CN102547852A

CN102547852A - Packet scheduling method and device

Info

Publication number: CN102547852A
Application number: CN2010106040907A
Authority: CN
Inventors: 杨菊花; 张毅; 徐�明; 袁仟共; 区旭青; 郭婉云
Original assignee: China Mobile Group Guangdong Co Ltd
Current assignee: China Mobile Group Guangdong Co Ltd
Priority date: 2010-12-14
Filing date: 2010-12-14
Publication date: 2012-07-04

Abstract

The present invention provides a method and device for group scheduling, wherein the method includes: obtaining the service data throughput of the user, and estimating the average throughput of the user according to the service data throughput of the user; if the average throughput of the user is less than If the minimum throughput is lower than the minimum throughput, the scheduling priority of the user is increased so that the user can be served. The solution of the invention can increase the fairness among users in the system.

Description

A group scheduling method and device

技术领域 technical field

本发明涉及HSDPA(High Speed Downlink Package Access，高速下行链路分组接入)调度算法领域，特别是指一种兼顾用户公平和优先的分组调度方法及装置。The present invention relates to the field of HSDPA (High Speed Downlink Package Access, High Speed Downlink Packet Access) scheduling algorithm, in particular to a packet scheduling method and device that takes into account user fairness and priority.

背景技术 Background technique

在HSDPA系统中，由分组调度算法来实现无线资源的管理功能，研究先进的分组高度算法是提高数据业务吞吐量、保证用户间的公平性、满足业务Qos(Quality of Service，服务质量)的根本。In the HSDPA system, the wireless resource management function is realized by the packet scheduling algorithm, and the study of the advanced packet height algorithm is fundamental to improve the data service throughput, ensure the fairness among users, and satisfy the service Qos (Quality of Service) .

当前，比例公平调度算法(PF)是一种在无线移动网络中广为使用的调度算法，它可以为系统最大吞吐量和用户间的公平性提供很好的平衡，能够很好地满足非实时业务的要求。但是由于这种算法并没有考虑数据包的时延，因此，对于时延敏感的实时业务效果并不理想。Currently, Proportional Fair Scheduling Algorithm (PF) is a scheduling algorithm widely used in wireless mobile networks. It can provide a good balance between the maximum throughput of the system and the fairness among users, and can well meet the non-real-time business requirements. However, since this algorithm does not consider the time delay of data packets, it is not ideal for delay-sensitive real-time services.

针对此问题，现有技术中最大加权时延优先(M-LWDF)调度算法做了进一步改进，它广泛用于HSDPA、EVDO、4G等系统中。该算法引入了用户Qos参数和数据包时延，它的主要思想是将分组数据包的时延和如何有效利用信道信息平衡考虑进去，其用户优先级的计算不仅和用户当前的信道质量有关，还和包的队列时延有关。用户i在第n个TTI(Transmission Time Interval，传输时间间隔)的优先级计算公式如下：i＝1，2，......nTo solve this problem, the maximum weighted delay first (M-LWDF) scheduling algorithm in the prior art has been further improved, and it is widely used in HSDPA, EVDO, 4G and other systems. This algorithm introduces user QoS parameters and data packet delay. Its main idea is to take into account the delay of packet data packets and how to effectively use channel information balance. The calculation of user priority is not only related to the user's current channel quality, but also It is also related to the queue delay of the packet. The formula for calculating the priority of user i at the nth TTI (Transmission Time Interval, transmission time interval) is as follows: i=1, 2,...n

其中，δ代表用户的Qos参数，R[n]是用户在第n个TTI的最大数据速率，λ是用户i的平均吞吐量，D[n]是用户i所在包的队列延时，T是用户i能容忍的最大时延(丢弃时间参数)Among them, δ represents the Qos parameter of the user, R[n] is the maximum data rate of the user at the nth TTI, λ is the average throughput of user i, D[n] is the queue delay of the packet of user i, and T is The maximum delay that user i can tolerate (discard time parameter)

该算法对小区吞吐量和基站队列所产生的时延抖动做了有效折衷，是一种非公平算法，用户信道条件好时会有更好的Qos，一般有2-3秒的的队列延时，然而对信道条件差的用户来说，该算法会造成这些用户的数据包在基站侧有较大的时延，当时延超过用户的最大容忍时间时就会被抛弃。This algorithm makes an effective compromise between the cell throughput and the delay jitter generated by the base station queue. It is an unfair algorithm. When the user channel condition is good, there will be better Qos, and there is usually a queue delay of 2-3 seconds. , however, for users with poor channel conditions, this algorithm will cause the data packets of these users to have a large delay on the base station side, and when the delay exceeds the user's maximum tolerance time, it will be discarded.

