技术领域technical field
本发明涉及使用通用处理器(GPP，General Purpose Processors)处理实时信号技术，尤其涉及一种基于GPP的资源调度方法、装置及系统。The present invention relates to the technology of using general purpose processors (GPP, General Purpose Processors) to process real-time signals, in particular to a resource scheduling method, device and system based on GPP.
背景技术Background technique
随着近几年能源和电力价格的上涨，全球的移动通信网络运营商面临日渐严重的成本压力，同时运营商获取站址和机房的难度也在不断加大。由于全球大多数主流运营商通常会同时拥有2-3个不同通信制式的网络，为保证网络的服务质量，就需要部署大量的基站以解决网络覆盖的问题。站址和机房资源的相对稀缺与不断增长的基站数量的矛盾在一定时期内无法协调，目前已成为运营商无法回避的难题。如今，电信市场的激烈竞争使得单用户平均收入(ARPU)增长缓慢甚至下降，严重消弱了移动运营商的盈利能力。而运营商收入的下降必将会导致建网和设备采购投资的压缩，进而影响到全行业的整体发展。面对这种局面，出于行业持续盈利和长期发展考虑，移动通信产业界提出了新型的绿色演进的无线网络(C-RAN)系统架构，以引导未来集中式基带处理网络架构技术的发展。With the rise of energy and power prices in recent years, global mobile communication network operators are facing increasingly serious cost pressures, and it is also becoming more difficult for operators to obtain site sites and computer rooms. Since most mainstream operators in the world usually have 2-3 networks with different communication standards at the same time, in order to ensure the quality of service of the network, it is necessary to deploy a large number of base stations to solve the problem of network coverage. The contradiction between the relative scarcity of site and equipment room resources and the ever-increasing number of base stations cannot be coordinated within a certain period of time, which has become an unavoidable problem for operators. Today's fierce competition in the telecom market has slowed or even declined the average revenue per user (ARPU), seriously weakening the profitability of mobile operators. The decline in operators' income will inevitably lead to the compression of investment in network construction and equipment procurement, which in turn will affect the overall development of the entire industry. Faced with this situation, out of consideration for the industry's sustainable profitability and long-term development, the mobile communication industry has proposed a new green evolved radio network (C-RAN) system architecture to guide the development of future centralized baseband processing network architecture technology.
C-RAN系统主要由远端的无线射频单元(RRU)与天线组成的分布式无线网络、具有高带宽、低延迟特性的光传输网络、以及集中式基带处理池三大部分组成。其中，集中式基带处理池由集中在一个物理站点的多个基带单元(BBU)构成，所有BBU和RRU通过具有高带宽、低延迟特性的光传输网络连接起来；并且，集中式基带处理池中多个BBU之间交叉连接。实现集中式基带处理池的功能需要应用基站虚拟化技术，具体地，需要支持集中式基带处理池的物理资源和计算能力的虚拟分配和组合。The C-RAN system is mainly composed of a distributed wireless network composed of remote radio frequency units (RRU) and antennas, an optical transmission network with high bandwidth and low delay characteristics, and a centralized baseband processing pool. Among them, the centralized baseband processing pool is composed of multiple baseband units (BBUs) concentrated in one physical site, and all BBUs and RRUs are connected through an optical transmission network with high bandwidth and low delay characteristics; and, the centralized baseband processing pool Cross connection between multiple BBUs. Realizing the function of the centralized baseband processing pool requires the application of base station virtualization technology, specifically, it needs to support virtual allocation and combination of physical resources and computing capabilities of the centralized baseband processing pool.
集中式基带处理池中多个BBU之间互联互通，构成高容量、低延迟的互联架构，而远端的RRU通过互联架构即光传输网络可以与集中式基带处理池中任意一个BBU进行交互，这种方式可以认为是对现有BBU进行集中化集成处理，如此，有效地实现了载波负载均衡、容灾备份，同时还能实现提高设备利用率、减少基站机房数量、以及降低能耗的目的。Multiple BBUs in the centralized baseband processing pool are interconnected to form a high-capacity, low-latency interconnection architecture, and the remote RRU can interact with any BBU in the centralized baseband processing pool through the interconnection architecture, that is, the optical transmission network. This method can be considered as a centralized integrated processing of the existing BBU. In this way, carrier load balancing, disaster recovery and backup are effectively realized, and at the same time, it can also achieve the purpose of improving equipment utilization, reducing the number of base station equipment rooms, and reducing energy consumption. .
现有实现集中式基带处理池的功能的技术方案中，多采用传统的数字信号处理器(DSP，Digital Signal Processing)+现场可编程门阵列(FPGA，FieldProgrammable Gate Array)的架构处理基带信号，并采用额外的附加交换设备实现多个BBU之间的同相/正交相位(I/Q，In-phase/Quadrature-phase)路数据的交换。但是，以DSP为基带处理核心的硬件架构必然会有着对应的软件架构特点，具体地，由于DSP的操作系统相对简单，当DSP内部多个核之间协作处理信号时，操作系统很难自动化分配处理资源，因此，DSP的软件架构设计往往基于固定的时序图。如此，当系统配置发生变化，比如：当从时分同步码分多址接入(TD-SCDMA，Time Division-Synchronous Code Division MultipleAccess)软升级为时分双工-长期演进(TDD-LTE，Time Division Duplexing-LongTerm Evolution)时，往往需要重启离线更改DSP的软件配置。并且，当DSP的软件配置组合复杂时，固定时序设计会给软件设计带来极高的复杂度，如此，DSP软件的升级和维护会大大增大成本。Among the existing technical solutions for realizing the function of the centralized baseband processing pool, the traditional digital signal processor (DSP, Digital Signal Processing) + field programmable gate array (FPGA, Field Programmable Gate Array) architecture is mostly used to process the baseband signal, and The exchange of in-phase/quadrature-phase (I/Q, In-phase/Quadrature-phase) data between multiple BBUs is realized by using an additional additional switching device. However, the hardware architecture with DSP as the baseband processing core will inevitably have corresponding software architecture characteristics. Specifically, because the DSP operating system is relatively simple, when multiple cores in the DSP cooperate to process signals, it is difficult for the operating system to automatically allocate Therefore, the software architecture design of DSP is often based on fixed timing diagrams. In this way, when the system configuration changes, for example: when softly upgrading from Time Division-Synchronous Code Division Multiple Access (TD-SCDMA, Time Division-Synchronous Code Division Multiple Access) to Time Division Duplex-Long Term Evolution (TDD-LTE, Time Division Duplexing -LongTerm Evolution), it is often necessary to restart and change the DSP software configuration offline. Moreover, when the combination of DSP software configuration is complex, the fixed timing design will bring extremely high complexity to the software design, so the upgrade and maintenance of DSP software will greatly increase the cost.
