WO2018119830A1

WO2018119830A1 - Method and system for constructing task processing path

Info

Publication number: WO2018119830A1
Application number: PCT/CN2016/112919
Authority: WO
Inventors: 田霖; 卓蕊潋; 周一青; 石晶林
Original assignee: 中国科学院计算技术研究所
Priority date: 2016-12-29
Filing date: 2016-12-29
Publication date: 2018-07-05

Abstract

Disclosed is a method for constructing a task processing path in a plurality of processing resource pools. The method comprises: determining a task processing path based on a task requirement, wherein the task processing path includes a processing unit selected from a plurality of resource pools; searching for a destination address corresponding to the processing unit based on the task processing path; and sending a notification message to the processing unit, wherein the notification message includes a destination address to be sent by a data flow of the processing unit. The task processing path can be effectively constructed as needed, so that the configuration of a switching path for transmitting a data flow is completed.

Description

构建任务处理路径的方法及***Method and system for constructing task processing path

技术领域Technical field

本发明涉及通信技术领域，尤其涉及一种构建任务处理路径的方法和***。The present invention relates to the field of communications technologies, and in particular, to a method and system for constructing a task processing path.

背景技术Background technique

传统的蜂窝网络采用垂直式的网络架构模型，小区的资源采用垂直独立的资源分配方式，如图1a所示，在GSM、TDSCDMA、LTE的基站中，分别将协议处理、基带处理、射频处理组成小区，并通常以小区的峰值负载需求作为小区的容量指标来分配资源。在这种架构下，不同基站之间不能共享处理资源，并且各小区的实时负载在时间上和空间上具有巨大的不均匀性。空闲的网络资源往往不能为其他基站所用，造成网络资源利用率低，只能通过增加基站数目，密集部署基站来应对日益增长的移动数据量的需求。The traditional cellular network adopts a vertical network architecture model, and the resources of the cell adopt a vertical independent resource allocation manner. As shown in FIG. 1a, in the GSM, TDSCDMA, and LTE base stations, protocol processing, baseband processing, and radio frequency processing are respectively performed. The cell, and usually allocates resources as a capacity indicator of the cell with the peak load demand of the cell. Under this architecture, processing resources cannot be shared between different base stations, and the real-time load of each cell has great non-uniformity in time and space. Idle network resources are often not used by other base stations, resulting in low network resource utilization. Only by increasing the number of base stations and densely deploying base stations to cope with the increasing demand for mobile data.

为了提高资源利用率，节约移动网络运营成本，人们开始研究水平共享式网络架构，如图1b所示，在该架构下，采取资源水平池化的概念，将各种类型的基站资源构成处理资源池，接受统一的资源管控，进行灵活的调配。例如，可以采用多模基站的通用架构将GSM、TDSCDMA、LTE中的资源作为统一的资源池，如形成射频处理资源池(射频处理1、射频处理2和射频处理3)、基带处理资源池和协议处理资源池，以通过动态组合射频处理资源、基带处理资源和协议处理资源形成小区，完成对目标区域的覆盖。通过这种方式，可以提高网络资源的利用率，并且只需要对相应的处理单元进行配置加载或升级定制，就可以重用现有的基础设施，从而满足网络升级和发展的需求。In order to improve resource utilization and save mobile network operation costs, people began to study the horizontal shared network architecture, as shown in Figure 1b. Under this architecture, the concept of resource level pooling is adopted to construct various types of base station resources into processing resources. Pool, accept unified resource management and control, and carry out flexible deployment. For example, the general architecture of the multi-mode base station can be used as a unified resource pool in the GSM, TDSCDMA, and LTE resources, such as forming a radio frequency processing resource pool (RF processing 1, RF processing 2, and RF processing 3), a baseband processing resource pool, and The protocol processes the resource pool to form a cell by dynamically combining radio frequency processing resources, baseband processing resources, and protocol processing resources to complete coverage of the target area. In this way, the utilization of network resources can be improved, and only the configuration processing load or upgrade customization of the corresponding processing unit can be used, and the existing infrastructure can be reused to meet the network upgrade and development requirements.