M-LWDF算法是吞吐量较佳算法，它将不同用户分组数据包的时延和当前信道质量信息综合考虑，但M-LWDF算法未兼顾用户公平性，它的吞吐量性能的提升是以牺牲用户公平性为代价的，信道状态较差的用户的分组会在基站队列中等待，而一旦超过时间限制，这些分组将会丢掉而不再被服务。这样对用户的公平性较差。The M-LWDF algorithm is an algorithm with better throughput, which comprehensively considers the delay of different user packet data packets and the current channel quality information, but the M-LWDF algorithm does not take into account user fairness, and its throughput performance is improved at the expense of At the expense of user fairness, the packets of users with poor channel conditions will wait in the base station queue, and once the time limit is exceeded, these packets will be discarded and no longer be served. This is less fair to users.

发明内容 Contents of the invention

本发明要解决的技术问题是提供一种对用户较为公平的分组调度方法及装置。The technical problem to be solved by the present invention is to provide a group scheduling method and device that are fair to users.

为解决上述技术问题，本发明的实施例提供一种分组调度方法，包括：In order to solve the above technical problems, an embodiment of the present invention provides a packet scheduling method, including:

获取用户的业务数据吞吐量，并根据所述用户的业务数据吞吐量估计用户的平均吞吐量；Obtaining the user's business data throughput, and estimating the user's average throughput according to the user's business data throughput;

若所述用户的平均吞吐量小于最小吞吐量，则提高所述用户的调度优先级，使所述用户得到服务。If the average throughput of the user is less than the minimum throughput, the scheduling priority of the user is increased, so that the user is served.

其中，所述提高所述用户的调度优先级具体为：Wherein, the raising of the scheduling priority of the user is specifically:

通过公式：

提高所述用户的调度优先级；By formula:

increasing the scheduling priority of the user;

其中，P_i为被服务的所述用户i的优先级，j为拥有最高优先级的用户的优先级，K为整数。Wherein, P _i is the priority of the user i being served, j is the priority of the user with the highest priority, and K is an integer.

其中，上述分组调度方法还包括：Wherein, the above group scheduling method also includes:

对于实时业务，对在基站队列中等待的平均吞吐量小于最小吞吐量的至少一个所述用户进行优先级和数据吞吐量的实时双重补偿。For real-time services, real-time dual compensation of priority and data throughput is performed for at least one user whose average throughput waiting in the queue of the base station is less than the minimum throughput.

其中，对在基站队列中等待的平均吞吐量小于最小吞吐量的至少一个所述用户进行优先级和数据吞吐量的实时双重补偿具体为：Wherein, the real-time dual compensation of priority and data throughput for at least one user whose average throughput waiting in the base station queue is less than the minimum throughput is specifically:

通过公式： $P_{i} [n] = \{\begin{matrix} - \log (δ_{i}) \cdot \frac{R_{i} [n]}{λ_{i} [n]} \cdot \frac{D_{i} [n]}{T_{i}} \cdot L_{i} [n] & if λ_{i} &GreaterEqual; k \\ - \log (δ_{i}) \cdot \frac{R_{i} [n]}{λ_{i} [n]} \cdot \frac{D_{i} [n]}{T_{i}} \cdot L_{i} [n] \cdot Z_{i} [n] & if λ_{i} < k \end{matrix}$ By formula: $P_{i} [no] = \{\begin{matrix} - \log (δ_{i}) \cdot \frac{R_{i} [no]}{λ_{i} [no]} \cdot \frac{{D.}_{i} [no]}{T_{i}} &Center Dot; L_{i} [no] & if λ_{i} &Greater Equal; k \\ - \log (δ_{i}) \cdot \frac{R_{i} [no]}{λ_{i} [no]} &Center Dot; \frac{{D.}_{i} [no]}{T_{i}} \cdot L_{i} [no] &Center Dot; Z_{i} [no] & if λ_{i} < k \end{matrix}$