另外，当多个DSP间相互协作完成信号处理时，各DSP间的信号交互与DSP内多核间的信号交互会采用不同的方式。由于协作处理时额外需要考虑信号发送时延、以及对方DSP工作的正常性，因此，在这种情况下，底层信息交互的维护会相对复杂。In addition, when multiple DSPs cooperate with each other to complete signal processing, the signal interaction between each DSP and the signal interaction between multiple cores in the DSP will adopt different methods. Since the signal transmission delay and the normality of the other party's DSP work need to be considered during cooperative processing, in this case, the maintenance of the underlying information exchange will be relatively complicated.
为解决DSP+FPGA架构带来的缺陷，提出了一种基于GPP、通过软件无线电技术实现多标准的统一开放的实现集中式基带处理池功能的技术方案，采用GPP的优点是：首先，GPP具有后向兼容好的优点，因此有利于网络系统的平滑演进；其次，在GPP中，多核间的通信有成熟的方式和方法，操作系统也较适合自动化地任务分配，对编程模型的要求比较简单。In order to solve the defects caused by the DSP+FPGA architecture, a technical solution based on GPP and realizing multi-standard unified and open centralized baseband processing pool function through software radio technology is proposed. The advantages of using GPP are: First, GPP has The advantage of backward compatibility is good, so it is conducive to the smooth evolution of the network system; secondly, in GPP, there are mature methods and methods for multi-core communication, and the operating system is more suitable for automatic task allocation, and the requirements for programming models are relatively simple. .
但是，目前，在实现集中式基带处理池功能时，即：在处理实时的数字信号时，传统的基于操作系统自动分配核资源的处理方式影响了处理效率。However, at present, when realizing the function of a centralized baseband processing pool, that is, when processing real-time digital signals, the traditional processing method of automatically allocating core resources based on an operating system affects processing efficiency.
发明内容Contents of the invention
有鉴于此，本发明的主要目的在于提供一种资源调度方法、装置及系统，在处理实时信号时，能有效地实现资源的动态调度。In view of this, the main purpose of the present invention is to provide a resource scheduling method, device and system, which can effectively realize dynamic scheduling of resources when processing real-time signals.
为达到上述目的，本发明的技术方案是这样实现的：In order to achieve the above object, technical solution of the present invention is achieved in that way:
本发明提供了一种资源调度方法，设置业务调度单元及环境调度单元；该方法还包括：The present invention provides a resource scheduling method, setting a service scheduling unit and an environment scheduling unit; the method also includes:
将要执行的业务划分为一个以上子业务，在执行所述业务的各子业务过程中，当需要使用相应资源时，所述业务调度单元通知所述环境调度单元分配所需资源。The service to be executed is divided into more than one sub-service, and during the execution of each sub-service of the service, when the corresponding resource needs to be used, the service scheduling unit notifies the environment scheduling unit to allocate the required resource.
上述方案中，所述将要执行的业务划分为一个以上子业务，包括：In the above scheme, the business to be executed is divided into more than one sub-business, including:
将所述业务划分为一个以上能独立运行的子业务；Divide the business into more than one sub-business that can operate independently;
确定各子业务所需的参数信息；Determine the parameter information required by each sub-service;
确定各子业务之间的关联关系。Determine the relationship between each sub-business.
上述方案中，所述将所述业务划分为一个以上能独立运行的子业务，为：In the above scheme, the division of the business into more than one sub-services that can operate independently is:
依据划分出的子业务之间的耦合相关性小的策略，将所述业务划分为一个以上能独立运行的子业务。According to the policy that the coupling correlation between the divided sub-services is small, the service is divided into more than one sub-services that can operate independently.
上述方案中，在执行所述业务的各子业务之前，该方法还包括：In the above solution, before executing each sub-service of the service, the method also includes:
根据各子业务的启动条件及处理时延，确定环境调度单元的触发机制；According to the starting conditions and processing delay of each sub-service, determine the trigger mechanism of the environmental scheduling unit;
相应的，所述业务调度单元根据确定的环境调度单元的触发机制通知所述环境调度单元分配所述所需资源。Correspondingly, the service scheduling unit notifies the environment scheduling unit to allocate the required resources according to the determined trigger mechanism of the environment scheduling unit.
上述方案中，该方法还包括：In the above-mentioned scheme, the method also includes:
在执行所述业务的各子业务过程中，所述业务调度单元与模块资源的算法模块进行接口适配，完成各子业务处理。During the execution of each sub-service of the service, the service scheduling unit performs interface adaptation with the algorithm module of the module resource to complete the processing of each sub-service.
上述方案中，在分配所述所需资源时，该方法进一步包括：In the above solution, when allocating the required resources, the method further includes:
所述环境调度单元根据资源的使用情况及所述所需资源的使用时间片，分配所述所需资源。The environment scheduling unit allocates the required resources according to the usage of the resources and the usage time slice of the required resources.
本发明又提供了一种资源调度装置，该装置包括：业务调度单元、以及环境调度单元；其中，The present invention further provides a resource scheduling device, which includes: a service scheduling unit, and an environment scheduling unit; wherein,
业务调度单元，用于在执行划分为一个以上子业务的业务的各子业务过程中，当需要使用相应资源时，通知环境调度单元；The service scheduling unit is used to notify the environment scheduling unit when corresponding resources need to be used during the execution of each sub-service that is divided into more than one sub-service;
环境调度单元，用于收到业务调度单元的通知后，分配所需资源。The environment scheduling unit is configured to allocate required resources after receiving the notification from the business scheduling unit.
上述方案中，所述环境调度单元，具体用于：在分配所述所需资源时，根据资源的使用情况及所述所需资源的使用时间片，分配所述所需资源。In the above solution, the environment scheduling unit is specifically configured to: when allocating the required resources, allocate the required resources according to the resource usage and the usage time slice of the required resources.
本发明还提供了一种资源调度系统，该系统包括：应用子系统、平台资源、硬件资源、底层资源、以及模块资源；所述应用子系统进一步包括资源调度装置，所述资源调度装置包括：业务调度单元、以及环境调度单元；其中，The present invention also provides a resource scheduling system, which includes: an application subsystem, platform resources, hardware resources, underlying resources, and module resources; the application subsystem further includes a resource scheduling device, and the resource scheduling device includes: business scheduling unit, and environment scheduling unit; wherein,
业务调度单元，用于在执行划分为一个以上子业务的业务的各子业务过程中，当需要使用相应资源时，通知环境调度单元；The service scheduling unit is used to notify the environment scheduling unit when corresponding resources need to be used during the execution of each sub-service that is divided into more than one sub-service;
环境调度单元，用于收到业务调度单元的通知后，分配所需资源。The environment scheduling unit is configured to allocate required resources after receiving the notification from the business scheduling unit.
上述方案中，所述平台资源、所述硬件资源、以及所述底层资源组成资源调度系统的资源，用于被所述环境调度单元控制。In the above solution, the platform resources, the hardware resources, and the underlying resources constitute the resources of the resource scheduling system, and are used to be controlled by the environment scheduling unit.
上述方案中，所述环境调度单元，具体用于：在分配所述所需资源时，根据资源的使用情况及所述所需资源的使用时间片，分配所述所需资源。In the above solution, the environment scheduling unit is specifically configured to: when allocating the required resources, allocate the required resources according to the resource usage and the usage time slice of the required resources.
