ping-pong handover effect within a cellular network according to the present invention. The cells of the cellular network are grouped into a plurality of location areas. As shown in FIG. 1, the flow comprises the steps of:

• • 11 collecting network traffic statistic data of a plurality of items;
  • 12 determining whether a set of criteria is met according to the network traffic statistic data; and
  • 13 detecting the ping-pong handover effect between the cells according to the result of the step 12.

In the step 11, each of the items corresponds to a state change of a plurality of mobile units (e.g. cellular phones) within the network. The state change means a certain change of the connection between the mobile unit and the network. Furthermore, the network traffic statistical data of the items are generated based on a certain period of time determined by the network equipments. In the step 12, the set of criteria provides a systematic way to determine which cell pairs have high probability of suffering from the ping-pong handover effect. Each of the criteria may involve a comparison between a related parameter, whose value is determined by the network traffic statistical data collected in the step 11, and a corresponding threshold. Then, according to whether the set of criteria is met, the ping-pong effect can be detected efficiently and objectively in the step 13.

Based on the flow of FIG. 1, three more detailed embodiments are described below.

The First Embodiment

In the first embodiment, the step 11 collects the below items: (1) Directional handover between each cell pair. (2) Location area update of each cell. (3) Mobile terminating call (MTC) and mobile originating call (MOC) within each cell. Directional handover means the times of the connection between the network and the mobile unit changing from via a source cell to via a neighboring destination cell. Location area update (LAU) of a cell means the times of the mobile units moving to the cell from its neighboring cells belonging to other different location areas. In a network unable to collect the item of LAU, a more general but less accurate item of location update (LU) is collected. MTC represents the times of the mobile units within a cell receiving calls, and MOC represents the times of the mobile units calling out. There is usually a proportional relation between the MTC and MOC of the cell.

In the step 12, the set of criteria as described below can be applied to any cell pair (e.g. CELL_i and CELL_j) of the network:

• • (1-1) The directional handover from CELL_i to CELL_j (denoted by HO_ij) is higher than a threshold TH₂. This criterion means that a cell pair with a relatively low directional handover is not taken into consideration.
  • (1-2) A parameter P1_ij is higher than a threshold TH₁. Here P1_ij is defined by a ratio between HO_ij and the sum of MTC and MOC within CELL_i (denoted by MTC_i and MOC_i respectively). That is, P1_ij=HO_ij/(MTC_i+MOC_i)  Eq. (1)

Another workable version of P1_ij is defined by a ratio between HO_ij and the sum of MTC and MOC within CELL_j. This varied version can also work since it brings a result similar to Eq. (1). That is, P1_ij=HO_ij/(MTC_j+MOC_j)

A high P1_ij represents that a significantly higher directional handover is generated based on the same quantity of the mobile units on the phone, that is, the ping-pong handover effect is probable to exist between CELL_i and CELL_j.

• • (1-3) A parameter P2_ij is higher than a threshold TH₃. Here P2_ij is defined by a ratio of the times of signal movement from CELL_i to CELL_j in active mode to those in idle mode. The signal movement from CELL_i to CELL_j means that a mobile unit moves from CELL_i; to CELL_j, or the medium via which the mobile unit connects with the network is changed from CELL_i to CELL_j. The active mode means that the mobile unit is accessing the network for voice or data service, and the idle mode means that the mobile unit is not in the active mode while still connecting to the network.

In this criterion, we can interpret the signal movement from CELL_i to CELL_j in active mode as HO_ij. Thus, a high P2_ij represents that a significantly larger portion of the moving mobile units from CELL_i to CELL_j is in the active mode (e.g. on the phone), that is, the ping-pong handover effect is probable to exist between CELL_i and CELL_j.

If CELL_j is a border cell of a location area, then P2_ij is equivalently determined by a ratio of the sum of all HO_kj to LAU of CELL_j (denoted by LAU_j), that is, $P{2}_{\mathrm{ij}}=\sum _{k}H{O}_{kj}/LA{U}_j$

• • wherein each CELL_k is a neighboring cell of CELL_j and belongs to a location area different from that of CELL_j.

If CELL_j is an inner cell (i.e. not a border cell), then P2_ij is equivalently determined by all P2_kj, where CELL_k is a neighboring cell of CELL_j. With regard to the way of determination, please refer to the part of A_i calculation in the U.S. patent application “SYSTEM FOR CONSTRUCTING A MOBILITY MODEL FOR USE IN MOBILITY MANAGEMENT IN A WIRELESS COMMUNICATION SYSTEM AND METHOD THEREOF” (filed on Feb. 21, 2003 with Ser. No. 10/369,678), which is incorporated herein by reference.

