WO2004057882A1 - Forward-link rate scheduling method and scheduler - Google Patents

Abstract

The present invention discloses a forward-link rate scheduling method and a scheduler, which method comprises the step of: gathering the information of the maximal scheduling resource available to the system, as well as the data service request information from the scheduling users, wherein the user scheduling measurement includes the short-term scheduling measurement, long-term scheduling measurement and normalized long-term scheduling measurement; creating the scheduling user queue in accordance with the scheduling measurement; allocating the rate and power in accordance with the scheduling user queue and the system-available scheduling resource; clearing the users which have obtained the resource from the scheduling user queue and notifying the allocation result to selectors and channel units. The invention allows for the wireless environment and available resource of the system comprehensively, which is capable of improving the utilization efficiency of the system resource, as well as reducing the transmission delay of scheduling user data, while allowing for the fairness of scheduling.

Description

 TECHNICAL FIELD The present invention relates to the field of data communications, and more particularly to a code division multiple access system forward link rate scheduling method and scheduler for providing high speed data services. BACKGROUND With the continuous expansion of the wireless service field and the rapid development of the Internet service, a simple call cannot meet the requirements of the user, and many users hope that the mobile communication system can provide the vast majority of the existing communication network (including ISDN, ADSL). Part of the business, therefore, in the new generation of mobile communication systems, the data service represented by multimedia services will dominate, which requires the system to provide more powerful support for data services and data and voice hybrid services, according to data services. The characteristics of reasonable and efficient use of wireless resources are the core to achieve this goal.

 When the base station has a large amount of data to be transmitted to the remote users, the base station scheduling the rate allocated to these users is an important part of rational and effective use of the wireless resources. The paper "Downlink Scheduling in CDMA Data Networks" states that if the base station allocates available system resources to only one user at a time during scheduling, the utilization of the forward link can be maximized. However, this scheduling method tends to favor users with better channel environments. Those with relatively poor channel environments will have longer delays. This extreme unfairness will be unacceptable to users. Therefore, a good rate scheduling method should not only improve the utilization efficiency of wireless resources, but also ensure the fairness of users in the system.

US Patent 6, 229, 795B1, "System For Allocating Resources In a Communication System" proposes a scheduling method that takes into account resource utilization efficiency and user fairness: Let the instantaneous rate assigned by the base station to the user be R, and the average rate is C, if R > C, the base station assigns a smaller weight to the user; if R < C, the base station assigns the user a larger weight, and the user's weight can be accumulated. The principle of base station scheduling is: The smaller the user's weight, the higher the user priority, and vice versa. However, the rate scheduling method requires the base station to allocate the rate only according to the channel environment in which the user is located: a user with a good channel environment allocates a higher rate, and a user with a poor channel environment allocates a lower rate. In fact, when the system has more resources available Users with poor channel conditions can also assign higher rates. Therefore, when scheduling the rate, considering the wireless environment in which the user is located and the available resources of the system, the system resources can be utilized more effectively.

 When the system schedules the user's rate, the smoothness of the scheduling is also a factor that cannot be ignored. The smoothness of scheduling refers to: When raising the priority of some users, the priority of other users in the system should not be drastically reduced, that is, the promotion and reduction of user priorities should be a smooth process.

 It should be said that the existing rate scheduling method fails to comprehensively consider the wireless environment in which the user is located and the smoothness of the available resources and scheduling of the system, which must be considered and improved. SUMMARY OF THE INVENTION One object of the present invention is to provide a method for forward link rate scheduling in a communication network;

 Another object of the present invention is to provide a means for forward link rate scheduling in a communication network.

 In order to achieve the object of the present invention, the present invention proposes the following technical solutions.

 The method for forward link rate scheduling of the present invention includes the following steps:

 a, collecting the maximum scheduling resource information available to the system and scheduling the data service request information of the user; b, calculating the scheduling metric of the scheduling user, and the scheduling metric of the user includes the short-term scheduling metric, the long-term metric, and the normalized long-term scheduling metric;

 c, create a scheduling user queue according to the scheduling metric;

 d, according to the scheduling user queue and system available scheduling resource allocation rate, power;

 e, clearing the user who obtained the resource from the scheduled user queue, and notifying the selector and the channel unit of the allocation result;

 f, one scheduling period ends, wait and schedule the next cycle as described above.

