WO2011038529A1 - Scheduling method and scheduler - Google Patents

Abstract

A scheduling method and scheduler are provided in the present invention, which can take into account traffic balancing in relay nodes in order to avoid the problem of the occurrence of the buffer overflow or buffer underflow in relay nodes. The scheduling method includes the steps of: from each relay node, receiving a channel quality indicator of a backhaul link, buffer state information corresponding to each user equipment served by the relay node and an average rate of an access link; according to the channel quality indicator of the backhaul link, estimating an average rate of the backhaul link and determining a scheduling priority of the backhaul link; according to the buffer state information, determining whether there is an instance of the buffer overflow in the relay node, and according to the estimated average rate of the backhaul link and the average rate of the access link, determining whether there is a possibility of the buffer underflow in the relay node; according to predetermined rules, adjusting the scheduling priority in order to degrade the scheduling priority of the backhaul link in the case that it is determined that the buffer overflow occurs and to improve the scheduling priority of the backhaul link in the case that it is determined that there is a possibility of the buffer underflow; and scheduling the backhaul link with a higher priority. Use of the method in the present invention can effectively solve the problem of traffic misallocation in relay nodes, thus effectively avoiding waste of radio resources of the backhaul link or the access link, and improving performance of the whole system.

Description

 Scheduling method and scheduler

 The present invention relates to the field of mobile communications, and more particularly to a scheduling method and scheduler in a multi-hop cellular system, wherein traffic balancing on a relay node is considered to avoid buffer overflow on the relay node. Or underflow problems. Background technique

In recent years, research on relay technology has gradually attracted people's attention. Relay technology is considered as a good candidate for improving cell capacity and extending coverage. Here, the relay node RN is a network node dedicated to storing and forwarding data packets from the base station eNodeB (eNB) to the user equipment UE, and, in contrast, data packets from the user equipment UE The storage is performed and forwarded to the base station _e NB. Based on specific access criteria, such as maximum received power, some user equipment is associated with the relay node and is not directly served by the eNB.

Unlike traditional cellular networks, data packets from/to user equipment within the coverage area of the relay node are buffered on the relay node before reaching the destination. However, in this case, there are some problems with resource allocation. For multi-hop cellular systems, the radio resources associated with the serving eNB and its cooperating relay nodes are sometimes fixed for a specific period to avoid frequent resource allocation and information interaction. That is, the resources of the serving eNB and the relay node connected thereto are allocated in a semi-static manner. In this case, each node (eNB or relay node) typically runs a per-hop scheduler, ie each service node only schedules its secondary link directly. However, the general hop-by-hop scheduling scheme does not consider the relationship between the relay backhaul link and the relay access link. Since the transmission capability of the user equipment UE-R directly served by the relay node is different on different relay hops on the available bandwidth, on the one hand, if the data packet received on the serving relay node of the UE-R is More than the number of data packets transmitted, it will cause a buffer overflow problem. On the other hand, if the buffer of the relay node corresponding to UE-R becomes empty when scheduling UE-R, it causes a problem of buffer underflow. That is to say, the mismatch of traffic between the backhaul link of UE_R and its access link may cause the system spectrum efficiency to decrease. In the case where the multi-hop cellular system adopts a hop-by-hop scheduling scheme, it is desirable to find a relay buffer control method capable of solving the problem of traffic mismatch on the relay node. Summary of the invention In the present invention, an effective relay buffer control mechanism in a multi-hop cellular system is proposed, which can overcome the above drawbacks and avoid waste of radio resources on the backhaul link and the access link, thereby improving the overall system. performance.

 The main concepts of the invention are described below:

 1. During each given relay state feedback period, each relay node reports its buffer status corresponding to each served user equipment to its serving node (eNB) (eg, one bit feedback is used for It is checked whether the buffer size of each UE-R is greater than the buffer threshold) and the average rate on the access link.

 2. Using the feedback of the relay status corresponding to the UEJR, the eNB first estimates whether the relay buffer has an overflow or underflow, and then increases or decreases the scheduling opportunity of the UE-R on the backhaul link by adjusting the scheduling priority. .

 3. The eNB may schedule one UE with a higher priority for each resource partition or a relay node of the serving UE at any time.

 The proposed method can utilize the feedback from each relay node to effectively solve the problem of traffic mismatch on the relay node, which is acceptable.