然而，在目前的水平共享式的网络架构中，资源池中各处理单元之间一般采用较为成熟的互连网络设计，如图2所示，基带处理资源池中的DSP处理单元之间、不同机箱之间通过串行RapidIO(sRIO)实现互联，射频处理资源池与基带处理资源池之间通过CPRI(通用公共无线电接口)协议实现互联，基带处理资源池与协议处理资源池之间通过光纤或者以太网进行互联。在此示例中，如果任务处理的路径(或称数据流的传输路径)是从射频处理单元2(简称RRU2)到基带处理单元1(简称BBU1)，再到协议处理单元1(简称PPU1)，则中间需要经过CPRI交换机、sRIO交换机和以太网交换机。为了完成数据流的传输，需要配置处理单元之间的数据交换路径。例如，在现有技术中，通常是通过RRU2的配置接口，手动为RRU2配置BBU1的地址；通过BBU1的配置接口，为BBU1配置PPU1的地址，然后，RRU2在数据包生成后填充BBU1的地址为目的地址，BBU1在数据报生成后填充PPU1的地址为目的地址。However, in the current horizontal shared network architecture, a relatively mature interconnection network design is generally adopted between the processing units in the resource pool, as shown in FIG. 2, the DSP processing units in the baseband processing resource pool are different. The chassis is interconnected by serial RapidIO (sRIO). The RF processing resource pool and the baseband processing resource pool are interconnected by the CPRI (Common Public Radio Interface) protocol. The baseband processing resource pool and the protocol processing resource pool are connected through the optical fiber or the Ethernet. The network is interconnected. In this example, if the path processed by the task (or the transmission path of the data stream) is from the radio frequency processing unit 2 (referred to as RRU2) to the baseband processing unit 1 (abbreviated as BBU1), to the protocol processing unit 1 (referred to as PPU1), Then you need to go through the CPRI switch, sRIO switch and Ethernet switch. In order to complete the transmission of the data stream, it is necessary to configure a data exchange path between the processing units. For example, in the prior art, the address of the BBU1 is manually configured for the RRU2 through the configuration interface of the RRU2; the address of the PPU1 is configured for the BBU1 through the configuration interface of the BBU1, and then the address of the BBU1 is filled in after the data packet is generated by the RRU2. The destination address, BBU1 fills the address of PPU1 as the destination address after the datagram is generated.

因此，现有技术中的这种方法是针对各个互联网络分别配置连接关系，任务处理路径中各处理单元的地址需要分别获取。其存在的主要问题是：一是地址的获取时间可能不同步，从而可导致任务处理路径初期的建立错误；二是由于各个交换机的互联和交换机制不同，如图2所示，RRU2与BBU1之间的交换机制为CPRI交换，BBU1与PPU1之间的交换机制为RapidIO交换，其采用不同的地址定义，使得配置操作复杂，实现困难；三是当任务处理路径改变时，需要重新手动进行配置。Therefore, the method in the prior art is to separately configure a connection relationship for each interconnection network, and the addresses of the processing units in the task processing path need to be separately acquired. The main problems are: First, the acquisition time of the address may not be synchronized, which may lead to the establishment error of the initial processing path; Second, due to the interconnection and switch mechanism of each switch, as shown in Figure 2, RRU2 and BBU1 The exchange mechanism between the two switches is CPRI exchange. The exchange mechanism between BBU1 and PPU1 is RapidIO exchange. It uses different address definitions, which makes the configuration operation complicated and difficult to implement. Third, when the task processing path changes, it needs to be manually configured again.

发明内容Summary of the invention

本发明的目的在于克服上述现有技术的缺陷，提供一种能够灵活动态地构建任务处理路径的方法。It is an object of the present invention to overcome the above-discussed deficiencies of the prior art and to provide a method for flexibly and dynamically constructing a task processing path.

根据本发明的第一方面，提供了一种用于从多个资源池中构建任务处理路径的方法。该方法包括：According to a first aspect of the present invention, a method for constructing a task processing path from a plurality of resource pools is provided. The method includes:

步骤1：基于任务需求确定任务处理路径，其中，所述任务处理路径包含从所述多个资源池中选择的处理单元；Step 1: determining a task processing path based on a task requirement, where the task processing path includes a processing unit selected from the plurality of resource pools;

步骤2：基于所述任务处理路径查找所述处理单元的对应的目的地址；Step 2: Search for a corresponding destination address of the processing unit based on the task processing path.

步骤3：向所述处理单元发送通知消息，其中所述通知消息中包含所述处理单元的数据流将要发送的目的地址。Step 3: Send a notification message to the processing unit, where the notification message includes a destination address to be sent by the data stream of the processing unit.

在一个实施例中，该方法还包括从所述处理单元接收反馈消息。In one embodiment, the method also includes receiving a feedback message from the processing unit.

在一个实施例中，所述反馈消息包括下列项中的至少一项：目的地址是否配置成功的标志、处理单元的状态、处理单元的资源利用率、处理单元的地址变化。In one embodiment, the feedback message includes at least one of the following: a flag indicating whether the destination address is successfully configured, a status of the processing unit, a resource utilization of the processing unit, and an address change of the processing unit.

在一个实施例中，该方法还包括基于下列项中的至少一项来改变所述任务处理路径：处理单元的资源利用率；处理单元的状态；网络负载；任务需求的变化。In one embodiment, the method further comprises changing the task processing path based on at least one of: resource utilization of the processing unit; state of the processing unit; network load; Changes in demand.

在一个实施例中，所述处理单元的类型包括处理板、DSP、CPU、内存、硬盘、虚拟内存。In one embodiment, the types of processing units include a processing board, a DSP, a CPU, a memory, a hard disk, and a virtual memory.

在一个实施例中，所述任务需求包括任务优先级、带宽、时延、吞吐量、载波数量中一项或多项。In one embodiment, the task requirements include one or more of task priority, bandwidth, latency, throughput, number of carriers.

在一个实施例中，在步骤2中，通过预先配置的各个处理单元的地址表来查找所述处理单元的对应的目的地址。In one embodiment, in step 2, the corresponding destination address of the processing unit is looked up by a pre-configured address table of each processing unit.