上述方案中，所述模块资源进一步包括算法模块；所述业务调度单元，还用于在执行所述业务的各子业务过程中，所述业务调度单元与所述算法模块进行接口适配，完成各子业务处理。In the above solution, the module resource further includes an algorithm module; the service scheduling unit is also used to perform interface adaptation between the service scheduling unit and the algorithm module during the execution of each sub-service of the service, to complete Processing of each sub-business.
本发明提供的资源调度方法、装置及系统，设置业务调度单元及环境调度单元；将要执行的业务划分为一个以上子业务，在执行所述业务的各子业务过程中，当需要使用相应资源时，所述业务调度单元通知所述环境调度单元分配所需资源，业务调度和资源调度进行了分离，由环境调度单元统一分配资源，如此，在处理实时信号时，能有效地实现资源的动态调度。并且，由于业务调度和资源调度进行了分离，资源能最大化的共享。The resource scheduling method, device and system provided by the present invention are provided with a business scheduling unit and an environment scheduling unit; the business to be executed is divided into more than one sub-service, and during the execution of each sub-service of the business, when corresponding resources are needed , the business scheduling unit notifies the environment scheduling unit to allocate required resources, the business scheduling and resource scheduling are separated, and the environment scheduling unit allocates resources uniformly, so that when processing real-time signals, the dynamic scheduling of resources can be effectively realized . Moreover, since service scheduling and resource scheduling are separated, resources can be shared to the maximum extent.
另外，依据划分出的子业务之间的耦合相关性小的策略，将所述业务划分为一个以上能独立运行的子业务，如此，能使环境调度单元更加灵活地调度资源。In addition, according to the policy that the coupling correlation between the divided sub-services is small, the service is divided into more than one sub-service that can run independently, so that the environment scheduling unit can more flexibly schedule resources.
本发明中，在执行所述业务的各子业务过程中，所述业务调度单元与模块资源的算法模块进行接口适配，完成各子业务处理，业务调度单元不关心通信机制，也不关心各处理资源之间的相互处理时序关系，仅调用算法模块，即仅完成某一空口制式的算法接口的适配和信号处理的优化，如此，提高了代码的可重用性，使得软件升级简单化，从而有利于系统的平滑演进。In the present invention, during the execution of each sub-service of the service, the service scheduling unit performs interface adaptation with the algorithm module of the module resource to complete the processing of each sub-service, and the service scheduling unit does not care about the communication mechanism or each sub-service. The mutual processing timing relationship between processing resources only calls the algorithm module, that is, only completes the adaptation of the algorithm interface of a certain air interface standard and the optimization of signal processing. In this way, the reusability of the code is improved, and the software upgrade is simplified. This is conducive to the smooth evolution of the system.
本发明中，在执行所述业务的各子业务过程中，所述环境调度单元根据确定的维护机制，实时维护各处理资源的任务队列，如此，最大化地保证了系统异常的可恢复性，从而具有更高的可靠性。In the present invention, during the execution of each sub-business process of the business, the environment scheduling unit maintains the task queue of each processing resource in real time according to the determined maintenance mechanism, so that the recoverability of the system exception is maximized, Thereby having higher reliability.
本发明中，在执行所述业务的各子业务过程中，所述环境调度单元释放未使用的处理资源，如此，能有效地节约能源。In the present invention, during the execution of each sub-service of the service, the environment dispatching unit releases unused processing resources, thus effectively saving energy.
本发明中，所述环境调度单元根据资源的使用情况及所述所需资源的使用时间片，分配所述所需资源，减少了GPP平台的操作系统任务调度的额外开销，如此，有效地提高了处理效率。In the present invention, the environment scheduling unit allocates the required resources according to the resource usage and the usage time slice of the required resources, which reduces the additional overhead of the operating system task scheduling of the GPP platform, thus effectively improving the processing efficiency.
附图说明Description of drawings
图1为基于GPP的实现集中式基带处理池功能的系统架构示意图；FIG. 1 is a schematic diagram of a system architecture based on GPP to realize a centralized baseband processing pool function;
图2为本发明资源调度方法流程示意图；FIG. 2 is a schematic flow chart of the resource scheduling method of the present invention;
图3为本发明实施例二PUSCH划分出的子业务示意图；FIG. 3 is a schematic diagram of subservices divided by PUSCH according to Embodiment 2 of the present invention;
图4为本发明实施例二各处理核间动态共享处理资源示意图；FIG. 4 is a schematic diagram of dynamically sharing processing resources between processing cores in Embodiment 2 of the present invention;
图5为本发明资源调度装置结构示意图；FIG. 5 is a schematic structural diagram of a resource scheduling device according to the present invention;
图6为本发明资源调度系统结构示意图。FIG. 6 is a schematic structural diagram of the resource scheduling system of the present invention.
具体实施方式Detailed ways
下面结合附图及具体实施例对本发明再作进一步详细的说明。The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
图1为基于GPP的实现集中式基带处理池功能的系统架构示意图，如图1所示，一个完整的基于GPP的实现集中式基带处理池功能的系统包括：应用子系统、平台资源、硬件资源、底层资源、以及模块资源；其中，平台资源为GPP平台，平台资源和硬件资源共同组成整个系统赖以构建的软硬件开发平台，这部分的资源随着系统使用的GPP平台的不同而改变，比如：不同的CPU等；包含操作系统的底层资源封装了软硬件开发平台的全部硬件资源和大部分软件资源；应用子系统和模块资源是整个系统的主要组成部分，应用子系统可根据不同的平台资源、系统需求而改变，且可混合多种不同的空口标准，模块资源可以在不同的平台资源间平滑移植。在图1中，同一资源层次中，位置在上的资源单元对位置在下的资源单元具有依赖关系，具体地讲，应用子系统对模块资源具有依赖关系，应用子系统还对底层资源具有依赖关系，模块资源对底层资源具有依赖关系，底层资源对硬件资源具有依赖关系。由于GPP在处理实时信号时需要更高的处理效率，传统的基于操作系统自动分配核资源的处理方式影响了处理效率，基于此，本发明资源调度方法，如图2所示，包括以下步骤：Figure 1 is a schematic diagram of the system architecture based on GPP to realize the function of the centralized baseband processing pool. As shown in Figure 1, a complete system based on GPP to realize the function of the centralized baseband processing pool includes: application subsystems, platform resources, and hardware resources , underlying resources, and module resources; among them, the platform resource is the GPP platform, and the platform resource and the hardware resource together form the software and hardware development platform on which the entire system is built. This part of the resources varies with the GPP platform used by the system. For example: different CPUs, etc.; the underlying resources including the operating system encapsulate all the hardware resources and most of the software resources of the software and hardware development platform; the application subsystem and module resources are the main components of the entire system, and the application subsystem can Platform resources and system requirements change, and a variety of different air interface standards can be mixed, and module resources can be smoothly transplanted between different platform resources. In Figure 1, in the same resource hierarchy, the upper resource unit has a dependency relationship with the lower resource unit. Specifically, the application subsystem has a dependency relationship with the module resources, and the application subsystem also has a dependency relationship with the underlying resources. , module resources have dependencies on underlying resources, and underlying resources have dependencies on hardware resources. Since GPP requires higher processing efficiency when processing real-time signals, the traditional processing method based on the automatic allocation of core resources by the operating system has affected the processing efficiency. Based on this, the resource scheduling method of the present invention, as shown in Figure 2, includes the following steps:
步骤200：设置业务调度单元及环境调度单元；Step 200: setting up a business scheduling unit and an environment scheduling unit;
这里，在实际应用时，本步骤的具体实现是指将整体实时信号处理软件中实现调度的调度逻辑分成业务调度逻辑和环境调度逻辑。Here, in actual application, the specific implementation of this step refers to dividing the scheduling logic for implementing scheduling in the overall real-time signal processing software into business scheduling logic and environment scheduling logic.