 The step c of creating a scheduled user queue further includes the following steps:

 Cl, create an active scheduling user queue;

C2, creating a non-activated scheduling user queue; C3. detecting whether the long-term scheduling metric of the activated scheduling user is greater than a fixed threshold constant T1;

 C4, detecting whether the difference between the long-term scheduling metric of the activated scheduling user and the long-term scheduling metric of the inactive scheduling user is greater than a fixed threshold constant T2;

 C5. Insert the inactive scheduling user in front of the active scheduling user, and clear the inactive user from the inactive scheduling user queue;

 C6. Re-detect whether all active scheduling users have been traversed, and if so, end; otherwise, perform steps cl to c5 repeatedly until all active scheduling users are traversed.

 In the step c3, when detecting that the long-term scheduling metric of the active scheduling user is greater than the fixed threshold constant T1, proceed to the next step c4; otherwise, step c6 is directly performed.

 In the step c4, when the difference between the long-term scheduling metric of the activated scheduled user and the long-term scheduling metric of the inactive scheduling user is greater than the fixed threshold constant T2, step c5 is performed; otherwise, step c6 is directly performed.

 After the step e, it is detected whether the system has available scheduling resources or the scheduling queue is empty. If it is not empty, if the returning step is empty, the scheduling metric of the scheduling user is updated first, and then the step is performed. The scheduling metric of the scheduling user is updated, and the following steps are further included in order:

 El, update the target rate of the scheduled user, the target rate may be the request rate of the latest packet call arrival, or the maximum value of the latest scheduled rate of the scheduled user or the average value of the latest scheduled rate of the scheduled user;

 E2, updating the short-term scheduling metric of the scheduled user;

 E3. Update the long-term scheduling metric of the scheduled user;

 E4. Update the normalized long-term scheduling metric of the scheduled user.

 The scheduler for forward link rate scheduling of the present invention is included in a base station of a code division multiple access communication system, and the scheduler includes a memory, a timer, and a controller.

 The memory receives and saves data service request information of the scheduling user and maximum scheduling resource information of the system;

 The output of the timer is connected to the controller, and the timer provides the required timing for the controller to perform forward link rate scheduling;

The controller is bidirectionally connected to the memory, and the controller is configured to calculate a scheduling metric of the scheduling user, A scheduling user queue is constructed, a scheduling resource allocation rate is obtained according to the scheduling user queue and the system, and the user who obtains the resource is cleared from the scheduling user queue, and the allocation result is sent to the selector and the channel unit as the output of the scheduler. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a schematic diagram of a simplified code division multiple access communication system.

 2 is a schematic diagram showing the structure of a base station including the scheduler of the present invention.

 3 is a schematic structural view of a scheduler of the present invention.

 4 is a schematic flow chart of a scheduling method of the present invention.

 FIG. 5 is a schematic flowchart of a method for creating a scheduling user queue according to the present invention.

 6 is a schematic flow chart of a scheduling metric updating method of the present invention. BEST MODE FOR CARRYING OUT THE INVENTION Code Division Multiple Access (CDMA) is the most competitive multiple access modulation technique in next generation mobile communications. In order to explain the present invention more clearly, an introduction to the code division multiple access communication system is first described. Referring to FIG. 1, the mobile station (MS) 107A or 107B is controlled by a base transceiver station (BTS) IOIA or 101B and a base station. The device (BSC) 105 A communicates, and at some point the mobile station may be in contact with both base stations at the same time. For example, the mobile station 107B simultaneously communicates with the base transceiver stations 101A and 101B. The base station controller 105A in turn communicates with the PSTN or ISDN via the Mobile Switching Center (MSC) 103, and an MSC 103 can also be connected to other base station controllers 105B and the like.