根据本发明的第二方面，提供了一种用于从多个处理资源池中构建任务处理路径的***。该***包括：用于基于任务需求确定任务处理路径的装置，其中，所述任务处理路径包含从所述多个处理资源池中选择的处理单元；用于基于所述任务处理路径查找所述处理单元的对应的目的地址的装置；用于向所述处理单元发送通知消息的装置，其中所述通知消息中包含所述处理单元的数据流将要发送的目的地址。According to a second aspect of the present invention, a system for constructing a task processing path from a plurality of processing resource pools is provided. The system includes: means for determining a task processing path based on a task requirement, wherein the task processing path includes a processing unit selected from the plurality of processing resource pools; and wherein the processing is performed based on the task processing path Means for a corresponding destination address of the unit; means for transmitting a notification message to the processing unit, wherein the notification message includes a destination address to be sent by the data stream of the processing unit.

在一个实施例中，该***还包括用于配置或生成所述任务需求的装置。In one embodiment, the system further includes means for configuring or generating the task requirements.

在一个实施例中，该***还包括管理代理装置，用于向所述处理单元统一发送通知消息和/或为所述处理单元的数据流填充目的地址。In one embodiment, the system further includes management agent means for uniformly transmitting the notification message to the processing unit and/or populating the data stream of the processing unit with a destination address.

与现有技术相比，本发明的优点在于：通过统一配置各处理单元的目的地址，能够按需、高效地构建任务处理路径；通过动态的组合资源池中的处理单元，能够完成特定功能的处理，并灵活的扩展新功能。Compared with the prior art, the present invention has the advantages that the task processing path can be constructed on demand and efficiently by uniformly configuring the destination addresses of the processing units; the specific functions can be completed by dynamically processing the processing units in the resource pool. Handle and flexibly extend new features.

附图说明DRAWINGS

通过阅读参照以下附图所作的对非限制性实施例的详细描述，本申请的其它特征、目的和优点将会变得更明显。Other features, objects, and advantages of the present invention will become more apparent from the detailed description of the appended claims.

图1a示出了传统的垂直式网络架构的示意图Figure 1a shows a schematic diagram of a conventional vertical network architecture

图1b示出了水平共享式网络架构的示意图。Figure 1b shows a schematic diagram of a horizontal shared network architecture.

图2示出了基站中资源池的示意图。Figure 2 shows a schematic diagram of a resource pool in a base station.

图3示出了根据本发明一个实施例的构建任务处理路径的方法的流程图。FIG. 3 illustrates a flow chart of a method of constructing a task processing path in accordance with one embodiment of the present invention.

图4示出了根据本发明一个实施例的构建任务处理路径的***的示意图。4 shows a schematic diagram of a system for constructing a task processing path in accordance with one embodiment of the present invention.

图5示出了根据本发明的另一实施例的构建任务处理路径的***的示意图。FIG. 5 shows an illustration of a system for constructing a task processing path in accordance with another embodiment of the present invention. intention.

图6示出了适合于采用本发明的方法和***的资源池的结构图。Figure 6 shows a block diagram of a resource pool suitable for employing the method and system of the present invention.

具体实施方式detailed description

下面结合附图和实施例对本申请作进一步的详细说明。可以理解的是，此处所描述的具体实施例仅用于解释相关发明，而非对该发明的限定。另外，还需要说明的是，为了便于描述，附图中仅示出了与有关发明相关的部分。The present application will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the invention, rather than the invention. In addition, it should be noted that, for convenience of description, only parts related to the related invention are shown in the drawings.

需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。下面将参考附图并结合实施例来详细说明本申请。It should be noted that the embodiments in the present application and the features in the embodiments may be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings.

根据本发明一个实施例，提供一种构建任务处理路径的方法。图3示出了该方法的实施步骤。In accordance with an embodiment of the present invention, a method of constructing a task processing path is provided. Figure 3 shows the implementation steps of the method.

1)步骤S310，选取任务处理路径。1) Step S310, selecting a task processing path.

选取任务处理路径是指从资源池中选择处理能力满足任务需求的处理单元。Selecting a task processing path refers to selecting a processing unit from the resource pool that has the processing capability to meet the task requirements.

在本文中，所述处理单元是指包含硬件且具有一定处理能力的功能单元。例如，资源池中的处理单元可以是包含大级别的硬件单元，例如，基带处理板等，也可以包括相对小级别的处理单元，例如，DSP等。处理单元可以包括诸如CPU、DSP、内存、磁盘等的物理资源，也可以包括诸如虚拟内存等的虚拟资源。Herein, the processing unit refers to a functional unit that includes hardware and has certain processing capabilities. For example, the processing unit in the resource pool may be a hardware unit containing a large level, such as a baseband processing board, etc., and may also include a relatively small level of processing unit, such as a DSP or the like. The processing unit may include physical resources such as a CPU, DSP, memory, disk, etc., and may also include virtual resources such as virtual memory.