步骤201：将要执行的业务划分为一个以上子业务，在执行所述业务的各子业务过程中，当需要使用相应资源时，所述业务调度单元通知所述环境调度单元分配所需资源。Step 201: Divide the service to be executed into more than one sub-service. During the execution of each sub-service of the service, when the corresponding resource needs to be used, the service scheduling unit notifies the environment scheduling unit to allocate the required resource.
这里，所述将要执行的业务划分为一个以上子业务，具体包括：Here, the business to be executed is divided into more than one sub-business, specifically including:
将所述业务划分为一个以上能独立运行的子业务；Divide the business into more than one sub-business that can operate independently;
确定各子业务所需的参数信息；Determine the parameter information required by each sub-service;
确定各子业务之间的关联关系。Determine the relationship between each sub-business.
其中，所述将所述业务划分为一个以上能独立运行的子业务，具体为：Wherein, the division of the business into more than one sub-services that can operate independently is specifically:
依据划分出的子业务之间的耦合相关性小的策略，将所述业务划分为一个以上能独立运行的子业务。进一步地，在划分时，依据的策略还可以包括：功能单一且处理复杂度适度。According to the policy that the coupling correlation between the divided sub-services is small, the service is divided into more than one sub-services that can operate independently. Further, when dividing, the basis strategy may also include: single function and moderate processing complexity.
所述参数信息包括：配置信息和输入输出的接口信息。The parameter information includes: configuration information and input and output interface information.
在执行所述业务的各子业务之前，该方法还可以进一步包括：Before executing each sub-service of the service, the method may further include:
根据各子业务的启动条件及处理时延，确定环境调度单元的触发机制；According to the starting conditions and processing delay of each sub-service, determine the trigger mechanism of the environmental scheduling unit;
相应的，所述业务调度单元根据确定的环境调度单元的触发机制通知所述环境调度单元分配所述所需资源。Correspondingly, the service scheduling unit notifies the environment scheduling unit to allocate the required resources according to the determined trigger mechanism of the environment scheduling unit.
该方法还可以进一步包括：The method may further include:
在执行所述业务的各子业务过程中，所述业务调度单元与模块资源的算法模块进行接口适配，完成各子业务处理。During the execution of each sub-service of the service, the service scheduling unit performs interface adaptation with the algorithm module of the module resource to complete the processing of each sub-service.
这里，所述业务调度单元与模块资源的算法模块进行接口适配是指：所述业务调度单元完成某一空口制式的算法接口的适配。Here, the interface adaptation between the service scheduling unit and the algorithm module of the module resource refers to: the service scheduling unit completes the adaptation of the algorithm interface of a certain air interface standard.
在分配所述所需资源时，该方法还可以进一步包括：When allocating the required resources, the method may further include:
所述环境调度单元根据资源的使用情况及所述所需资源的使用时间片，分配所述所需资源。The environment scheduling unit allocates the required resources according to the usage of the resources and the usage time slice of the required resources.
该方法还可以进一步包括：The method may further include:
在执行所述业务的各子业务过程中，所述环境调度单元根据确定的维护机制，实时维护各处理资源的任务队列。During the execution of each sub-service process of the service, the environment scheduling unit maintains the task queue of each processing resource in real time according to the determined maintenance mechanism.
其中，所述处理资源是指已被各子业务使用的资源。Wherein, the processing resources refer to resources that have been used by each sub-service.
所述环境调度单元实时维护各处理资源的任务队列，包括：The environment scheduling unit maintains the task queue of each processing resource in real time, including:
当各处理资源出现执行各子业务的时间超时或通信故障的异常时，所述环境调度单元进行处理；这里，所述环境调度单元如何处理各处理资源出现的执行时间超时或通信故障的具体实现为本领域技术人员惯用技术手段，不再赘述。When the execution time of each processing resource exceeds the execution time of each sub-service or the abnormality of the communication failure occurs, the environment scheduling unit performs processing; here, the concrete realization of how the environment scheduling unit handles the execution time of each processing resource when the execution time exceeds or the communication failure occurs It is a technical means commonly used by those skilled in the art and will not be repeated here.
该方法还可以进一步包括：The method may further include:
在执行所述业务的各子业务过程中，所述环境调度单元释放未使用的处理资源。比如：将未使用的处理核进行降频节能处理等；这里，释放未使用的处理资源的具体实现本领域技术人员惯用技术手段，不再赘述。During execution of each sub-service process of the service, the environment scheduling unit releases unused processing resources. For example: reducing the frequency of unused processing cores for energy-saving processing, etc.; here, specific implementations of releasing unused processing resources are commonly used technical means by those skilled in the art, and will not be repeated here.
下面结合实施例对本发明再作进一步详细的描述。The present invention will be further described in detail below in conjunction with the examples.
实施例一Embodiment one
本实施例以长期演进(LTE，Long Term Evolution)系统为例，描述实现本发明资源调度方法的准备工作，准备工作主要包括以下步骤：This embodiment takes the Long Term Evolution (LTE, Long Term Evolution) system as an example to describe the preparatory work for implementing the resource scheduling method of the present invention. The preparatory work mainly includes the following steps:
步骤a：将业务划分为一个以上能独立运行的子业务；Step a: Divide the business into more than one sub-services that can operate independently;
在本实施例中，是将基带处理物理层信号的处理流程划分为一个以上能独立运行的子业务；In this embodiment, the processing flow of baseband processing physical layer signals is divided into more than one sub-services that can operate independently;
这里，在划分时，以划分出的子业务之间耦合相关性小、功能单一、处理复杂度适当为准则。Here, when dividing, the criteria for the divided sub-services are small coupling correlation, single function, and appropriate processing complexity.
步骤b：确定各子业务所需的参数信息，并将划分的各子业务分别封装成能独立运行的业务调度实例；Step b: Determine the parameter information required by each sub-service, and package the divided sub-services into service scheduling instances that can run independently;
这里，所述参数信息包括：配置信息和输入输出的接口信息。Here, the parameter information includes: configuration information and input/output interface information.