Please refer to FIG. 2, which illustrates a base station structure including a scheduler of the present invention. The data source 201 contains a large amount of packet data that needs to be transmitted to the remote mobile station 107. The data source 201 may also be located at the base station controller 105 or the mobile switching center 103 due to differences in actual system implementation. There are a large number of selector units 205 and channel units 207 inside the base station. Among them, the selector unit 205 is used to control the communication between the data source 201 and the channel unit 207. The selector unit 205 and the channel unit 207 are simultaneously controlled by the scheduler 203. Under the instruction of the scheduler 203, the selector unit 205 transmits the packet data stored in the data source 201 to the channel unit 207, and the channel unit 207 pairs the selector. The data sent by unit 205 performs a series of processing, These processes typically include: adding control bits, adding CRC bits, adding tail bits, encoding, interleaving, Wlash quadrature modulation, PN short code modulation, and the like. Finally, transmitter 209 transmits the signal processed by the channel unit to the remote mobile station.

 For convenience of explanation, here, the structure of the scheduler of the present invention will be described first. Referring first to FIG. 3, FIG. 3 illustrates the internal structure of the scheduler, and the scheduler 203 includes a memory 301, a timer 303, and a controller. 305. The memory 301 is configured to save base station available scheduling resource information and data service request information. The available scheduling resource information of the base station includes resources, rates, orthogonal codes, channel units, and the like that the base station can allocate to the mobile station requesting the data service, and the data service request information includes the identification number of the mobile station applying for the data service, and the data amount of the mobile station. Request rate, scheduling metrics for mobile stations, recent scheduling rate, etc. The timer 303 provides the timing required for the controller 305 to perform forward link rate scheduling. The timer 303 can be implemented using a system clock, an oscillator on a circuit board that is locked to an external signal, or a receiver that is driven by a memory unit that is synchronized with an external source. When the scheduling time expires, the controller 305 performs rate scheduling on the mobile station having the data service request according to the information provided by the memory 301, and notifies the corresponding selector unit 205 and the channel unit 207 of the result of the rate scheduling.

 The method of the present invention provides a method for scheduling forward link data transmission in a code division multiple access communication system, in which at least one cell and at least one scheduled user are included.

 The method of the invention comprises the following steps:

 a, collecting the maximum scheduling resource information available to the system and scheduling the data service request information of the user; b, calculating the scheduling metric of the scheduling user, and the scheduling metric of the user includes a short-term scheduling metric, a long-term metric, and a normalized long-term scheduling metric;

 c Create a scheduled user queue based on the scheduling metrics;

 d, according to the scheduling user queue and system available scheduling resource allocation rate, power;

 e, clearing the user who obtained the resource from the scheduled user queue, and notifying the selector and the channel unit of the allocation result;

 f, one scheduling period ends, wait and schedule the next cycle as described above.

After the step e, it is detected whether the system has available scheduling resources or the scheduling queue is empty. If it is not empty, it returns to step d; if it is empty, it first updates the scheduling metric of the scheduling user, and then performs step ^ The implementation of the above method can continue to refer to Figure 4, in Figure 4, since the method of the present invention is cured by The software in the controller 305 is completed. For convenience of description, each step is represented by a module. In the module 401, the controller 305 extracts the system maximum available scheduling resource information and data service request information from the memory 301, and the information is The controller performs the basis of rate scheduling. In the module 403, the controller 305 creates a scheduling user queue according to the scheduling metric of the user. The principle of creating a scheduling user queue is that the higher the priority of the scheduling user, the higher the position of the user in the scheduling user queue. In block 405, controller 305 assigns a rate to the scheduling user in the order in which the user queues are scheduled, which rate allocation is limited by the currently available scheduling resources. In block 407, the controller 305 clears the scheduled user that obtained the rate allocation from the dispatch queue. At the same time, the controller 305 notifies the corresponding selector unit 205 to transmit the scheduled user data to the channel unit 207, and the channel unit 207 is based on the controller. The rate assigned by 305 processes the data accordingly. In block 409, if the controller 305 detects that the system has available scheduling resources and the scheduling queue is not empty, the controller 305 repeats the flow of the modules 405-409; otherwise, directly executes the module 411, the scheduling metric for the scheduling user. Update. In block 413, controller 305 waits for the next scheduling cycle in accordance with the indication of timer 303 and repeats the flow of modules 401-411.