对在基站队列中等待的平均吞吐量小于最小吞吐量的至少一个所述用户进行优先级和数据吞吐量的实时双重补偿；performing real-time dual compensation of priority and data throughput for at least one user whose average throughput waiting in the base station queue is less than the minimum throughput;

其中，δ_i为用户i的Qos参数，R_i[n]为用户i在第n个传输时间间隔TTI的最大数据速率，λ_i[n]为用户i的平均吞吐量，D_i[n]为用户i所在包的队列延时，T_i为用户i能容忍的最大时延，k为最小吞吐量，

L_i[n]为用户i的待传输数据量，P_i[n]为用户i在第n个传输时间间隔TTI的优先级。其中，在每次分组调度开始时，统计当前L_i[n]，令

每完成一个子载波分配后，按如下公式更新λi_[n]和L_i[n]：Among them, δ _i is the Qos parameter of user i, R _i [n] is the maximum data rate of user i in the nth transmission time interval TTI, λ _i [n] is the average throughput of user i, D _i [n] is the queue delay of user i's packet, T _i is the maximum delay that user i can tolerate, k is the minimum throughput,

L _i [n] is the amount of data to be transmitted by user i, and P _i [n] is the priority of user i in the nth transmission time interval TTI. Among them, at the beginning of each packet scheduling, the current L _i [n] is counted, so that

After each subcarrier allocation is completed, update λi _[ n] and L _i [n] according to the following formula:

${λ λ}_{i i} [[n no]] = = {λ λ}_{i i} [[n no]] + + \frac{11}{{n no}_{r r}} \cdot &Center Dot; {ρ ρ}_{i i} \cdot &Center Dot; {R R}_{i i} [[n no]]$

L_i[n]＝L_i[n]-L_ii L _i [n] = L _i [n] - L _ii

其中，R_i[n]为当前用户i在载波上的最大传输速率，L_ii为用户i在载波上实际传输的数据量，当载波上发送用户i的数据时，ρ_i＝1，否则ρ_i＝0。Among them, R _i [n] is the maximum transmission rate of current user i on the carrier, L _ii is the amount of data actually transmitted by user i on the carrier, when the data of user i is sent on the carrier, ρ _i =1, otherwise ρ _i =0.

本发明的实施例还提供一种分组调度装置，包括：Embodiments of the present invention also provide a packet scheduling device, including:

第一处理模块，用于获取用户的业务数据吞吐量，并根据所述用户的业务数据吞吐量估计用户的平均吞吐量；The first processing module is used to obtain the user's business data throughput, and estimate the user's average throughput according to the user's business data throughput;

第二处理模块，用于若所述用户的平均吞吐量小于最小吞吐量，则提高所述用户的调度优先级，使所述用户得到服务。The second processing module is configured to increase the scheduling priority of the user if the average throughput of the user is less than the minimum throughput, so that the user can be served.

其中，所述第二处理模块具体通过公式：

提高所述用户的调度优先级；Wherein, the second processing module specifically adopts the formula:

increasing the scheduling priority of the user;

其中，上述分组调度装置还包括：Wherein, the above-mentioned packet scheduling device also includes:

第三处理模块，用于对于实时业务，对在基站队列中等待的平均吞吐量小于最小吞吐量的至少一个所述用户进行优先级和数据吞吐量的实时双重补偿。The third processing module is configured to perform real-time double compensation of priority and data throughput for at least one user whose average throughput waiting in the queue of the base station is less than the minimum throughput for real-time services.