步骤c：确定各业务调度实例之间的关联关系；Step c: Determine the association relationship between each service scheduling instance;
具体地，画出数据在各业务调度实例间的处理流图、以及各业务调度实例之间的相互协作关系。Specifically, draw the data processing flow diagram among the business scheduling instances, and the mutual cooperation relationship among the business scheduling instances.
步骤d：根据各业务调度实例触发启动的条件和各业务调度实例的处理时延，确定环境调度单元的触发机制；Step d: Determine the triggering mechanism of the environment scheduling unit according to the conditions for triggering and starting of each service scheduling instance and the processing delay of each service scheduling instance;
这里，举个例子来说，根据各业务调度实例触发启动的条件和各业务调度实例的处理时延，确定环境调度单元的触发机制为：当某个业务调度实例触发启动的条件满足后，触发环境调度单元。Here, as an example, according to the triggering conditions of each business scheduling instance and the processing delay of each business scheduling instance, the triggering mechanism of the environment scheduling unit is determined as follows: when the triggering condition of a certain business scheduling instance is met, trigger Environment scheduling unit.
步骤e：确定环境调度单元的环境调度实现及维护机制；Step e: Determine the environment scheduling implementation and maintenance mechanism of the environment scheduling unit;
这里，确定的环境调度单元的环境调度实现机制为：环境调度单元根据当前各处理核正在执行的业务调度实例和各处理核待执行的业务调度实例，判断当前处理的子业务下发的目标核，将当前处理的子业务分配给可以及时完成相关处理的核执行。Here, the environment scheduling implementation mechanism of the determined environment scheduling unit is as follows: the environment scheduling unit judges the target core issued by the currently processed sub-service according to the business scheduling instance currently being executed by each processing core and the business scheduling instance to be executed by each processing core. , assign the currently processed sub-business to the core execution that can complete the relevant processing in time.
确定的维护机制为：实时维护各处理核的任务队列，处理下发的各业务调度实例执行时间片超时或通信故障等异常处理，从而保证环境调度单元的健壮性。The determined maintenance mechanism is: maintain the task queues of each processing core in real time, and handle exceptions such as time slice timeout or communication failure for each dispatched business scheduling instance, so as to ensure the robustness of the environmental scheduling unit.
步骤f：确定环境调度单元的节能机制；Step f: Determine the energy-saving mechanism of the environment scheduling unit;
具体地，将部分未使用的处理核进行降频节能处理。Specifically, some unused processing cores are subjected to frequency reduction and energy-saving processing.
实施例二Embodiment two
为了更好地说明实施例一中实现本发明资源调度方法的准备工作，本实施例以LTE的上行处理流程为例进行描述；其中，LTE的上行处理流程是演进型节点B(eNB，EvolvedNode B)物理层处理的主要部分。In order to better illustrate the preparatory work for implementing the resource scheduling method of the present invention in Embodiment 1, this embodiment takes the uplink processing flow of LTE as an example for description; wherein, the uplink processing flow of LTE is an evolved node B (eNB, EvolvedNode B ) The main part of the physical layer processing.
首先，描述子业务的划分、以及各子业务调度触发机制的确定。First, the division of sub-services and the determination of the scheduling trigger mechanism for each sub-service are described.
LTE的上行处理流程可包括：前端处理、物理随机接入信道(PRACH，Physical Random Access Channel)处理、SRS处理、信道估计、符号(Symbol)处理、以及比特(Bit)处理。其中，前端处理及PRACH处理可认为与用户设备(UE，User Equipment)配置无关，而SRS处理、信道估计、Symbol处理、以及Bit处理均与UE配置直接相关，UE调度的数量决定了这些处理模块的处理负载。因此，在划分子业务时，可以以处理的信道对象为粒度进行划分，每类不同的信道还可进一步划分子业务。下面以物理上行共享信道(PUSCH，Physical Uplink Shared Channel)为例，详细描述信道划分子业务的方式。图3为PUSCH划分出的子业务示意图；如图3所示，划分出的子业务包括：前端抽样级处理子业务、Symbol级处理子业务、以及Bit级处理子业务。其中，前端抽样级处理子业务的处理包括降采样和傅立叶变换(FFT，Fourier Transform)等操作，这部分操作与当前子帧的UE配置无关，因此可按照天线进行均分，单任务处理降采样、每个FFT模块处理2个天线。前端抽样级处理子业务的待处理数据为天线口数据，处理后的数据为FFT操作后的结果，调度触发机制为：单正交频分复用(OFDM，Orthogonal Frequency Division Multiplexing)符号完整接收。对于PUSCH，Symbol级处理子业务的处理主要包括：信道估计、解调、离散傅里叶反变换(IDFT，Inverse Discrete Fourier Transform)、频域均衡等操作，处理目标是多组UE配置资源块(RB，Resource Block)组；Symbol级处理子业务的处理复杂度与处理的RB块大小和块数量近似线性相关。因此可按照RB组为粒度进行划分；Symbol级处理子业务的待处理数据为某单个UE分配的RB资源块，处理后的数据为单UE数据的解调后的结果。调度触发机制为：单子帧的多个OFDM符号已被处理完毕，且具备各UE在当前子帧的调度配置信息。Bit级操作子业务的处理包括：解扰码、解速率匹配、信道译码等操作，处理复杂度与编码块的大小及码块数量直接相关。由于按照UE承载的传输块大小、以及当前信道的信噪比(SNR，Signal Noise Ratio)可动态预测出每个UE的处理延时，因此以编码块(CB，Code Block)为处理单位粒度设计子业务。Bit级处理子业务的待处理数据为单UE的单个编码块的解调后信息，处理后的数据为译码后的结果。The uplink processing flow of LTE may include: front-end processing, Physical Random Access Channel (PRACH, Physical Random Access Channel) processing, SRS processing, channel estimation, symbol (Symbol) processing, and bit (Bit) processing. Among them, front-end processing and PRACH processing can be considered to have nothing to do with user equipment (UE, User Equipment) configuration, while SRS processing, channel estimation, Symbol processing, and Bit processing are all directly related to UE configuration, and the number of UE scheduling determines these processing modules processing load. Therefore, when dividing sub-services, the granularity of processed channel objects can be divided, and each type of channel can be further divided into sub-services. The following takes Physical Uplink Shared Channel (PUSCH, Physical Uplink Shared Channel) as an example to describe in detail how channels are divided into sub-services. FIG. 3 is a schematic diagram of sub-services divided by PUSCH; as shown in FIG. 3 , the divided sub-services include: front-end sampling-level processing sub-service, Symbol-level processing sub-service, and Bit-level processing sub-service. Among them, the processing of sub-services at the front-end sampling level includes operations such as downsampling and Fourier Transform (FFT, Fourier Transform). This part of the operation has nothing to do with the UE configuration of the current subframe, so it can be divided equally according to the antenna, and the downsampling is processed by a single task. , Each FFT module handles 2 antennas. The data to be processed at the front-end sampling level processing sub-service is the antenna port data, and the processed data is the result of the FFT operation. The scheduling trigger mechanism is: a single Orthogonal Frequency Division Multiplexing (OFDM, Orthogonal Frequency Division Multiplexing) symbol is completely received. For PUSCH, the processing of Symbol-level processing sub-services mainly includes: channel estimation, demodulation, inverse discrete Fourier transform (IDFT, Inverse Discrete Fourier Transform), frequency domain equalization and other operations, and the processing target is multiple groups of UE configuration resource blocks ( RB, Resource Block) group; the processing complexity of Symbol-level processing sub-services is approximately linearly related to the size of the processed RB block and the number of blocks. Therefore, it can be divided according to the granularity of the RB group; the data to be processed of the Symbol-level processing sub-service is the RB resource block allocated by a single UE, and the processed data is the demodulated result of the single UE data. The scheduling trigger mechanism is as follows: multiple OFDM symbols in a single subframe have been processed, and the scheduling configuration information of each UE in the current subframe is available. The processing of bit-level operation sub-services includes: descrambling, derate matching, channel decoding and other operations, and the processing complexity is directly related to the size of the code block and the number of code blocks. Since the processing delay of each UE can be dynamically predicted according to the size of the transmission block carried by the UE and the SNR (Signal Noise Ratio) of the current channel, the granularity design is based on the code block (CB, Code Block) sub business. The data to be processed in the bit-level processing sub-service is the demodulated information of a single coding block of a single UE, and the processed data is the decoded result.