 According to a first aspect of the present invention, a scheduling method is provided, comprising the steps of: receiving, from each relay node, a channel quality indicator of a backhaul link, a buffer corresponding to each user equipment served by the relay node State information and an average rate of the access link; estimating an average rate of the backhaul link according to a channel quality indicator of the backhaul link, determining a scheduling priority of the backhaul link; determining whether the relay node exists on the buffer according to the buffer status information Overflow condition, and determining whether the relay node has buffer underflow according to the estimated average rate of the backhaul link and the average rate of the access link; adjusting the scheduling priority according to a predetermined rule, so that the buffer is determined to exist In the case of overflow, the scheduling priority of the backhaul link is lowered, and in the case where it is determined that there is a possibility of buffer underflow, the scheduling priority of the backhaul link is increased; and the backhaul link with higher priority is scheduled. .

 Preferably, the step of determining whether the relay node has buffer underflow according to the average rate of the estimated backhaul link and the average rate of the access link comprises: an average transmission traffic on the available bandwidth of the backhaul link at the base station When the average transmission traffic of the access link on the available bandwidth of the relay node is smaller, it is determined that the relay node has buffer underflow.

 Preferably, the scheduling method is performed for each resource partition.

Preferably, the predetermined rule comprises: adjusting the scheduling priority to zero in the case that it is determined that there is a buffer overflow. Preferably, the predetermined rule comprises: exponentially increasing the scheduling priority in association with an average rate of the backhaul link and an average rate of the access link in the case of determining that there is a possibility of buffer underflow.

 Preferably, the predetermined rule comprises: adjusting the scheduling priority without determining that there is no buffer overflow and there is no buffer underflow.

 Preferably, the scheduling priority is determined according to a proportional fairness criterion or a polling criterion.

 Preferably, the average rate of the access link received from each relay node corresponding to each user equipment served by the relay node is a channel of the access link fed back by the relay node according to the user equipment. The quality indicator is estimated.

 According to a second aspect of the present invention, a scheduler is provided, including: a scheduling information receiving module, configured to receive scheduling required information from a relay node connected thereto, including a channel quality indicator of a backhaul link, and a buffering status information corresponding to each user equipment served by the relay node and an average rate of the access link; a scheduling priority calculation module, configured to calculate a scheduling priority of the backhaul link according to a channel quality indicator of the backhaul link a relay node buffer information determining unit, configured to determine, according to the buffer status information, whether the relay node has a buffer overflow condition, and determine the medium speed according to the estimated average rate of the backhaul link and the average rate of the access link. Whether the node has the possibility of buffer underflow; the scheduling priority adjustment module is configured to adjust the scheduling priority according to a predetermined rule, so that in the case that it is determined that there is a buffer overflow, the scheduling priority of the backhaul link is lowered, and Increase the scheduling priority of the backhaul link if there is a possibility of buffer underflow; and resource scheduling Block for the backhaul link having a higher priority scheduling.

 Preferably, the predetermined rule comprises: in the case that it is determined that there is a buffer overflow, the scheduling priority adjustment module adjusts the scheduling priority to zero.

 Preferably, the predetermined rule is: in the case that it is determined that there is a possibility of buffer underflow, the scheduling priority adjustment module associates the scheduling priority with the average rate of the backhaul link and the average rate of the access link. The index has increased.

 Preferably, the predetermined rule comprises: in the case that it is determined that there is no buffer overflow and there is no buffer underflow, the scheduling priority adjustment module does not adjust the scheduling priority.

 The beneficial effects of the invention:

 The present invention discloses a scheduling method for a multi-hop cellular system, which can effectively control a relay buffer, which achieves the following beneficial effects with respect to the prior art -

1. Effectively solve the problem of traffic mismatch on the relay node. Therefore, the waste of radio resources on the backhaul link or the access link can be effectively avoided, thereby improving the performance of the entire system. 2. The proposed method is easy to implement and only needs to be modified for the scheduling scheme at the eNB.

 3. The proposed method is widely applicable to multi-hop cellular networks, including two-hop systems and more hop systems. DRAWINGS

 The above and other objects, features and advantages of the present invention will become more apparent from the Detailed Description of Description

 1 is a schematic diagram showing a relay system implementing a scheduling method according to an embodiment of the present invention; FIG. 2 is a block diagram showing a system for implementing a scheduling method according to an embodiment of the present invention; A flowchart of a scheduling method of an embodiment of the invention;

 Fig. 4 is a view showing a comparison of simulation results of the scheduling method and the existing scheduling method according to the present invention. detailed description

 Hereinafter, the present invention will be described based on the drawings. In the following description, some specific embodiments are for illustrative purposes only and should not be construed as limiting the invention. Conventional structures or configurations will be omitted when it may result in obscuring the understanding of the present invention.