其中，所述第三处理模块具体通过公式：Wherein, the third processing module specifically adopts the formula:

${P P}_{i i} [[n no]] = = \{\begin{matrix} - - log log (({δ δ}_{i i})) \cdot &Center Dot; \frac{{R R}_{i i} [[n no]]}{{λ λ}_{i i} [[n no]]} \cdot &Center Dot; \frac{{D D.}_{i i} [[n no]]}{{T T}_{i i}} \cdot &Center Dot; {L L}_{i i} [[n no]] & if if {λ λ}_{i i} &GreaterEqual; &Greater Equal; k k \\ - - log log (({δ δ}_{i i})) \cdot &Center Dot; \frac{{R R}_{i i} [[n no]]}{{λ λ}_{i i} [[n no]]} \cdot &Center Dot; \frac{{D D.}_{i i} [[n no]]}{{T T}_{i i}} \cdot &Center Dot; {L L}_{i i} [[n no]] \cdot &Center Dot; {Z Z}_{i i} [[n no]] & if if {λ λ}_{i i} < < k k \end{matrix}$

为用户i的待传输数据量，P_i[n]为用户i在第n个传输时间间隔TTI的优先级。Among them, δ _i is the Qos parameter of user i, R _i [n] is the maximum data rate of user i in the nth transmission time interval TTI, λ _i [n] is the average throughput of user i, D _i [n] is the queue delay of user i's packet, T _i is the maximum delay that user i can tolerate, k is the minimum throughput,

is the amount of data to be transmitted by user i, and P _i [n] is the priority of user i in the nth transmission time interval TTI.

其中，所述第三处理模块进一步用于：在每次分组调度开始时，统计当前L_i[n]，令

Wherein, the third processing module is further configured to: at the beginning of each packet scheduling, count the current L _i [n], so that

每完成一个子载波分配后，按如下公式更新λ_i[n]和L_i[n]：After each subcarrier allocation is completed, λ _i [n] and L _i [n] are updated according to the following formula:

${λ λ}_{i i} [[n no]] = = {λ λ}_{i i} [[n no]] + + \frac{11}{{n no}_{r r}} \cdot \cdot {ρ ρ}_{i i} \cdot \cdot {R R}_{i i} [[n no]]$

L_i[n]＝L_i[n]-L_ii L _i [n] = L _i [n] - L _ii

本发明的上述技术方案的有益效果如下：The beneficial effects of above-mentioned technical scheme of the present invention are as follows:

上述方案中，通过估计用户的平均吞吐量，并在用户的平均吞吐量小于最小吞吐量时，则提高所述用户的调度优先级，使所述用户得到服务；本方法考虑到了用户最小吞吐量保证，以便更好地衡量用户队列的负载情况和提高低吞吐量用户的优先级，增加了系统中用户之间的公平性。In the above scheme, by estimating the average throughput of the user, and when the average throughput of the user is less than the minimum throughput, the scheduling priority of the user is increased so that the user can be served; this method takes into account the minimum throughput of the user Guaranteed to better measure the load on user queues and improve the priority of low-throughput users, increasing the fairness among users in the system.

附图说明 Description of drawings

图1为本发明的分组调度方法第一实施例流程图；Fig. 1 is the flowchart of the first embodiment of the packet scheduling method of the present invention;

图2为本发明的分组调度方法第二实施例流程图。Fig. 2 is a flow chart of the second embodiment of the packet scheduling method of the present invention.

具体实施方式 Detailed ways

为使本发明要解决的技术问题、技术方案和优点更加清楚，下面将结合附图及具体实施例进行详细描述。In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will describe in detail with reference to the drawings and specific embodiments.

本发明针对现有调度算法对用户的公平性较差问题，提供一种对用户较为公平的分组调度方法及装置。Aiming at the problem that the existing scheduling algorithm has poor fairness to users, the invention provides a grouping scheduling method and device which are relatively fair to users.

如图1所示，本发明的分组调度方法，包括：As shown in Figure 1, the packet scheduling method of the present invention includes:

步骤11，获取用户的业务数据吞吐量，并根据所述用户的业务数据吞吐量估计用户的平均吞吐量；Step 11, obtaining the service data throughput of the user, and estimating the average throughput of the user according to the service data throughput of the user;

步骤12，若所述用户的平均吞吐量小于最小吞吐量，则提高所述用户的调度优先级，使所述用户得到服务。Step 12, if the average throughput of the user is less than the minimum throughput, increase the scheduling priority of the user so that the user can be served.