估算各子业务的处理时延，其中，各子业务的处理时延如下：Estimate the processing delay of each sub-service, where the processing delay of each sub-service is as follows:
表1前端抽样级处理子业务的处理时延Table 1 Processing delay of front-end sampling level processing sub-services
表2Symbol级处理子业务的处理时延Table 2 Processing delay of Symbol-level processing sub-services
表3Bit级处理子业务的处理时延Table 3 Bit-level processing delay of processing sub-services
其次，描述环境调度单元的设置及功能，具体包括：时序关系排布、接口数据传输等。Next, describe the settings and functions of the environment scheduling unit, specifically including: timing relationship arrangement, interface data transmission, etc.
在实际应用时，设置的环境调度单元对应的环境调度逻辑可以绑定在主核上，负责环境调度中查询业务处理状态和分配处理任务的功能，其他核主要处理业务的各子业务。在设置时，环境调度单元能根据每个子业务的关键信息如：处理天线数、目标频域数据RB大小、CB块大小、以及SNR等可以估算出每个子业务的处理时延，因此可合理地分配给处理子业务的核处理。In actual application, the environment scheduling logic corresponding to the set environment scheduling unit can be bound to the main core, which is responsible for querying the business processing status and assigning processing tasks in the environment scheduling function, and other cores mainly process the sub-services of the business. When setting, the environment scheduling unit can estimate the processing delay of each sub-service according to the key information of each sub-service, such as the number of processing antennas, the target frequency domain data RB size, the CB block size, and the SNR, so it can be reasonably A core process assigned to process sub-services.
使用环境调度和业务调度分离的方法可以在各处理核间动态共享处理资源，如图4所示，假设环境调度逻辑承载在主核上，称为调度核，即为环境调度单元，具有环境调度中查询子业务处理状态和分配处理子业务的功能，其它两个核负责处理所述业务的各子业务，称为业务核1和业务核2。在子帧开始后，在执行前端抽样级处理子业务的过程中，环境调度单元即承载在主核上的环境调度逻辑收到业务调度单元的指示后，分配业务核1处理天线0(Ant0)和天线1(Ant1)的前端抽样级处理子业务，分配业务核2处理天线2(Ant2)和天线4(Ant4)的前端抽样级处理子业务，处理时延为94us，业务核1和业务核2处理完成后，向环境调度单元返回前端抽样级处理子业务完成的指示；在执行Symbol级处理子业务的过程中，环境调度单元收到业务调度单元的指示后，分配业务核1处理UE1的20RB及UE3的20RB，分配业务核2处理UE2的40RB，处理时延为72us，业务核1和业务核2处理完成后，向环境调度单元返回Symbol级处理子业务完成的指示。在执行Bit级处理子业务的过程中，环境调度单元收到业务调度单元的指示后，分配业务核1处理UE1的相关CB，分配业务核2处理UE3的相关CB，同时分配业务核1核和业务核2分别处理UE2的部分相关CB，换句话说，由业务核1和业务核2共同处理UE2的相关CB，处理时延为800us，业务核1和业务核2处理完成后，向环境调度单元返回Bit级处理子业务完成的指示，从而实现各子业务的处理。其中，在图4中，表示与UE1相关的处理，表示与UE2相关的处理，表示与UE3相关的处理。Using the method of separating environment scheduling and business scheduling can dynamically share processing resources between processing cores, as shown in Figure 4, assuming that the environment scheduling logic is carried on the main core, it is called the scheduling core, which is the environment scheduling unit and has environment scheduling The function of inquiring sub-service processing status and assigning and processing sub-services in the center, and the other two cores are responsible for processing each sub-service of the service, called service core 1 and service core 2. After the start of the subframe, in the process of processing the sub-services at the front-end sampling level, the environment scheduling unit, that is, the environment scheduling logic carried on the main core, receives instructions from the service scheduling unit, and assigns service core 1 to handle antenna 0 (Ant0) And the front-end sampling level of antenna 1 (Ant1) processes sub-services, assigns service core 2 to process the front-end sampling level of antenna 2 (Ant2) and antenna 4 (Ant4) to process sub-services, the processing delay is 94us, service core 1 and service core 2 After the processing is completed, return to the environment dispatching unit the indication of the completion of the front-end sampling level processing sub-service; in the process of executing the symbol-level processing sub-service, after the environment dispatching unit receives the instruction from the service dispatching unit, it assigns service core 1 to process UE1’s 20RB and 20RB of UE3 are allocated to service core 2 to process 40RB of UE2, and the processing delay is 72us. After the processing of service core 1 and service core 2 is completed, the symbol-level processing sub-service completion instruction is returned to the environment scheduling unit. In the process of executing Bit-level processing sub-services, after receiving the instruction from the service scheduling unit, the environment scheduling unit assigns service core 1 to process the relevant CB of UE1, allocates service core 2 to process the relevant CB of UE3, and at the same time allocates service core 1 core and Service core 2 separately processes some relevant CBs of UE2. In other words, service core 1 and service core 2 jointly process UE2's relevant CBs. The processing delay is 800us. After the processing of service core 1 and service core 2 is completed, it is dispatched to the environment The unit returns an indication of the completion of the Bit-level processing sub-service, thereby realizing the processing of each sub-service. Among them, in Figure 4, Indicates the processing related to UE1, Indicates the processing related to UE2, Indicates the processing related to UE3.