 Resource scheduling method for eNB and its cooperative relay node in a specific scheduling period

It is assumed that there are ^ and ^ resource partitions on the available bandwidth _{Q of the} eNB and the available bandwidth of the relay node (the subscript k indicates the index of the relay node). In this case, each relay node runs a hop-by-hop scheduling scheme, ie the eNB schedules resources only for the direct link and the backhaul link, and each relay node schedules resources only for its access link. The proportional fair scheduling scheme and the polling scheduling scheme are common algorithms for controlling packet transmission in the field of wireless communications. Here, the description is made by taking a proportional fair scheduling scheme as an example.

 FIG. 1 is a schematic diagram showing a relay system that implements a scheduling method according to an embodiment of the present invention. As shown in FIG. 1, taking the downlink transmission of the two-hop relay system using the proportional fair algorithm at the eNB as an example, it is assumed that each cell has T relay nodes with /. The user equipments are connected to the eNB and there are ^ user equipments connected to the respective relay nodes RN ( : = 1,..., ).

在现有技术中， 首先计算调度优先级，然后进行资源分配。通常优生级越大 的用户设备将获得调度机会。现有比例公平调度方案的调度优先级可通过下式来 计算： Μ( ) = ~ί=^^，其中 7 ^)是在 t时刻用户设备的瞬时速率， T_mer{t) Assume that if the user equipment z ₀ (z _Q =l,...,J ₀ )ieNB " ₍₎ (« ₍₎ = l .., N _Q ) ± direct service (hereinafter referred to as "macro user equipment"), then Create a direct link from the eNB to the macro user device ^. And if the user equipment 4 is indirectly served through the relay node, respectively create a resource partition "a backhaul link from the eNB to the serving relay node on _{Q >n"} . And resources to the partitions from the serving relay node UE R is a relay access link _"t. Further, eNB from all The user equipment and the relay node obtain channel quality indicator CQI feedback such as MCS. Using the feedback CQI information, the eNB estimates the instantaneous rate of the direct link of the macro user equipment at time t, the resource partition, and the instantaneous rate of the backhaul link corresponding to the user equipment.

In the prior art, the scheduling priority is first calculated, and then resource allocation is performed. Usually user equipment with a higher eugenic level will get a scheduling opportunity. The scheduling priority of the existing proportional fair scheduling scheme can be calculated by: Μ( ) = ~ί=^^, where 7 ^) is the instantaneous rate of the user equipment at time t, T _mer {t)

 T_aver t)

The average rate of user equipment at time t. For example, for a macro user device ^, it is at resource time at time t

_≠

The scheduling priority on the area ^ ₎ is ^(0 .

在现有的比例公平调度方案中，也是采用逐跳调度方案，但是并不考虑中继 回程链路和中继接入链路之间的关系， 存在中继节点流量不匹配的问题， 即中继 缓冲上溢和中继缓冲下溢的问题。

In the existing proportional fair scheduling scheme, the hop-by-hop scheduling scheme is also adopted, but the relationship between the relay backhaul link and the relay access link is not considered, and there is a problem that the relay node traffic does not match, that is, Following the problem of buffer overflow and relay buffer underflow.

 In the present invention, in order to solve the problem of traffic mismatch on the relay node, a scaling factor is added.

After 6 „.( ), the scheduling priority of the proportional fair scheduling scheme is redesigned as follows:

其中， T—ave ( }表示在 t 时刻用户 在回程链路/上的平均速率. 在 t 时刻和资源分区^ )上用户设备 UEJ _A的回程链路 ¾_>/¾的缩放因子 > (^表示如下：

柳 （2) 其中， c^^ )是避免中继缓冲上溢的调度幵关，这里，如果中继缓冲大小 大于用户设备 UE一 的给定中继节点缓冲阈值 , 则《 ，„ (t) =0, 否则 αζ„_ο (ή =ι； >¾是一个小的固定参数。 与用户速率有关， 通常选取使 kT_ik^ [ή 为小数的一个常数， 其值越大调节速度越快， 但效果较差； 反之， 调节速度慢，但效果好。 T_h ( )是在 t时刻用户设备 UEJ _A的令牌计数器， T_ik (k) 按照下面的表达式进行更新：

Where T_ave ( } represents the average rate of the user on the backhaul link at time t. At time t and the resource partition ^ ), the backhaul link of the user equipment UEJ _A is 3⁄4 _>/ 3⁄4 scaling factor > (^ indicates as follows:

Willow (2) Where c^^) is a scheduling switch to avoid relay buffer overflow. Here, if the relay buffer size is greater than the buffer threshold of a given relay node of the user equipment UE, then, , (t) =0, otherwise Αζ„ _ο (ή =ι; >3⁄4 is a small fixed parameter. It is related to the user's rate. Usually, kT _ik ^ [ή is a constant of a decimal. The larger the value, the faster the adjustment, but the effect is poor; Conversely, the adjustment speed is slow, but the effect is good. T _h ( ) is the token counter of the user equipment UEJ _A at time t, and T _ik (k) is updated according to the following expression:

( - 1)表示在 t-l时刻在中继节点的一个资源分区上与用户设备 UE—R相对应的接入链路的平均速率。 T _ik (t) = max |θ, T _ik {t ~ l) + B _k - T _ aver!: (Bu 1) - Β ₀ · Τ - averj: (i)J (3) where Γ

(-1) indicates the average rate of access links corresponding to user equipment UE-R on one resource partition of the relay node at time t1.

In general, the average rate on the relay backhaul link / ⁶ and the direct link /, Γ is estimated by the eNB, while the average rate of the relay access link is estimated by the relay node it serves. After all resources have been allocated, the average rate 用户 of the user z' on the link at time t can be calculated by:

T_aver! (t) = 1 . T _inst _n (ή) (4)

Where r _PF is the proportional fair scheduling time window size (unit: second), Np is the number of resource partitions of its service node, and is the subframe length (unit: second). Is a scheduling indicator that is equal to 1 when the user's link/scheduled when allocating resources, and 0 otherwise.

 In the present invention, the token mechanism is used in the present invention to ensure that the minimum average transmission traffic of the backhaul link corresponding to the UE-R over the available bandwidth of the eNB is not less than the availability of its access link at its serving relay node. The minimum average transmission traffic over the bandwidth to avoid buffer underflow. Specifically, when the average rate of the relay backhaul link of a certain user is less than the relay access rate thereof, it can be known from the expressions (1)-(3) that the backhaul link allocation resource will be increased, thereby avoiding the relay buffer in the There is no data transmission situation on the relay access link.

效用函数即调度策略， 可以通过其来决定该调度哪个用户设备来使用资源。 在表 达式 （5 ) 中 ro)和 f (ί)分别是在 t时刻被调度的链路和被调度的用户设备或 中继节点。 r—ave^ )是通过低通滤波器平滑后的平均传输速度。 在分配完所 有资源后， 用户 在 t时刻在链路 /上的平均速率 Γ— ave^ ( )可以通过下式来计 算：

The utility function of the proportional fair scheduling scheme of the eNB according to the present invention is expressed as

The utility function is the scheduling policy, which can be used to determine which user equipment to schedule to use the resource. In the expression (5), ro) and f(ί) are the link scheduled at time t and the scheduled user equipment or relay node, respectively. R-ave^) is the average transmission speed smoothed by a low-pass filter. After all the resources have been allocated, the average rate 用户 ave^ ( ) of the user at link t at time t can be calculated by:

among them, ? _F is the proportional fair scheduling time window size (unit: second), Np is the number of resource partitions of its service node, and 7 is the subframe length. ( ) is a scheduling indicator that is equal to 1 when the resource is allocated when the user's link/scheduled, otherwise 0.

 Next, you can assign all resource partitions in turn.

 It can be seen from the expression (2) that when the buffer overflows, the user's backhaul link will not participate in resource allocation, so on the one hand, the buffer data overflow is avoided, and on the other hand, the macro UE can be more Scheduling opportunities.

 The backhaul link scaling factor as shown in Expression (2) ensures that no buffer overflow occurs by using a scheduling switch for buffer overflow conditions, and uses an exponential scaling factor for buffer underflow conditions. The occurrence of buffer underflow can be well avoided.

 Of course, this is just a concrete example. According to an embodiment of the invention, the expression of the selected scaling factor only needs to be such that the scheduling priority is reduced when the relay buffer size is greater than a buffering threshold (but no buffer overflow has occurred), while the average rate of the backhaul link is low. The scheduling priority is increased at the average rate of access links and is suitable for the purposes of the present invention.

 2 is a block diagram showing a system for implementing a scheduling method in accordance with an embodiment of the present invention.