 Please continue to see Figure 5,

 When you create a scheduled user queue in step c, follow these steps:

 Cl, create an active scheduling user queue;

 C2, create a non-activated scheduling user queue,

 C3, detecting whether the long-term scheduling metric of the activated scheduling user is greater than a fixed threshold constant T1;

 C4, detecting whether the difference between the long-term scheduling metric of the activated scheduling user and the long-term scheduling metric of the inactive scheduling user is greater than a fixed threshold constant T2;

 C5. Insert the inactive scheduling user in front of the active scheduling user, and clear the inactive user from the inactive scheduling user queue;

 C6. Re-detect whether all active scheduling users have been traversed, and if so, end; otherwise, perform steps cl to c5 repeatedly until all active scheduling users are traversed.

 In the step c3, when detecting that the long-term scheduling metric of the active scheduling user is greater than the fixed threshold constant T1, proceed to the next step c4; otherwise, step c6 is directly performed.

In the step c4, when it is detected that the difference between the long-term scheduling metric of the active scheduling user and the long-term scheduling metric of the inactive scheduling user is greater than the fixed threshold constant T2, proceed to step c5; otherwise, the step is directly performed. Step c6.

Based on the same considerations above, for convenience of explanation, we classify the mobile stations participating in the scheduling into two categories: In the most recent rate scheduling process, the scheduling user who obtains the rate allocation of the controller 305 is called an active scheduling user; The scheduled user of the rate allocation is called the inactive scheduling user. In block 501, the controller 305 creates an active scheduling user queue according to the normalized long-term scheduling metric of the active scheduling user. The smaller the normalized long-term scheduling metric of the scheduling user, the more the scheduling user is in the active scheduling user queue. before. In block 503, the controller 305 creates an inactive scheduled user queue based on the normalized long-term scheduling metric of the inactive scheduling user, the principle of which is the same as the module 501. In block 505, controller 305 executes module 507 if it detects that the long-term scheduling metric for the active scheduled user satisfies equation (1); otherwise, executes module 513. In block 507, the controller 305 executes the module 509 if it detects that the long-term scheduling metric of the active scheduled user and some inactive scheduled user satisfies equation (2); otherwise, executes the module 511. Both Tl and Τ2 in equations (1) and (2) are fixed threshold constants. Careful selection of thresholds T1 and Τ2 allows the system to smoothly schedule the user's rate. In block 509, the controller 305 inserts the inactive scheduled user in front of the corresponding active scheduled user and simultaneously clears the inactive scheduled user from the inactive scheduled user queue. At block 511, controller 305 detects if all inactive scheduled users have been traversed, and if so, executes module 513, and if not, repeats the flow of modules 507-511. At block 513, controller 305 detects if all active scheduled users have been traversed, and if so, executes module 515, otherwise, the flow of modules 503-513 is repeated. At block 505, the controller 305 merges the active scheduled user queue with the inactive scheduled user queue (the active scheduled user queue is first) as the scheduled user queue, ending the process of creating the scheduled user queue. Activate the long-term scheduling metric of the scheduled user > T1 ( 1 ) Activate the long-term scheduling metric of the scheduled user - the long-term scheduling metric of the inactive scheduling user > Τ 2 (2) It should be noted that the method of creating the scheduling user queue is different due to different system implementations. It should be different. For example, when the impact of the signaling overhead of the data traffic channel established by the system on the forward link utilization is negligible, the controller can only create a scheduling user queue according to the normalized long-term scheduling metric, and schedule the user's return. The smaller the long-term scheduling metric is, the higher the position of the scheduling user in the scheduling user queue. These methods are all in the hair Within the scope of the inclusion.