从上面的描述可以得出，基于本发明的方法，在图1所示的基于GPP的实现集中式基带处理池功能的系统中，调度模块包含了业务调度单元及环境调度单元；调度模块的功能主要包括：资源分配以及接口适配；其中，从分配资源和数字处理的角度来看，业务调度单元的处理只与具体的空口制式相关，而不涉及分配各处理核的处理任务，而环境调度单元负责分配各处理核的处理任务，以满足空口协议规定的实时性要求。这样划分后，使得业务调度单元更加专注于某种空口制式的算法接口适配和处理流程，从而脱离了具体的服务器，而环境调度单元则需要考虑各子业务调度处理流程的处理时间和分配原则等，环境调度单元在处理时，需要考虑平台资源的主频、处理器内核、以及所选操作系统等。其中，所述服务器是指业务调度单元所在的服务器，也就是说，所述服务器为承载业务调度单元对应的业务调度逻辑的硬件设备。From the above description, it can be drawn that based on the method of the present invention, in the system for realizing the centralized baseband processing pool function based on GPP shown in Fig. 1, the scheduling module includes a service scheduling unit and an environment scheduling unit; It mainly includes: resource allocation and interface adaptation; among them, from the perspective of resource allocation and digital processing, the processing of the service scheduling unit is only related to the specific air interface standard, and does not involve the allocation of processing tasks for each processing core, while the environment scheduling unit The unit is responsible for assigning the processing tasks of each processing core to meet the real-time requirements stipulated by the air interface protocol. After this division, the service scheduling unit is more focused on the algorithm interface adaptation and processing flow of a certain air interface standard, thus breaking away from the specific server, while the environment scheduling unit needs to consider the processing time and allocation principles of each sub-service scheduling process etc. When the environment scheduling unit is processing, it needs to consider the main frequency of the platform resource, the processor core, and the selected operating system. Wherein, the server refers to the server where the service scheduling unit is located, that is to say, the server is a hardware device that bears the service scheduling logic corresponding to the service scheduling unit.
具体地讲，业务调度单元与模块资源接口，它完成的主要功能包括：算法模块接口的适配、以及向环境调度单元申请资源分配给算法模块使用。业务调度单元不允许直接使用硬件资源，也不允许使用栈以外的任何存储器空间。在业务调度单元需要使用硬件资源和/或存储器空间时，需要通过环境调度单元进行分配并通过环境调度单元才能使用。Specifically, the business scheduling unit interfaces with the module resources, and its main functions include: adapting the interface of the algorithm module, and applying to the environment scheduling unit for resource allocation to the algorithm module. The service scheduling unit is not allowed to directly use hardware resources, nor is it allowed to use any memory space other than the stack. When the service scheduling unit needs to use hardware resources and/or memory space, it needs to be allocated by the environment scheduling unit and can only be used by the environment scheduling unit.
因此，在实际应用时，业务调度单元对应的业务调度逻辑可以在任意GPP平台和操作系统上使用，它仅和所实现的上层应用绑定。不同的空口协议比如全球移动通讯系统(GSM，Global System of Mobile communication)、TD-SCDMA、TDD-LTE等，可以有不同的业务调度逻辑，这些业务调度逻辑可以在同一个业务环境中混合执行，当新的处理器或操作系统出现时，这些业务调度逻辑可以以lib库的形式移植到新的GPP平台上而不作修改。Therefore, in actual application, the service scheduling logic corresponding to the service scheduling unit can be used on any GPP platform and operating system, and it is only bound to the implemented upper-layer application. Different air interface protocols, such as Global System of Mobile Communication (GSM, Global System of Mobile communication), TD-SCDMA, TDD-LTE, etc., can have different business scheduling logics, and these business scheduling logics can be mixed and executed in the same business environment. When a new processor or operating system appears, these business scheduling logics can be transplanted to the new GPP platform in the form of lib library without modification.
具体而言，业务调度单元可能完成的功能包括：事件管理、配置管理、业务数据统计、以及数据异常保护。其中，所述事件管理包括：算法实例的创建和释放、事件比如测量事件的启动和终止、算法实例的调用等；所述配置管理包括：在不同的信令或后台配置下的各算法实例配置消息的分发、和所执行业务的配置管理；所述业务数据统计包括：需上报的业务数据统计比如流量等、以及，某些需要做二次统计的业务数据的二次统计等；所述数据异常保护包括各类数据异常的异常处理，例如数据不配套等。Specifically, the functions that may be completed by the business scheduling unit include: event management, configuration management, business data statistics, and data exception protection. Wherein, the event management includes: the creation and release of algorithm instances, the start and termination of events such as measurement events, the invocation of algorithm instances, etc.; the configuration management includes: the configuration of each algorithm instance under different signaling or background configurations The distribution of messages, and the configuration management of the executed business; the business data statistics include: business data statistics that need to be reported, such as traffic, etc., and secondary statistics of some business data that need to be counted twice; the data Exception protection includes exception handling for various data exceptions, such as mismatched data.
环境调度单元与完成业务所需的所有资源接口，它完成的主要功能包括：平台资源、底层资源和硬件资源的接口适配、以及对资源包括时间片的分配和管理。环境调度单元不关心上层业务的具体应用，仅关心资源的使用情况。由于这种特性，环境调度单元需要控制所有的资源，因此需要提供特殊的应用程序编程接口(API，Application Programming Interface)供上层应用即业务调度单元和算法模块使用，以便业务调度单元和/或算法模块可以在需要的时候申请包括外部加速器和时间片在内的资源，同时由环境调度单元分配这些资源。The environment scheduling unit interfaces with all resources required to complete the business. Its main functions include: interface adaptation of platform resources, underlying resources and hardware resources, and allocation and management of resources including time slices. The environment scheduling unit does not care about the specific application of the upper layer business, but only cares about the usage of resources. Due to this characteristic, the environment scheduling unit needs to control all resources, so it is necessary to provide a special application programming interface (API, Application Programming Interface) for use by the upper layer application, that is, the business scheduling unit and algorithm module, so that the business scheduling unit and/or algorithm Modules can apply for resources including external accelerators and time slices when needed, and these resources are allocated by the environment scheduling unit.
从原理上来说，环境调度单元的功能相当于手动实现操作系统在业务管理方面的部分功能，这主要是基于无线协议物理层对算法处理的特殊实时性要求的考虑，比如：尽量减少任务切换、以及单核单线程绑定等特殊需求。In principle, the function of the environment scheduling unit is equivalent to manually implementing some of the functions of the operating system in terms of business management. This is mainly based on the consideration of the special real-time requirements for algorithm processing at the physical layer of the wireless protocol, such as: minimizing task switching, And special requirements such as single-core single-thread binding.