所述任务需求可以由用户配置或者根据任务要求的带宽、时延、吞吐量、载波数量等确定。例如，任务需求可以确定为是一个载波或者是最大传输速率50Mbps，或者是两者的结合。本发明对任务需求的类型不作限制。The task requirements may be determined by the user or determined according to the bandwidth, delay, throughput, number of carriers, etc. required by the task. For example, the mission requirement can be determined to be a carrier or a maximum transmission rate of 50 Mbps, or a combination of the two. The present invention does not limit the type of task requirements.

所述任务处理路径是指资源池中的处理单元的组合，用于完成数据流的传输。任务处理路径中处理单元的数量可以是两个或两个以上。The task processing path refers to a combination of processing units in a resource pool for completing transmission of a data stream. The number of processing units in the task processing path may be two or more.

任务处理路径的选择方法有多种。例如，可以简单地从资源池中分别选择处理能力满足任务需要的处理单元，参见图2所示，如果当前的任务需求是一个LTE载波(20M，2*2MIMO)，而在射频处理资源池、基带处理资源池和射频处理资源池中，还存在的剩余资源分别是RRU2(射频处理单元2)、基带处理单元1(BBU1)和协议处理单元1(PPU1)，并且通过计算可知RRU2、BBU1和PPU1可以承担该LTE载波的处理任务。因此，选择的任务处理路径为：RRU2→BBU1→PPU1。具体而言，如果基带处理资源池中的BBU1为一个型号为TI TCI6618的DSP，能够支持双通道20MHz、300Mbps的下行链接和150Mbps上行链接的2x2多输入多输出(MIMO)解决方案，则选择BBU1即可处理当前任务的基带部分。RRU和PPU的选择可以采用类似的方法来分析。。There are several ways to select a task processing path. For example, the processing unit that needs the processing capability to meet the task needs can be selected from the resource pool separately. Referring to FIG. 2, if the current task requirement is an LTE carrier (20M, 2*2 MIMO), and the radio processing resource pool, In the baseband processing resource pool and the radio processing resource pool, the remaining resources still exist are RRU2 (radio frequency processing unit 2), baseband processing unit 1 (BBU1), and protocol processing unit 1 (PPU1), and RRU2, BBU1, and The PPU 1 can assume the processing tasks of the LTE carrier. Cause Therefore, the selected task processing path is: RRU2→BBU1→PPU1. Specifically, if the BBU1 in the baseband processing resource pool is a TI TCI6618 DSP capable of supporting dual 20MHz, 300Mbps downlink and 150Mbps uplink 2x2 multiple input multiple output (MIMO) solutions, select BBU1. The baseband portion of the current task can be processed. The choice of RRU and PPU can be analyzed in a similar way. .

在任务处理路径确定之后，也可以根据网络的监测数据或资源池中处理单元的实际状态而动态地改变该任务处理路径。例如，如果上述LTE载波的负载增加，PPU1处理能力不能满足，则可以从协议资源池中再选择合适的处理单元(如PPU2)增加到该任务处理路径中。同理，如果该LTE载波的负载减小，则可以回收某些处理单元，以用于其他的虚拟基站。通过这种方式，可以有效的提高处理单元的利用率。After the task processing path is determined, the task processing path may also be dynamically changed according to the monitoring data of the network or the actual state of the processing unit in the resource pool. For example, if the load of the LTE carrier is increased and the processing capability of the PPU1 cannot be satisfied, an appropriate processing unit (such as PPU2) may be selected from the protocol resource pool to be added to the task processing path. Similarly, if the load of the LTE carrier is reduced, some processing units can be reclaimed for use in other virtual base stations. In this way, the utilization rate of the processing unit can be effectively improved.

由上述可知，任务处理路径中的各处理单元之间的关系可以是一对一、一对多或者多对多的情况。例如，BBU1可以对应两个处理单元PPU1和PPU2。As can be seen from the above, the relationship between the processing units in the task processing path may be one-to-one, one-to-many, or many-to-many. For example, BBU1 may correspond to two processing units PPU1 and PPU2.

2)步骤S320，查找任务处理路径中处理单元的数据流的目的地址。2) Step S320, searching for the destination address of the data stream of the processing unit in the task processing path.

此步骤的目的在于，根据任务处理路径中处理单元之间的连接关系，获得各个处理单元其处理完的数据流需转发的下一个处理单元。The purpose of this step is to obtain the next processing unit to be forwarded by the processing unit for each of the processing units according to the connection relationship between the processing units in the task processing path.

为了实现上述目的，可以预先存储反映各个处理单元的连接关系的信息表，以获知整个资源池的处理单元的组成情况，例如，对于每个处理单元，可以保存该处理单元的类型、标志、所属硬件板卡、所属机箱、连接情况等(如所连接的交换机及交换机的地址)。每个处理单元可能存在多个连接关系，例如，图2中的基带处理资源池，既通过CPRI交换网络与射频处理单元连接，又通过sRIO交换网络与协议处理单元连接。In order to achieve the above object, an information table reflecting the connection relationship of each processing unit may be stored in advance to know the composition of the processing unit of the entire resource pool. For example, for each processing unit, the type, flag, and belonging of the processing unit may be saved. Hardware board, its own chassis, connection status, etc. (such as the address of the connected switch and switch). There may be multiple connection relationships in each processing unit. For example, the baseband processing resource pool in FIG. 2 is connected to the radio processing unit through the CPRI switching network and to the protocol processing unit through the sRIO switching network.