 Referring to FIG. 6 again, in updating the scheduling metric of the scheduled user, the following steps are further included in order:

 El. Update the target rate of the scheduled user. The target rate may be the request rate of the latest packet call arrival, or the maximum value of the latest scheduled scheduling rate of the user or the average of the latest scheduled scheduling rate of the scheduling user.

 E2, update the short-term scheduling metric of the scheduled user;

 E3. Update the long-term scheduling metric of the scheduled user;

 E4. Update the normalized long-term scheduling metric of the scheduled user.

 Based on the foregoing method and convenience of description, the scheduling metric step of the above-mentioned update scheduling user is described as follows.

In block 603, the controller 305 updates the short-term scheduling metric of the scheduled user. The short-term scheduling metric is defined as shown in equation (3): represents the instantaneous rate obtained by the scheduling user; R represents the target rate of the scheduling user; "represents the weighting factor corresponding to the scheduling user. The weighting factor" and the instantaneous rate obtained by the scheduling user, The resources consumed are all related. For example, for a forward link power limited CDMA cellular system, it is assumed that the instantaneous rate obtained by the scheduling user is that the nominal power consumed by the base station 101 to support the rate is ¹ ^ ^r) . The power actually consumed by the base station 101 to support the rate is ^p (^, then the weighting factor can be obtained by the equation (4), where is the control factor, reflecting the degree to which the base station 101 tends to schedule the user with a better wireless channel environment. The larger the channel users with better wireless channel environment, the more opportunities they have to obtain the speed* allocation, and the greater the capacity of the forward link resources. However, when the system resources are tight, the wireless channel environment is not good due to excessive conference conditions. The user can't get the rate for a long time. Therefore, when selecting the factor, the system engineer should balance the following two aspects:

 (1) increase the capacity of the forward link;

(2) The scheduling user who avoids the bad wireless environment cannot obtain the rate allocation for a long time or only obtains a very low rate allocation.

In block 605, the controller 305 updates the long-term scheduling metric of the scheduled user. The long-term scheduling metric reflects the degree of satisfaction of the scheduling user to the resource allocation in multiple scheduling times. The short-term scheduling metric of the scheduling user in the nth scheduling is ( _n ), and the user is scheduled for long-term scheduling within n schedulings. The measure &r (n) is:

∑S _S r (i) n≤N

 ί=1

S _LT (n) = (5)

∑S _ST (i) n > N where is a fixed constant.

In block 607, the controller 305 updates the normalized long-term scheduling metric of the scheduled user. The normalized long-term scheduling metric reflects the average satisfaction of the scheduling users to the resource allocation in multiple scheduling periods. The long-term scheduling metric of a scheduling user in n scheduling is r (n), then the user's normalization. long-term scheduling is a measure ^ύ,

 For a scheduled user to receive the rate allocation for the first time, we assume that the initial long-term scheduling metric and the normalized long-term scheduling metric of the user before the rate allocation is 0.

 In the foregoing description of the flow of the module 607, the soft handover is not discussed. As shown in FIG. 1, when the mobile station 107A stops communicating with the base station 101 and starts to communicate with the base station 101B, in order to ensure the smoothness of the base station 101B rate scheduling, the base station 101B shall be randomly assigned to mobile station 107A - an initial normalized long-term scheduling metric value that is between the maximum normalized long-term scheduling metric value and the minimum normalized long-term scheduling metric within the system.

Update the target rate of the scheduled user. The target rate reflects the expected rate of the user. The target rate may have different reflection forms due to different system implementations. For example, it may schedule the request rate of the user's last packet call, or schedule the user. The maximum or average value of the recent scheduling rates is replaced by the number of reflections, all of which are within the scope of the present invention. The foregoing is a specific embodiment to illustrate and describe the present invention. Those skilled in the art can make modifications without departing from the spirit and scope of the present invention. For example, changing the transmit power in equation (4) to receive power, the present invention can also be used for rate scheduling of reverse link data services. It will not be described in detail here.