本发明的上述方案通过估计用户的平均吞吐量，并在用户的平均吞吐量小于最小吞吐量时，则提高所述用户的调度优先级，使所述用户得到服务；本方法考虑到了用户最小吞吐量保证，以便更好地衡量用户队列的负载情况和提高低吞吐量用户的优先级，增加了系统中用户之间的公平性。The above scheme of the present invention estimates the average throughput of the user, and when the average throughput of the user is less than the minimum throughput, then increases the scheduling priority of the user, so that the user is served; this method takes into account the minimum throughput of the user Capacity guarantee, in order to better measure the load of user queues and improve the priority of low-throughput users, increasing the fairness among users in the system.

其中，为了降低信道质量差的用户的数据包被丢弃的概率，提高用户的公平性，本实施例中采用了新的优先级公式对所述用户的优先级进行提高，对吞吐量较低的用户进行了补偿；具体来讲，上述步骤12中，提高所述用户的调度优先级具体可以为：Among them, in order to reduce the probability that the data packets of users with poor channel quality are discarded and improve the fairness of users, a new priority formula is used in this embodiment to improve the priority of the users, and the user with lower throughput The user has compensated; specifically, in the above step 12, increasing the scheduling priority of the user can specifically be:

通过公式：

提高所述用户的调度优先级；By formula:

increasing the scheduling priority of the user;

其中，P_i为被服务的所述用户i的优先级，j为拥有最高优先级的用户的优先级，K为整数；Wherein, _Pi is the priority of the user i being served, j is the priority of the user with the highest priority, and K is an integer;

最小吞吐量保证是QoS(服务质量)要求的重要组成部分，如果每个用户都有最小吞吐量保证，那么所有满足此条件的用户将被服务；对不满足此条件的用户，可以通过改变调度规则(如采用上述公式提高用户的调度优先级)，使其调度优先级得到提高，以保证低吞吐量用户的公平性。The minimum throughput guarantee is an important part of QoS (Quality of Service) requirements. If each user has a minimum throughput guarantee, then all users who meet this condition will be served; for users who do not meet this condition, you can change the scheduling Rules (such as using the above formula to increase the user's scheduling priority), so that its scheduling priority is increased to ensure the fairness of low-throughput users.

另外，在上述图1所示调度方法的实施例的基础上，如图2所示，上述调度方法还可以进一步包括：In addition, on the basis of the embodiment of the scheduling method shown in Figure 1 above, as shown in Figure 2, the above scheduling method may further include:

步骤13，对于实时业务，对在基站队列中等待的平均吞吐量小于最小吞吐量的至少一个所述用户进行优先级和数据吞吐量的实时双重补偿；Step 13, for real-time services, perform real-time double compensation of priority and data throughput for at least one user whose average throughput waiting in the base station queue is less than the minimum throughput;

这样可以降低Node B侧数据流包被丢弃的概率，保证用户公平性，在保证用户分组的最小吞吐量的条件下最大化实时业务的用户数，具体来讲，可采用下述方式对信道条件差的用户进行优先级和数据吞吐量的实时双重补偿：This can reduce the probability of data flow packets being discarded on the Node B side, ensure user fairness, and maximize the number of real-time service users under the condition of ensuring the minimum throughput of user groups. Specifically, the channel conditions can be adjusted in the following way Poor users perform real-time dual compensation of priority and data throughput:

即对在基站队列中等待的平均吞吐量小于最小吞吐量的至少一个所述用户进行优先级和数据吞吐量的实时双重补偿具体为：That is, the real-time dual compensation of priority and data throughput for at least one user whose average throughput waiting in the base station queue is less than the minimum throughput is specifically:

通过公式： $P_{i} [n] = \{\begin{matrix} - \log (δ_{i}) \cdot \frac{R_{i} [n]}{λ_{i} [n]} \cdot \frac{D_{i} [n]}{T_{i}} \cdot L_{i} [n] & if λ_{i} &GreaterEqual; k \\ - \log (δ_{i}) \cdot \frac{R_{i} [n]}{λ_{i} [n]} \cdot \frac{D_{i} [n]}{T_{i}} \cdot L_{i} [n] \cdot Z_{i} [n] & if λ_{i} < k \end{matrix}$ By formula: $P_{i} [no] = \{\begin{matrix} - \log (δ_{i}) &Center Dot; \frac{R_{i} [no]}{λ_{i} [no]} &Center Dot; \frac{{D.}_{i} [no]}{T_{i}} &Center Dot; L_{i} [no] & if λ_{i} &Greater Equal; k \\ - \log (δ_{i}) \cdot \frac{R_{i} [no]}{λ_{i} [no]} &Center Dot; \frac{{D.}_{i} [no]}{T_{i}} &Center Dot; L_{i} [no] &Center Dot; Z_{i} [no] & if λ_{i} < k \end{matrix}$