表1以BBU为例，示意了信息表的所包含的内容。通过该信息表即可以获取每个处理单元的连接关系以及其地址。Table 1 uses the BBU as an example to illustrate the contents of the information table. Through the information table, the connection relationship of each processing unit and its address can be obtained.

表1BBU的信息表Table 1BBU information table

类型Types of	标志Sign	所属板卡Own card	所属机箱Own chassis	连接的交换机Connected switch	地址(2进制表示)Address (in binary)
BBU BBU	11	Board 1 Board 1	Box 1 Box 1	CPRI switch 1 CPRI switch 1	0000000100000001
BBU BBU	11	Board 1 Board 1	Box 1 Box 1	sRIO switch 1 sRIO switch 1	0001000100010001
BBUBBU	22	Board 1 Board 1	Box 1 Box 1	CPRI switch 1 CPRI switch 1	0000001000000010
BBUBBU	22	Board 1 Board 1	Box 1 Box 1	sRIO switch 1 sRIO switch 1	0001001000010010
…...	…...	…...	…...	…...	…...

BBUBBU

kk

Board xBoard x

Box yBox y

CPRI switch nCPRI switch n

0000101000001010

因此，在此步骤中，根据任务处理路径的组成，通过查找处理单元的信息表，即可获得相关处理单元数据流转发的目的地址。Therefore, in this step, according to the composition of the task processing path, by looking up the information table of the processing unit, the destination address of the data stream forwarding of the relevant processing unit can be obtained.

3)步骤S330，将目的地址通知相应的处理单元。3) Step S330, notifying the corresponding processing unit of the destination address.

在此步骤中，向处理单元发送通知消息，以执行通知目的地址的行为，例如，如果RRU2的数据流需要转发给BBU1，则BBU1的地址就是RRU2上数据流的目的地址，为完成数据流的传输，需要将BBU1对应的地址通知RRU2。In this step, a notification message is sent to the processing unit to perform the action of notifying the destination address. For example, if the data stream of the RRU 2 needs to be forwarded to the BBU 1 , the address of the BBU 1 is the destination address of the data stream on the RRU 2 , and the data stream is completed. To transmit, you need to notify RRU2 of the address corresponding to BBU1.

例如，上述通知消息可以包括但不限于以下项：For example, the above notification message may include but is not limited to the following items:

信息类型：用于表明该信息是通知目的地址的信息。Information Type: Information used to indicate that the information is the destination address of the notification.

信息编号：是该信息的唯一标识，可以由***统一生成。Information ID: is the unique identifier of the information, which can be generated by the system.

处理单元标识：是需要通知的处理单元的唯一标识，例如，DSP 1、CPU 1等。Processing unit identification: is a unique identifier of the processing unit that needs to be notified, for example, DSP 1, CPU 1, and the like.

目的地址：用于标识数据流需要发送到的地址。根据处理单元的类型不同，目的地址也可以包括多种类型，例如IP地址、MAC地址等，DSP编号、内存编号等。根据处理单元的对应关系不同，在一条通知消息中可以包括多个目的地址。进一步地，还需要在通知消息中指明地址类型，以便处理单元能够正确解析。Destination Address: Used to identify the address to which the data stream needs to be sent. Depending on the type of processing unit, the destination address may also include multiple types, such as an IP address, a MAC address, etc., a DSP number, a memory number, and the like. Depending on the correspondence of the processing units, a plurality of destination addresses may be included in one notification message. Further, it is also necessary to specify the address type in the notification message so that the processing unit can correctly parse.

4)步骤S340，判断是否获得所有处理单元的反馈。4) Step S340, determining whether feedback of all processing units is obtained.

为了知悉各个处理单元是否已接收到目的地址并已经成功地完成相关的配置，在启动任务处理之前，可以进一步从所有的处理单元接收针对目的地址通知行为的反馈消息。如果指示目的地址配置成功，则执行步骤S350，否则可以选择继续等待一段时间、重传或者考虑更新任务处理路径。通过这种方式，能够保证在各处理单元均成功配置之后，再同步地进行数据流的传输，从而避免因处理单元异常而造成的传输数据错误的情况。In order to know whether each processing unit has received the destination address and has successfully completed the relevant configuration, a feedback message for the destination address notification behavior may be further received from all processing units before the task processing is initiated. If the destination address configuration is successful, step S350 is performed; otherwise, it may choose to continue waiting for a period of time, retransmit, or consider updating the task processing path. In this way, it is ensured that the data streams are resynchronized after each processing unit is successfully configured, thereby avoiding transmission data errors caused by processing unit abnormalities.

在一个实施例中，从处理单元接收的反馈信息包括但不限于下列项：In one embodiment, the feedback information received from the processing unit includes, but is not limited to, the following:

信息类型：用于表明该信息为通知目的地址的反馈信息。Information Type: The feedback information used to indicate that the information is the destination address of the notification.