• • (1-4) If the pair of CELL_i and CELL_j has the ping-pong handover effect, HO_ij, P1_ij, and P2_ij used in the criteria (1-1)-(1-3) should be symmetric to HO_ji, P1_ji, and P2_ji respectively. HO_ji is the directional handover from CELL_j to CELL_i. P1_ji is a reverse parameter of P_ij and defined by the ratio of HO_ji to the sum of MTC_j and MOC_j. P2_ji is a reverse parameter of P2_ij and defined by the ratio of the times of signal movement from CELL_j to CELL_i in active mode to those in idle mode. This criterion is straightforward since the ping-pong handover means “two-way” handover. In this embodiment, it is notable that a range of difference (e.g. 10%-15%) between two parameters is allowable for these two parameters to be symmetric.

If the criteria (1-1) to (1-4) are met for the pair of CELL_i and CELL_j according to the collected statistical data, then in the step 13, the cell pair can be judged to have a high probability of ping-pong handover effect. By adjusting the thresholds used in the criteria (1-1) to (1-3), we can change the degree of strictness in these criteria. For example, if these thresholds are adjusted higher, the cell pair with more serious ping-pong effect will be detected.

In a varied version of the first embodiment, the weight of the criteria (1-1) to (1-3) is considered. There are two ways to execute the step 13 of FIG. 1:

• • (1) For each of the criteria (1-1) to (1-3), a rank list of all the qualified cell pairs is first generated according to the value of the related parameter. For example, for the criterion (1-1), the qualified cell pairs are ranked according to their HO_ij values. Next, a respective weighting score is given to each qualified cell pair according to its position in the respective rank list of the criteria (1-1) to (1-3). The weighting score for the same position of the rank list may be different for different criteria. Then, the sum of all the weighting scores for each qualified cell pair is calculated, and we can compare the sum with a threshold to decide whether the qualified cell pair has the ping-pong handover effect.
  • (2) For each of the criteria (1-1) to (1-3), the related parameter of each qualified cell pair is first multiplied with a weighting coefficient to generate a weighting score. The weighting coefficient for different criteria may be different. Next, the sum of all the weighting scores for each qualified cell pair is calculated. Finally, we can compare the sum with a threshold to decide whether the qualified cell pair has the ping-pong handover effect.

The Second Embodiment

The second embodiment is applied to the network in which the statistical data of handover is non-directional. The non-directional handover of CELL_j (denoted by HO_j) means the times of the connection between the network and the mobile unit changing from via any neighboring cell of CELL_j to via CELL_j, without differentiating the source cell. Thus, in the second embodiment, the items collected in the step 11 include: (1) non-directional handover of each cell. (2) MTC and MOC within each cell.

In the step 12, the set of criteria as described below is applied to each cell (e.g. CELL_j):

• • (2-1) HO_j is higher than a threshold TH₄.
  • (2-2) A parameter P3 is higher than a threshold TH₅. P3 is defined by a ratio between HO_j and the sum of MTC and MOC within CELL_j. That is, P3=HO_j/(MTC_j+MOC_j).

If the criteria (2-1) and (2-2) are met for CELL_j according to the collected statistical data, the step 13 can only determine that there is a high probability of the ping-pong effect to exist between CELL_j and one or more neighboring cells of CELL_j, since the related parameters of the criteria (2-1) and (2-2) are non-directional. Similarly, the degree of strictness in the criteria (2-1) and (2-2) can be changed by adjusting the thresholds used in these criteria.

The Third Embodiment

The third embodiment is applied to the network in which the statistical data of handover back can be collected. The handover back means the times of backward handover occurring within a preset period of time after forward handover. The notation of HO_ij—back is used to represent the handover back of HO_ij. In the third embodiment, the items collected in the step 11 include: (1) Directional handover between each cell pair. (2) Handover back between each cell pair.

In the step 12, the set of criteria as described below is applied to each cell pair (e.g. CELL_i and CELL_j):

• • (3-1) A parameter P4 defined by a ratio of HO_ij—back to HO_ij is higher than a threshold TH₆.
  • (3-2) HO_ij—back is higher than a threshold TH₇. This criterion means that a cell pair with a relatively low handover back is not taken into consideration.

If the criteria (3-1) and (3-2) are met for the pair of CELL_i and CELL_j according to the collected statistical data, then the cell pair can be judged to have a very high probability of the ping-pong handover effect in the step 13. Similarly, the degree of strictness in the criteria (3-1) and (3-2) can be changed by adjusting the thresholds used in these criteria.

There is an assisting criterion in the step 12 for helping to detect the cell pair with the ping-pong effect. The assisting criterion is that if the environment of a cell causes signal fluctuation higher than a level (e.g. an allowable or average level of the network), then the cell is more probable to suffer from the ping-pong effect than other cells. This assisting criterion can be applied to the first to third embodiments mentioned above.

FIG. 2 is a flow chart of a preferred embodiment of the method for reducing the ping-pong handover effect within the cellular network according to the present invention. As shown in FIG. 2, the flow comprises the steps of:

• • 21 collecting network traffic statistic data of a plurality of items;
  • 22 determining whether a set of criteria is met according to the network traffic statistic data;
  • 23 detecting the ping-pong handover effect between the cells according to the result of the step 22; and
  • 24 changing a handover parameter for a cell pair with the ping-pong handover effect.