As shown in FIG. 2, the relay node includes an access link average rate estimation module 201, configured to estimate an average rate of access links of each UE-R according to CQI information from the user equipment UE-R; a module 202, configured to monitor its own buffer status for each UE_R; and an information sending module 203. Used to feed back to the eNodeB its own buffer state information for each UEJR and the average rate of the estimated access link, and the CQI of its backhaul link.

 The eNodeB includes a scheduling information receiving module 101, configured to receive, from the relay node connected thereto and the served macro user equipment, information required for scheduling, including CQI of the backhaul link and CQI of the direct link of the macro user equipment. The buffering status information of the relay node and the average rate of the access link; the scheduling priority calculation module 102 is configured to estimate an average rate according to the CQI of the backhaul link and the CQI of the direct link, thereby determining a scheduling priority; The following node buffer information determining unit 103 is configured to determine, according to the buffer status information, whether the relay node has a buffer overflow condition, and according to the average rate of the estimated backhaul link and the average rate of the access link fed back by the relay node, Determining whether the relay node has a buffer underflow; the scheduling priority adjustment module 104 is configured to adjust the scheduling priority according to a predetermined rule, and reduce the scheduling priority of the backhaul link if it is determined that there is a buffer overflow. Increase the scheduling priority of the backhaul link if it is determined that there is a possibility of buffer underflow; The source scheduling module 105 schedules user equipments with higher priorities.

 In the above example, the predetermined rule is the scaling factor shown by the expression (2).

 FIG. 3 shows a flow chart of a scheduling method in accordance with an embodiment of the present invention.

 As shown in FIG. 3, at step S101, each macro user equipment feeds back its CQI to the base station. Meanwhile, at step S102, the user equipment served by the relay node feeds back the CQI to its serving relay node, and then at step S103, the relay node estimates the average rate of its access link based on the CQI fed back by the user equipment. At step S104, the relay node feeds back to the base station the average rate of the access link estimated for the user, and its CQI regarding the buffer status of the user, the backhaul link. At step S105, the base station estimates the average rate of the backhaul link of the user equipment according to the CQI of the backhaul link, and determines the scheduling priority. For macro user equipment, the base station estimates the average rate of its direct link based on its fed 'cQI' and determines its scheduling priority.

 For the macro user equipment, at step S106, the base station determines, according to the buffer status information, whether the relay node has a buffer overflow condition, and according to the average rate of the estimated backhaul link and the average rate of the access link fed back by the relay node. , determine if the relay node has buffer underflow.

 At step S107, the base station adjusts the scheduling priority according to a predetermined rule, reduces the scheduling priority of the backhaul link if it is determined that there is a buffer overflow, and increases the backhaul chain if it is determined that there is a possibility of buffer underflow. The scheduling priority of the road.

· At step S108, the base station has a higher scheduling priority according to the last determined scheduling priority The user equipment is scheduled.

 The downlink performance of the scheduling method considering the traffic balance on the relay node according to an embodiment of the present invention can be given by system level simulation. As a comparison, an existing scheme that does not consider the traffic balance on the relay node is also given. Here, Table 1 below shows the simulation parameters. Here, it is assumed that the semi-static resource allocation of the eNB and its cooperative relay node is associated with user access, and a hop-by-hop proportional fair scheduling scheme is employed. Table 1 System level simulation parameters

Fig. 4 is a view showing a comparison of simulation results of the scheduling method according to the present invention and the existing scheduling method under the simulation parameters shown in Table 1. In Fig. 4, system performance is shown as a cumulative score for normalized single-user throughput. Cloth function CDF. Table 2 below summarizes the results of the comparison. As can be seen from FIG. 4, the average cell user throughput of the scheduling scheme according to the present invention is increased by approximately 4.67% compared to the existing scheduling scheme that does not consider the traffic balance on the relay node. '

因此， 根据本发明的调度方法可灵活地应用于 LTE-A中继网络中， 以提高系 统性能。 Table 2 System performance comparison

Therefore, the scheduling method according to the present invention can be flexibly applied to an LTE-A relay network to improve system performance.

 Those skilled in the art will readily recognize that the different steps of the above methods can be implemented by a programmed computer. Herein, some embodiments also include a machine readable or computer readable program storage device (eg, a digital data storage medium) and encoding machine executable or computer executable program instructions, wherein the instructions perform some of the above methods or All steps. For example, the program storage device can be a digital memory, a magnetic storage medium (such as a disk and tape), a hardware or an optically readable digital data storage medium. The implementation also includes a programming computer that performs the steps of the above method.