信息编号：信息编号为通知目的地址信息的编号，以表明该反馈信息所对应的是哪一条通知目的地址的信息。Information No.: The information number is the number of the notification destination address information to indicate which of the notification destination addresses corresponds to the feedback information.

是否成功标志：用于向反馈该目的地址是否配置成功。Whether the success flag is used to feedback whether the destination address is configured successfully.

反馈消息还可以进一步包括处理单元是否正常工作、资源利用率、地址的变化等。The feedback message may further include whether the processing unit is working normally, resource utilization, and ground. Changes in the address, etc.

在实际的应用中，上述所列的项可以包含在一条消息中，或者分别存在于多条消息中。例如，采用状态报告消息(status report)单独指示处理单元异常。In practical applications, the items listed above may be included in one message or in multiple messages. For example, a processing unit exception is indicated separately using a status report message.

反馈信息的上报可以是事件触发式上报或者周期性的上报。例如，在目的地址配置成功、目的地址配置失败、处理单元异常、或处理单元的负载超过阈值时、处理单元的本身地址发生变化时立即上报，从而可以基于该反馈消息，决定是否更新任务处理路径。而对于资源利用率的报告可以采用周期性上报的方式。The reporting of the feedback information may be an event triggered report or a periodic report. For example, when the destination address is configured successfully, the destination address is configured incorrectly, the processing unit is abnormal, or the load of the processing unit exceeds the threshold, and the processing unit's own address is changed, the report is immediately reported, so that whether to update the task processing path may be determined based on the feedback message. . For the report of resource utilization, periodic reporting can be adopted.

在另一实施例中，为了减小信息的交互，也可以在发送通知信息之后，等待预定的时间阈值之后，直接执行步骤S350。In another embodiment, in order to reduce the interaction of the information, step S350 may be directly performed after waiting for the predetermined time threshold after transmitting the notification information.

5)步骤S350，启动任务。5) Step S350, start the task.

启动任务即开始实际的数据流传输，例如，如图2所示，对于任务处理路径RRU2至BBU1再至PPU1，其数据传输的基本过程是：需要处理的数据首先进入RRU2进行处理，处理完的数据根据CPRI协议封装，在目的地址中填入BBU1的地址，发送到CPRI交换机上；CPRI交换机根据数据包的目的地址，将其转发给BBU1；BBU1收到后，进行基带处理，处理完毕后将数据根据RapidIO协议封装，在目的地址中填入PPU1的地址，发送到sRIO交换机上；sRIO交换机根据数据包的目的地址，将其转发给PPU1；PPU1收到后，就可以继续进行协议处理，从而完成该任务。The actual data stream transmission is started when the task is started. For example, as shown in FIG. 2, for the task processing paths RRU2 to BBU1 and then to PPU1, the basic process of data transmission is: the data to be processed first enters RRU2 for processing, and the processed data is processed. The data is encapsulated according to the CPRI protocol, and the address of the BBU1 is filled in the destination address and sent to the CPRI switch. The CPRI switch forwards the packet to the BBU1 according to the destination address of the data packet. After receiving the data, the BBU1 performs baseband processing. The data is encapsulated according to the RapidIO protocol, and the address of the PPU1 is filled in the destination address and sent to the sRIO switch; the sRIO switch forwards the packet to the PPU1 according to the destination address of the data packet; after receiving the PPU1, the protocol processing can be continued. Complete the task.

由上可知，本发明的构建任务处理路径的方法能够向相关的处理单元集中通知其完成数据流之后转发的目的地址。这种方式可以灵活有效地实现任务处理路径的建立和更新。As can be seen from the above, the method for constructing a task processing path of the present invention can notify the relevant processing unit of the destination address forwarded after completing the data stream. This method can flexibly and effectively realize the establishment and update of the task processing path.

根据本发明的一个实施例，还提供了一种适合于执行本发明的构建任务处理路径的方法的***。如图4所示，该***包括配置层、控制层和物理资源层(即资源池)。配置层通过配置接口与控制层进行信息交互，控制层通过控制接口与资源池进行信息交互。配置层和控制层只是逻辑功能模块，可以实现在图4所示的管控中心中，也可以独立于管控中心。In accordance with an embodiment of the present invention, a system suitable for performing the method of constructing a task processing path of the present invention is also provided. As shown in FIG. 4, the system includes a configuration layer, a control layer, and a physical resource layer (ie, a resource pool). The configuration layer exchanges information with the control layer through the configuration interface. The control layer exchanges information with the resource pool through the control interface. The configuration layer and the control layer are only logical function modules, which can be implemented in the control center shown in Figure 4 or independently of the control center.

在此实施例中，配置层的作用是为控制层传递任务需求。例如，配置层的任务需求可以由用户来定义。配置层也可以提供接口来接收来自与其他外部模块的任务处理需求。In this embodiment, the role of the configuration layer is to pass the task requirements to the control layer. For example, the task requirements of the configuration layer can be defined by the user. The configuration layer can also provide an interface to receive task processing requirements from other external modules.