其中，δ_i为用户i的Qos参数，R_i[n]为用户i在第n个TTI的最大数据速率，λ_i[n]为用户i的平均吞吐量，D_i[n]为用户i所在包的队列延时，T_i为用户i能容忍的最大时延，k为最小吞吐量，L_i[n]为用户i的待传输数据量，P_i[n]为用户i在第n个传输时间间隔TTI的优先级。Among them, δ _i is the Qos parameter of user i, R _i [n] is the maximum data rate of user i in the nth TTI, λ _i [n] is the average throughput of user i, D _i [n] is the The queue delay of the package, T _i is the maximum delay that user i can tolerate, k is the minimum throughput, L _i [n] is the amount of data to be transmitted by user i, and P _i [n] is the priority of user i in the nth transmission time interval TTI.

在上述公式中，λ_i[n]和L_i[n]是关于用户和队列的统计量，需要进行及时的更新，即更新是在每个子载波资源完成分配后进行的；与原有的在每次分组调度结束后进行的更新方法相比，实时更新的方法时效性更强，能够更好地反映用户被服务程度以及队列信息的真实状态，并且在分组调度时具有更强的灵活性；In the above formula, λ _i [n] and L _i [n] are statistics about users and queues, which need to be updated in time, that is, the update is performed after each subcarrier resource is allocated; Compared with the update method performed after each group scheduling, the real-time update method is more time-sensitive, can better reflect the user's service level and the real status of the queue information, and has stronger flexibility in group scheduling;

具体来讲，在每次分组调度开始时，统计当前L_i[n]，令 Specifically, at the beginning of each packet scheduling, the current L _i [n] is counted, so that

${λ λ}_{i i} [[n no]] = = {λ λ}_{i i} [[n no]] + + \frac{11}{{n no}_{r r}} \cdot \cdot {ρ ρ}_{i i} \cdot &Center Dot; {R R}_{i i} [[n no]]$

L_i[n]＝L_i[n]-L_ii L _i [n] = L _i [n] - L _ii

该实时更新方法同时引入2个新的参量：用户i的待传输数据量L，更好的衡量了用户队列的负载情况，且采用实时更新状态参数的方法，具有更强的调度灵活性。对于平均吞吐量开始低于最小吞吐量的用户，通过引入一个权值因子Z来提高低吞吐量用户的优先级，增加了系统中用户之间的公平性。The real-time update method introduces two new parameters at the same time: the amount of data L to be transmitted by user i, which better measures the load of the user queue, and adopts the method of real-time update status parameters, which has stronger scheduling flexibility. For users whose average throughput starts to be lower than the minimum throughput, a weight factor Z is introduced to increase the priority of low-throughput users, which increases the fairness among users in the system.

对于实时业务，本发明的该新方法不仅考虑到信道条件和用户时延的影响，还考虑到了用户最小吞吐量保证，并且算法引入了新的衡量服务队列负载的信息，采用了实时更新状态参数的方法，具有更强的调度灵活性，以便更好地衡量用户队列的负载情况和提高低吞吐量用户的优先级，增加了系统中用户之间的公平性，有利于提高系统的整体性能。For real-time services, the new method of the present invention not only takes into account the influence of channel conditions and user delays, but also considers the minimum throughput guarantee of users, and the algorithm introduces new information to measure the load of service queues, and adopts real-time update status parameters The method has stronger scheduling flexibility, so as to better measure the load of user queues and improve the priority of low-throughput users, which increases the fairness among users in the system and is conducive to improving the overall performance of the system.