 The description and drawings merely illustrate the principles of the invention. It will be appreciated that those skilled in the art are able to devise various structures, and the various structures are not described or illustrated herein, but are intended to be within the spirit and scope of the invention. In addition, all of the examples mentioned herein are explicitly used primarily for teaching purposes to assist the reader in understanding the principles of the present invention and the concepts promoted by the inventors, and should be construed as not to the specific examples. And conditional restrictions. In addition, all statements herein reciting principles, aspects, and embodiments of the invention, as well as specific examples thereof,

 The embodiments described above are for illustrative purposes only and are not intended to limit the scope of the invention. It will be apparent to those skilled in the art that various modifications and changes in the form and details may be made without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims and their equivalents.

Claims

Rights request A scheduling method comprising the steps of: receiving, from each relay node, a channel quality indicator of a backhaul link, buffering status information corresponding to each user equipment served by the relay node, and an access link Average rate  Estimating an average rate of the backhaul link according to a channel quality indicator of the backhaul link, and determining a scheduling priority of the backhaul link;  Determining whether the relay node has a buffer overflow according to the buffer status information, and determining whether the relay node has buffer underflow according to the average rate of the estimated back link and the average rate of the access link;  Adjusting the scheduling priority according to a predetermined rule, so that in the case where it is determined that there is a buffer overflow, the scheduling priority of the backhaul link is lowered, and in the case of determining that there is a possibility of buffer underflow, the scheduling priority of the backhaul link is increased. ; as well as  Schedule a backhaul link with a higher priority.  2. The scheduling method according to claim 1, wherein the step of determining whether the relay node has buffer underflow according to the average rate of the estimated backhaul link and the average rate of the access link comprises: in the backhaul chain When the average transmission traffic on the available bandwidth of the base station is smaller than the average transmission traffic of the access link on the available bandwidth of the relay node, it is determined that the relay node has buffer underflow.  3. The scheduling method according to claim 1, wherein the scheduling method is performed for each resource partition.  The scheduling method according to claim 1, wherein the predetermined rule comprises: adjusting the scheduling priority to zero in the case of determining that the buffer overflow occurs.  5. The scheduling method according to claim 1, wherein the predetermined rule comprises: averaging a scheduling priority and a backhaul link and an average of an access link in a case where it is determined that there is a possibility of buffer underflow The rate increases exponentially in association with each other.  The scheduling method according to claim 1, wherein the predetermined rule comprises: adjusting the scheduling priority without determining that there is no buffer overflow and there is no buffer underflow.  7. Method according to one of the claims 1-6, wherein the scheduling priority is determined according to a proportional fairness criterion or a polling criterion. 8. The method according to claims 1-6, wherein the relay received from each relay node The average rate of the access link corresponding to each user equipment served by the node is estimated by the relay node according to the channel quality indicator of the access link fed back by the user equipment.  9. A scheduler comprising:  a scheduling information receiving module, configured to receive scheduling required information from a relay node connected thereto, including a channel quality indicator of the backhaul link, buffer state information corresponding to each user equipment served by the relay node, and Average rate of access links;  a scheduling priority calculation module, configured to calculate a scheduling priority of the backhaul link according to a channel quality indicator of the backhaul link;  The relay node buffer information determining unit is configured to determine, according to the buffer status information, whether the relay node has a buffer overflow condition, and determine the relay node according to the estimated average rate of the backhaul link and the average rate of the access link. Whether there is a possibility of buffer underflow;  a scheduling priority adjustment module, configured to adjust a scheduling priority according to a predetermined rule, so that in a case where it is determined that there is a buffer overflow, the scheduling priority of the backhaul link is reduced, and in a case where it is determined that there is a possibility of buffer underflow, Increase the scheduling priority of the backhaul link;  A resource scheduling module is configured to schedule a backhaul link with a higher priority.  10. The scheduler according to claim 9, wherein the predetermined rule comprises: in a case where it is determined that the buffer overflow occurs, the scheduling priority adjustment module adjusts the scheduling priority to zero.  11. The scheduler according to claim 9, wherein the predetermined rule comprises: in a case where it is determined that there is a possibility of buffer underflow, the scheduling priority adjustment module makes the scheduling priority and the average rate of the backhaul link The average rate of incoming links increases exponentially in association. .  12. The scheduler according to claim 9, wherein the predetermined rule comprises: in a case where it is determined that there is no buffer overflow and there is no buffer underflow, the scheduling priority adjustment module does not adjust the scheduling priority.  13. The scheduler according to any one of claims 9-12, wherein the scheduling priority calculation module determines the scheduling priority according to a proportional fairness criterion or a polling criterion.