控制层的作用包括但不限于：为某些处理单元分配地址；保存整个资源池的处理单元的组成；根据配置层传递的任务需求、物理资源层的实际情况和/或管控中心的资源管理策略，形成任务处理路径；与处理单元进行信息交互，以配置底层的物理资源，生成与任务处理路径匹配的交换路径；获取处理单元的状态信息等。The role of the control layer includes but is not limited to: assigning addresses to certain processing units; saving the entire capital The composition of the processing unit of the source pool; forming a task processing path according to the task requirements of the configuration layer, the actual situation of the physical resource layer, and/or the resource management policy of the control center; and interacting with the processing unit to configure the underlying physical resources And generating a switching path that matches the task processing path; acquiring state information of the processing unit, and the like.

控制层与处理单元之间的信息交互，包括向处理单元发送告知其目的地址的通知消息以及从处理单元接收反馈信息。The information interaction between the control layer and the processing unit includes sending a notification message to the processing unit notifying its destination address and receiving feedback information from the processing unit.

控制层还可以负责资源池中某些处理单元的目的地址的分配，例如，对于以太网交换机来说，所连接设备的MAC地址都是设备自带的，不需要额外分配。而对于如SRIO交换机来说，所连接设备的地址是需要分配的。在控制层为某些处理单元分配地址之后，通知给相应的交换机和处理单元，以使交换机完成处理单元地址与端口的对应。只要进行信息交互的双方可以根据目的地址唯一地识别目标处理单元，所分配的目的地址可以是多种形式，例如，对于DSP地址而言，可以通过简单的统一编号来标识，或者采用DSP所属的机架号+单板号+本地的DSP标号来唯一确定。The control layer can also be responsible for the allocation of destination addresses of some processing units in the resource pool. For example, for an Ethernet switch, the MAC address of the connected device is self-contained and does not require additional allocation. For an SRIO switch, for example, the address of the connected device needs to be assigned. After the control layer allocates addresses to some processing units, it notifies the corresponding switch and processing unit to enable the switch to complete the correspondence between the processing unit address and the port. As long as the two parties performing information interaction can uniquely identify the target processing unit according to the destination address, the assigned destination address can be in various forms, for example, for the DSP address, it can be identified by a simple unified number, or by using a DSP. The rack number + board number + local DSP label is uniquely determined.

控制层和处理单元之间的控制接口是逻辑接口，该控制接口也可以存在于控制层和交换网络之间。控制层和处理单元之间的信息交互可以由控制层分别与各个处理单元直接通信，也可以是控制层先通知管理代理，由管理代理再进一步通知处理单元，图5示意了管理代理的一个示例。The control interface between the control layer and the processing unit is a logical interface, which may also exist between the control layer and the switching network. The information interaction between the control layer and the processing unit may be directly communicated with each processing unit by the control layer, or the control layer may first notify the management agent, and the management agent further informs the processing unit. FIG. 5 illustrates an example of the management agent. .

在另一个实施例中，管理代理也可以存在于每个资源池内部，或者每个交换网络内部。例如，CPRI交换网络中存在一个管理代理，控制层将BBU1的地址通知到CPRI交换网络中的管理代理，管理代理进行保存。RRU2处理完的数据进入到CPRI交换网络中，其内部的管理代理将RRU2的数据流的目的地址中填入BBU1的地址；交换网络中的交换实体根据目的地址，将其转发给BBU1。通过这种由管理代理集中进行数据流的目的地址填充的方式，可以避免控制层和资源池中每个处理单元进行信息交互，便于集中管理处理单元。In another embodiment, a management agent may also exist within each resource pool, or within each switching network. For example, there is a management agent in the CPRI switching network, and the control layer notifies the management agent in the CPRI switching network of the address of the BBU1, and the management agent saves it. The RRU2 processed data enters the CPRI switching network, and its internal management agent fills the address of the BRU1 with the destination address of the RRU2 data stream; the switching entity in the switching network forwards it to BBU1 according to the destination address. Through the manner in which the management agent centralizes the destination address of the data stream, the information exchange between each control unit in the control layer and the resource pool can be avoided, and the processing unit can be conveniently managed.

由上述可知，在根据本发明的***中，配置层和控制层的功能相对单一、独立，从而能够降低模块之间的耦合度和实现的复杂性。通过控制层向处理单元告知其数据流转发的目的地址的方式，可以灵活动态地构建或改变任务处理路径，并避免处理单元获取目的地址的时间不同步的问题。此外，控制层通过从处理单元接收反馈信息可以及时获取各处理单元的状态(正常、异常、资源利用率等)，以便集中地管理资源池。 As can be seen from the above, in the system according to the present invention, the functions of the configuration layer and the control layer are relatively simple and independent, so that the degree of coupling between modules and the complexity of implementation can be reduced. Through the manner in which the control layer informs the processing unit of the destination address of the data stream forwarding, the task processing path can be flexibly and dynamically constructed, and the problem that the processing unit acquires the destination address is not synchronized. In addition, the control layer can timely acquire the status (normal, abnormal, resource utilization, etc.) of each processing unit by receiving feedback information from the processing unit, so as to centrally manage the resource pool.