Industrial Applicability Due to the adoption of the above scheduling method and scheduler, the present invention can comprehensively consider the wireless environment in which the user is located and the available resources of the system, can effectively improve the utilization efficiency of system resources, and reduce the transmission delay of scheduling user data. The scheduling method and the device also take into consideration the fairness of the scheduling user. At the same time, the smoothness of the scheduling is also considered. Considering the smoothing of the scheduling, on the one hand, the number of times the data traffic channel is established and released can be reduced, usually, the establishment of the traffic channel and The release requires more signaling overhead, and the reduction of the number of times the service channel is established and released means that the resource utilization rate is improved. On the other hand, the user's satisfaction with the system can be guaranteed to a certain extent.

Claims

Rights request A method for scheduling a forward link rate, the method comprising the steps of: a, collecting the maximum scheduling resource information available to the system and the data service request information of the scheduling user; b, calculating the scheduling user's The scheduling metric, the user's scheduling metrics include short-term scheduling metrics, long-term scheduling metrics, and normalized long-term scheduling metrics;  c Create a scheduled user queue based on the scheduling metrics;  d, according to the scheduling user queue and system available scheduling resource allocation rate, power;  e, clearing the user who obtained the resource from the scheduled user queue, and notifying the selector and the channel unit of the allocation result;  f, one scheduling period ends, wait and schedule the next cycle as described above.  2. The method for scheduling a forward link rate according to claim 1, wherein the step c of creating a scheduling user queue further comprises the following steps:  c Create an active scheduling user queue;  C2, creating a non-activated scheduling user queue;  C3. detecting whether the long-term scheduling metric of the activated scheduling user is greater than a fixed threshold constant T1;  C4, detecting whether the difference between the long-term scheduling metric of the activated scheduling user and the long-term scheduling metric of the inactive scheduling user is greater than a fixed threshold constant T2;  C5. Insert the inactive scheduling user in front of the active scheduling user, and clear the inactive user from the inactive scheduling user queue;  C6. Re-detect whether all active scheduling users have been traversed, and if so, end; otherwise, perform steps cl to c5 repeatedly until all active scheduling users are traversed.  3. The method for scheduling a forward link rate according to claim 2, wherein:  In the step c3, when detecting that the long-term scheduling metric of the active scheduling user is greater than the fixed threshold constant T1, proceed to the next step c4; otherwise, step c6 is directly performed.  4. The method for scheduling a forward link rate according to claim 2, wherein: In the step c4, when detecting the long-term scheduling metric of the activated scheduling user and the inactive scheduling user If the difference between the long-term scheduling metrics is greater than the fixed threshold constant T2, then step c5 is performed; otherwise, step c6 is directly performed.  The method for scheduling a forward link rate according to claim 1, wherein: after the step e, detecting whether the system has available scheduling resources or the scheduling queue is empty, if not, returning If the step mountain is empty, the scheduling metric of the scheduling user is updated first, and then the step is executed ^ The method for scheduling a forward link rate according to claim 5, wherein, in updating the scheduling metric of the scheduling user, the method further comprises the following steps:  El, update the target rate of the scheduled user, the target rate may be the request rate of the latest packet call arrival, or the maximum value of the latest scheduled rate of the scheduled user or the average value of the latest scheduled rate of the scheduled user;  E2, updating the short-term scheduling metric of the scheduled user;  E3. Update the long-term scheduling metric of the scheduled user;  E4. Update the normalized long-term scheduling metric of the scheduled user.  7. A forward link rate scheduler, the scheduler being included in a base station of a code division multiple access communication system, wherein:  The scheduler includes a memory, a timer, a controller,  The memory receives and saves data service request information of the scheduling user and maximum scheduling resource information of the system;  The output of the timer is connected to the controller, and the timer provides the required timing for the controller to perform forward link rate scheduling;  The controller is bidirectionally connected to the memory, and the controller is configured to calculate a scheduling metric of the scheduling user, create a scheduling user queue, allocate a rate according to the scheduling user queue and the system available scheduling resource, and clear the user who obtains the resource from the scheduling user queue. The result of the allocation is sent to the selector and channel unit as the output of the scheduler.