与上述图1所示方法相应的，本发明的方案还提供一种分组调度装置，包括：Corresponding to the above method shown in Figure 1, the solution of the present invention also provides a packet scheduling device, including:

其中，所述第二处理模块具体通过公式：提高所述用户的调度优先级；Wherein, the second processing module specifically adopts the formula: increasing the scheduling priority of the user;

进一步的，上述调度装置还包括：Further, the above scheduling device also includes:

具体地，所述第三处理模块具体通过公式：Specifically, the third processing module specifically uses the formula:

${P P}_{i i} [[n no]] = = \{\begin{matrix} - - log log (({δ δ}_{i i})) \cdot &Center Dot; \frac{{R R}_{i i} [[n no]]}{{λ λ}_{i i} [[n no]]} \cdot \cdot \frac{{D D.}_{i i} [[n no]]}{{T T}_{i i}} \cdot \cdot {L L}_{i i} [[n no]] & if if {λ λ}_{i i} &GreaterEqual; &Greater Equal; k k \\ - - log log (({δ δ}_{i i})) \cdot &Center Dot; \frac{{R R}_{i i} [[n no]]}{{λ λ}_{i i} [[n no]]} \cdot &Center Dot; \frac{{D D.}_{i i} [[n no]]}{{T T}_{i i}} \cdot &Center Dot; {L L}_{i i} [[n no]] \cdot &Center Dot; {Z Z}_{i i} [[n no]] & if if {λ λ}_{i i} < < k k \end{matrix}$

对在基站队列中等待的平均吞吐量小于最小吞吐量的至少一个所述用户进行优先级和数据吞吐量的实时双重补偿；performing real-time dual compensation of priority and data throughput for at least one of the users whose average throughput waiting in the base station queue is less than the minimum throughput;

L_i[n]为用户i的待传输数据量，P_i[n]为用户i在第n个传输时间间隔TTI的优先级。Among them, δ _i is the Qos parameter of user i, R _i [n] is the maximum data rate of user i in the nth transmission time interval TTI, λ _i [n] is the average throughput of user i, D _i [n] is the queue delay of user i's packet, T _i is the maximum delay that user i can tolerate, k is the minimum throughput,

L _i [n] is the amount of data to be transmitted by user i, and P _i [n] is the priority of user i in the nth transmission time interval TTI.

其中，所述第三处理模块进一步用于：在每次分组调度开始时，统计当前

Wherein, the third processing module is further used to: at the beginning of each packet scheduling, count the current

${λ λ}_{i i} [[n no]] = = {λ λ}_{i i} [[n no]] + + \frac{11}{{n no}_{r r}} \cdot &Center Dot; {ρ ρ}_{i i} \cdot \cdot {R R}_{i i} [[n no]]$

L_i[n]＝L_i[n]-L_ii L _i [n] = L _i [n] - L _ii

需要说明的是：本发明的上述方法中所有具体实现方案均适用于该装置实施例中相应的模块，也能达到相同的技术效果，且本发明的上述调度方法和装置可广泛用于HSDPA、EVDO、4G等系统中。It should be noted that all specific implementation schemes in the above-mentioned method of the present invention are applicable to the corresponding modules in the device embodiment, and can also achieve the same technical effect, and the above-mentioned scheduling method and device of the present invention can be widely used in HSDPA, In EVDO, 4G and other systems.

以上所述是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明所述原理的前提下，还可以作出若干改进和润饰，这些改进和润饰也应视为本发明的保护范围。The above description is a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications can also be made. It should be regarded as the protection scope of the present invention.

Claims

1. A grouping scheduling method, characterized in that, comprising:

Obtaining the user's business data throughput, and estimating the user's average throughput according to the user's business data throughput;

If the average throughput of the user is less than the minimum throughput, the scheduling priority of the user is increased, so that the user is served.

2. The group scheduling method according to claim 1, wherein said improving the scheduling priority of said user is specifically:

By formula:

increasing the scheduling priority of the user;

Wherein, P _i is the priority of the user i being served, j is the priority of the user with the highest priority, and K is an integer.

3. The packet scheduling method according to claim 1 or 2, further comprising:

For real-time services, real-time dual compensation of priority and data throughput is performed for at least one user whose average throughput waiting in the queue of the base station is less than the minimum throughput.