The steps 21 to 23 are similar to the steps 11 to 13 in FIG. 1. Thus, all the embodiments based on FIG. 1 can also be applied to the flow of FIG. 2 and would not be described again here. In the step 24, the handover parameter includes handover hysteresis, handover window timer or both. If a cell pair of CELL_i and CELL_j is detected as with the ping-pong effect in the step 23, then the handover hysteresis or handover window timer from CELL_i to CELL_j (or from CELL_j to CELL_i) are increased to reduce HO_ij (or HO_ji) in the step 24, thereby reducing the ping-pong effect between CELL_i and CELL_j. However, in the preferred embodiment of FIG. 2, the change of handover parameter in the step 24 must not cause negative effect on the wireless communication network. For example, if the handover parameter is adjusted higher than an allowable threshold, the network performance will be lowered since the signal quality of mobile units may be damaged without necessary handover. Another example of negative effect is that the change of the handover parameter may push the handover to another neighboring cell which is already overloaded.

While the present invention has been shown and described with reference to the preferred embodiments thereof and in terms of the illustrative drawings, it should not be considered as limited thereby. Various possible modifications and alterations could be conceived of by one skilled in the art to the form and the content of any particular embodiment, without departing from the scope and the spirit of the present invention.

Claims

1. A method for detecting ping-pong handover effect of a cellular network, comprising: collecting network traffic statistic data of a plurality of items, wherein each of the items corresponds to a state change of a plurality of mobile units within cells of the cellular network; determining whether a plurality of criteria are met according to the network traffic statistic data; and detecting the ping-pong handover effect between the cells according to a result of the determining step.

2. The method according to claim 1, wherein the items include handover between a first cell and a second cell of the cells.

3. The method according to claim 2, wherein the items further include at least one of mobile terminating call (MTC) and mobile originating call (MOC) within the first and second cells.

4. The method according to claim 3, wherein the criteria comprise comparing a first parameter with a first threshold, wherein the first parameter is defined according to the handover from the first cell to the second cell and at least one of the MTC and MOC within one of the first and second cells.

5. The method according to claim 4, wherein the criteria further comprise comparing the handover from the first cell to the second cell with a second threshold.

6. The method according to claim 5, wherein the criteria further comprise the handover from the first cell to the second cell is symmetric to that from the second cell to the first cell, and the first parameter is symmetric to a first reverse parameter defined according to the handover from the second cell to the first cell and at least one of the MTC and MOC within one of the first and second cells.

7. The method according to claim 2, wherein the cells of the cellular network are grouped into a plurality of location areas, and the items further include location area update of each cell and handover between the cells.

8. The method according to claim 7, wherein the criteria comprise comparing a second parameter with a third threshold, wherein the second parameter is defined according to signal movements from the first cell to the second cell in active mode and in idle mode.

9. The method according to claim 8, wherein the criteria further comprise the second parameter is symmetric to a second reverse parameter defined according to signal movements from the second cell to the first cell in active mode and in idle mode.

10. The method according to claim 1, wherein the items include non-directional handover between a first cell and neighboring cells thereof, and at least one of MTC and MOC within the first cell.

11. The method according to claim 10, wherein the criteria comprise: (1) comparing the non-directional handover with a fourth threshold; and (2) comparing a third parameter with a fifth threshold, wherein the third parameter is defined according to the non-directional handover and at least one of the MTC and MOC within the first cell.

12. The method according to claim 2, wherein the items further include handover back between the first and second cells.

13. The method according to claim 12, wherein the criteria comprise comparing a fourth parameter with a seventh threshold, wherein the fourth parameter is defined according to the handover and the handover back between the first and second cells.

14. The method according to claim 13, wherein the criteria comprise comparing the handover back between the first and second cells with a sixth threshold.

15. The method according to claim 1, wherein the criteria comprise comparing signal fluctuation caused by environment of the cells with a level.

16. The method according to claim 1, wherein each of the criteria has a related parameter which has a corresponding value for each of a plurality of cell pairs, wherein the corresponding value is determined according to the network traffic statistic data.

17. The method according to claim 16, wherein the detecting step comprises: generating a weighting score for each qualified cell pair of each of the criteria according to the corresponding value of the qualified cell pair; comparing a sum of the weighting scores of each qualified cell pair with a threshold; and deciding whether the qualified cell pair has the ping-pong handover effect according to a result of the comparing step.

18. A method for reducing ping-pong handover effect of a cellular network, comprising: collecting network traffic statistic data of a plurality of items, wherein each of the items corresponds to a state change of a plurality of mobile units within cells of the cellular network; determining whether a plurality of criteria are met according to the network traffic statistic data; detecting the ping-pong handover effect between the cells according to a result of the determining step; and changing a handover parameter for the cells with the ping-pong handover effect.

19. The method according to claim 18, wherein the handover parameter is handover hysteresis.

20. The method according to claim 18, wherein the handover parameter is handover window timer.

21. The method according to claim 18, wherein the change of the handover parameter does not cause negative effect on the cellular network.

22. The method according to claim 21, wherein the handover parameter is not higher than an allowable threshold.