为了清楚起见，在本申请中，主要以基站处理资源为例，介绍了本发明的方法和***，本领域的技术人员应当理解，上述的任务处理路径的构建方法和***也可以应用其它技术领域，例如，对于图6所示的资源池的示例，其包括第一处理资源池、第二处理资源池和第三处理资源池，而各处理资源池之间采用交换矩阵或交换网络连接，在这种情况下，可以应用根据本发明的方法和***来构建任务处理路径，同样可以避免传统地采用交换协议构建数据流传输路径时导致的时间不同步等问题。For the sake of clarity, in the present application, the method and system of the present invention are mainly described by taking a base station processing resource as an example. Those skilled in the art should understand that the above-mentioned task processing path construction method and system can also be applied to other technical fields. For example, for the example of the resource pool shown in FIG. 6, it includes a first processing resource pool, a second processing resource pool, and a third processing resource pool, and each processing resource pool uses a switching matrix or a switched network connection, where In this case, the method and system according to the present invention can be applied to construct a task processing path, and the problem of time synchronization caused when the data stream transmission path is traditionally constructed by using the exchange protocol can be avoided.

虽然已经通过例子对本发明的一些特定实施例进行了详细说明，但是本领域的技术人员应该理解，以上例子仅是为了进行说明，而不是为了限制本发明的范围。本领域的技术人员应该理解，可在不脱离本发明的范围和精神的情况下，对以上实施例进行修改。本发明的范围由所附权利要求来限定。 While the invention has been described in detail with reference to the preferred embodiments of the present invention, it is understood that It will be appreciated by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.

Claims

一种用于从多个处理资源池中构建任务处理路径的方法，包括：A method for constructing a task processing path from a plurality of processing resource pools, comprising:

步骤1：基于任务需求确定任务处理路径，其中，所述任务处理路径包含从所述多个处理资源池中选择的处理单元；Step 1: determining a task processing path based on task requirements, where the task processing path includes a processing unit selected from the plurality of processing resource pools;

步骤2：基于所述任务处理路径查找所述处理单元的对应的目的地址；Step 2: Search for a corresponding destination address of the processing unit based on the task processing path.

步骤3：向所述处理单元发送通知消息，其中所述通知消息中包含所述处理单元的数据流将要发送的目的地址。Step 3: Send a notification message to the processing unit, where the notification message includes a destination address to be sent by the data stream of the processing unit.
根据权利要求1所述的方法，其中，还包括：从所述处理单元接收反馈消息。The method of claim 1 further comprising receiving a feedback message from said processing unit.
根据权利要求2所述的方法，其中，所述反馈消息包括下列项中的至少一项：目的地址是否配置成功的标志、处理单元的状态、处理单元的资源利用率、处理单元的地址变化。The method of claim 2, wherein the feedback message comprises at least one of the following: a flag of whether the destination address is successfully configured, a status of the processing unit, a resource utilization of the processing unit, and an address change of the processing unit.
根据权利要求1所述的方法，还包括基于下列项中的至少一项来改变所述任务处理路径：处理单元的资源利用率；处理单元的状态；网络负载；任务需求的变化。The method of claim 1, further comprising changing the task processing path based on at least one of: resource utilization of the processing unit; state of the processing unit; network load; change in task requirements.
根据权利要求1所述的方法，其中，所述处理单元的类型包括处理板、DSP、CPU、内存、硬盘、虚拟内存。The method of claim 1, wherein the type of the processing unit comprises a processing board, a DSP, a CPU, a memory, a hard disk, and a virtual memory.
根据权利要求1所述的方法，其中，所述任务需求包括任务优先级、带宽、时延、吞吐量、载波数量中一项或多项。The method of claim 1 wherein the task requirements include one or more of task priority, bandwidth, latency, throughput, number of carriers.
根据权利要求1所述的方法，其中，在步骤2中，通过预先配置的各个处理单元的地址表来查找所述处理单元的对应的目的地址。The method according to claim 1, wherein in step 2, a corresponding destination address of the processing unit is searched by an address table of each processing unit configured in advance.
一种用于从多个处理资源池中构建任务处理路径的***，包括：A system for constructing a task processing path from a plurality of processing resource pools, comprising:

用于基于任务需求确定任务处理路径的装置，其中，所述任务处理路径包含从所述多个处理资源池中选择的处理单元；用于基于所述任务处理路径查找所述处理单元的对应的目的地址的装置；Means for determining a task processing path based on a task requirement, wherein the task processing path includes a processing unit selected from the plurality of processing resource pools; and searching for a corresponding one of the processing units based on the task processing path Device of destination address;

用于向所述处理单元发送通知消息的装置，其中所述通知消息中包含所述处理单元的数据流将要发送的目的地址。Means for transmitting a notification message to the processing unit, wherein the notification message includes a destination address to which the data stream of the processing unit is to be transmitted.
根据权利要求8所述的***，还包括用于配置或生成所述任务需求的装置。The system of claim 8 further comprising means for configuring or generating said task requirements.
根据权利要求8或9所述的***，还包括管理代理装置，用于向所述处理单元统一发送通知消息和/或为所述处理单元的数据流填充目的地址。 A system according to claim 8 or 9, further comprising management agent means for uniformly transmitting a notification message to said processing unit and/or for filling a data stream of said processing unit address.