4. The grouping scheduling method according to claim 3, wherein the real-time double compensation of priority and data throughput is carried out to at least one user whose average throughput waiting in the base station queue is less than the minimum throughput is specifically: :

By formula:

P_{i} [no] = \{\begin{matrix} - \log (δ_{i}) &Center Dot; \frac{R_{i} [no]}{λ_{i} [no]} &Center Dot; \frac{{D.}_{i} [no]}{T_{i}} &Center Dot; L_{i} [no] & if λ_{i} &Greater Equal; k \\ - \log (δ_{i}) \cdot \frac{R_{i} [no]}{λ_{i} [no]} &Center Dot; \frac{{D.}_{i} [no]}{T_{i}} &Center Dot; L_{i} [no] \cdot Z_{i} [no] & if λ_{i} < k \end{matrix}

performing real-time dual compensation of priority and data throughput for at least one of the users whose average throughput waiting in the base station queue is less than the minimum throughput;

Among them, δ _i is the Qos parameter of user i, R _i [n] is the maximum data rate of user i in the nth transmission time interval TTI, λ _i [n] is the average throughput of user i, D _i [n] is the queue delay of user i's packet, T _i is the maximum delay that user i can tolerate, k is the minimum throughput,

5. The packet scheduling method according to claim 4, characterized in that, at the beginning of each packet scheduling, the current L _i [n] is counted, so that

After each subcarrier allocation is completed, λ _i [n] and L _i [n] are updated according to the following formula:

{λ λ}_{i i} [[n no]] = = {λ λ}_{i i} [[n no]] + + \frac{11}{{n no}_{r r}} \cdot &Center Dot; {ρ ρ}_{i i} \cdot \cdot {R R}_{i i} [[n no]]

L _i [n] = L _i [n] - L _ii

Among them, R _i [n] is the maximum transmission rate of current user i on the carrier, Li _i is the amount of data actually transmitted by user i on the carrier, when the data of user i is sent on the carrier, ρ _i =1, otherwise ρ _i =0.

6. A packet scheduling device, characterized in that, comprising:

The first processing module is used to obtain the user's business data throughput, and estimate the user's average throughput according to the user's business data throughput;

The second processing module is configured to increase the scheduling priority of the user if the average throughput of the user is less than the minimum throughput, so that the user can be served.

7. The packet scheduling device according to claim 6, wherein the second processing module specifically uses the formula:

increasing the scheduling priority of the user;

8. The packet scheduling device according to claim 6 or 7, further comprising:

The third processing module is configured to perform real-time double compensation of priority and data throughput for at least one user whose average throughput waiting in the queue of the base station is less than the minimum throughput for real-time services.

9. The packet scheduling device according to claim 8, wherein the third processing module specifically uses the formula:

{P P}_{i i} [[n no]] = = \{\begin{matrix} - - log log (({δ δ}_{i i})) \cdot \cdot \frac{{R R}_{i i} [[n no]]}{{λ λ}_{i i} [[n no]]} \cdot \cdot \frac{{D D.}_{i i} [[n no]]}{{T T}_{i i}} \cdot &Center Dot; {L L}_{i i} [[n no]] & if if {λ λ}_{i i} &GreaterEqual; &Greater Equal; k k \\ - - log log (({δ δ}_{i i})) \cdot &Center Dot; \frac{{R R}_{i i} [[n no]]}{{λ λ}_{i i} [[n no]]} \cdot &Center Dot; \frac{{D D.}_{i i} [[n no]]}{{T T}_{i i}} \cdot &Center Dot; {L L}_{i i} [[n no]] \cdot &Center Dot; {Z Z}_{i i} [[n no]] & if if {λ λ}_{i i} < < k k \end{matrix}

performing real-time dual compensation of priority and data throughput for at least one user whose average throughput waiting in the base station queue is less than the minimum throughput;

10. The packet scheduling device according to claim 9, wherein the third processing module is further configured to: at the beginning of each packet scheduling, count the current L _i [n], so that

{λ λ}_{i i} [[n no]] = = {λ λ}_{i i} [[n no]] + + \frac{11}{{n no}_{r r}} \cdot &Center Dot; {ρ ρ}_{i i} \cdot &Center Dot; {R R}_{i i} [[n no]]

L _i [n] = L _i [n] - L _ii

Among them, R _i [n] is the maximum transmission rate of current user i on the carrier, L _ii is the amount of data actually transmitted by user i on the carrier, when the data of user i is sent on the carrier, ρ _i =1, otherwise ρ _i =0.