具体而言，环境调度单元可能完成的功能包括：消息管理、时序管理、任务管理、通信管理、时序和通信异常保护、以及运行状态统计。其中，所述消息管理包括管理消息队列，以驱动子业务的执行；所述时序管理包括：对各种资源的使用时序以及业务调度单元申请分配的时间片；所述任务管理包括：对各子业务、中断、信号量、以及邮箱等与操作系统相关的配置和管理；所述通信管理包括：CPU内部各核、CPU之间、以及跨服务器的各类通信的配置和管理；所述时序和通信异常保护包括：各类时序和通信的异常保护，比如子业务的执行时间片超出、数据传输丢失等；所述运行状态统计包括：对CPU运行的各类数据进行统计，具体包括消息触发次数、成功次数、以及失败次数等，以便调试需要。Specifically, the functions that the environment scheduling unit may complete include: message management, timing management, task management, communication management, timing and communication exception protection, and running status statistics. Wherein, the message management includes managing message queues to drive the execution of sub-services; the timing management includes: the usage timing of various resources and the time slices allocated by the service scheduling unit; the task management includes: The configuration and management related to the operating system such as business, interrupt, semaphore, and mailbox; the communication management includes: the configuration and management of various types of communication between CPU cores, between CPUs, and across servers; the timing and Communication anomaly protection includes: various timing and communication anomaly protections, such as sub-service execution time slice exceeding, data transmission loss, etc.; the operation status statistics include: statistics of various data of CPU operation, specifically including the number of message triggers , the number of successes, and the number of failures, etc., for debugging needs.
综上所述，本发明提供的资源调度方法，业务调度和资源调度进行了分离，由环境调度单元统一分配资源，如此，在处理实时信号时，能有效地实现资源的动态调度。并且，由于业务调度和资源调度进行了分离，因此，环境调度逻辑可以不局限于同一个空口协议，举个例子来说，LTE和TD-SCDMA的业务可以在同一个业务环境中混合执行，换句话说，在LTE和TD-SCDMA的业务处理过程中，当LTE或TD-SCDMA承载的用户量不足时，环境调度逻辑可以将资源调度给用户量大的业务来使用。To sum up, in the resource scheduling method provided by the present invention, business scheduling and resource scheduling are separated, and resources are allocated uniformly by the environment scheduling unit. In this way, dynamic scheduling of resources can be effectively realized when real-time signals are processed. Moreover, since service scheduling and resource scheduling are separated, the environment scheduling logic is not limited to the same air interface protocol. For example, LTE and TD-SCDMA services can be mixed and executed in the same service environment. In other words, during the service processing of LTE and TD-SCDMA, when the number of users carried by LTE or TD-SCDMA is insufficient, the environment scheduling logic can schedule resources for services with a large number of users.
另外，在划分子业务时，依据划分出的子业务之间的耦合相关性小的策略，将所述业务划分为一个以能独立运行的上子业务，如此，能使环境调度单元更加灵活地调度资源。In addition, when dividing the sub-services, according to the policy that the coupling correlation between the divided sub-services is small, the service is divided into an upper sub-service that can run independently. In this way, the environment scheduling unit can be more flexible Scheduling resources.
除此以外，在执行所述业务的各子业务过程中，所述环境调度单元根据确定的维护机制，实时维护各处理资源的任务队列，而业务调度单元不关心通信机制，也不关心各处理核之间的相互处理时序关系，仅调用算法模块，即仅完成某一空口制式的算法接口的适配和信号处理的优化，如此，提高了代码的可重用性，使得软件升级简单化，从而有利于系统的平滑演进。In addition, during the execution of each sub-business process of the business, the environment scheduling unit maintains the task queue of each processing resource in real time according to the determined maintenance mechanism, while the business scheduling unit does not care about the communication mechanism or each processing resource. The mutual processing timing relationship between the cores only calls the algorithm module, that is, only completes the adaptation of the algorithm interface of a certain air interface standard and the optimization of signal processing, thus improving the reusability of the code and simplifying the software upgrade. Conducive to the smooth evolution of the system.
环境调度单元实时维护各处理资源的任务队列，如此，最大化地保证了系统异常的可恢复性，从而具有更高的可靠性。The environment scheduling unit maintains the task queues of each processing resource in real time, thus maximizing the recoverability of system exceptions and thus having higher reliability.
由环境调度单元分配资源，减少了GPP平台的操作系统任务调度的额外开销，如此，有效地提高了处理效率。The resources are allocated by the environment scheduling unit, which reduces the additional overhead of the operating system task scheduling on the GPP platform, thus effectively improving the processing efficiency.
为实现上述方法，本发明还提供了一种资源调度装置，如图5所示，该装置包括业务调度单元51、以及环境调度单元52；其中，In order to realize the above method, the present invention also provides a resource scheduling device, as shown in FIG. 5 , the device includes a service scheduling unit 51 and an environment scheduling unit 52; wherein,
业务调度单元51，用于在执行划分为一个以上子业务的业务的各子业务过程中，当需要使用相应资源时，通知环境调度单元52；The business scheduling unit 51 is used to notify the environment scheduling unit 52 when corresponding resources need to be used during the execution of each sub-service of the business divided into more than one sub-service;
环境调度单元52，用于收到业务调度单元51的通知后，分配所需资源。The environment scheduling unit 52 is configured to allocate required resources after receiving the notification from the service scheduling unit 51 .
其中，所述环境调度单元52，具体用于：在分配所述所需资源时，根据资源的使用情况及所述所需资源的使用时间片，分配所述所需资源。Wherein, the environment scheduling unit 52 is specifically configured to: when allocating the required resources, allocate the required resources according to the resource usage and the usage time slice of the required resources.
所述环境调度单元52，还用于在执行所述业务的各子业务过程中，根据确定的维护机制，实时维护各处理资源的任务队列。The environment scheduling unit 52 is further configured to maintain the task queue of each processing resource in real time according to the determined maintenance mechanism during the execution of each sub-service of the service.
所述环境调度单元52，还用于在执行所述业务的各子业务过程中，释放未使用的处理资源。The environment scheduling unit 52 is further configured to release unused processing resources during execution of each sub-service process of the service.
基于上述资源调度装置，本发明还提供了一种资源调度系统，如图6所示，该系统包括：应用子系统61、平台资源62、硬件资源63、底层资源64、以及模块资源65；其中，应用方案61包括调度资源装置661，所述资源调度装置包括：业务调度单元6611、以及环境调度单元6612；其中，Based on the above resource scheduling device, the present invention also provides a resource scheduling system, as shown in Figure 6, the system includes: application subsystem 61, platform resource 62, hardware resource 63, underlying resource 64, and module resource 65; , the application solution 61 includes a scheduling resource device 661, and the resource scheduling device includes: a service scheduling unit 6611, and an environment scheduling unit 6612; wherein,
业务调度单元6611，用于在执行划分成一个以上子业务的业务的各子业务过程中，当需要使用相应资源时，通知环境调度单元6612；The business scheduling unit 6611 is configured to notify the environment scheduling unit 6612 when corresponding resources need to be used during the execution of each sub-service of the business divided into more than one sub-service;
环境调度单元6612，用于收到业务调度单元的通知后，分配所需资源。The environment scheduling unit 6612 is configured to allocate required resources after receiving the notification from the service scheduling unit.
其中，平台资源62、硬件资源63、底层资源64、以及模块资源65的功能与图1中相应的各模块的功能相同，这里不再赘述。这里，所述平台资源62为GPP平台。Among them, the functions of platform resource 62 , hardware resource 63 , underlying resource 64 , and module resource 65 are the same as those of the corresponding modules in FIG. 1 , and will not be repeated here. Here, the platform resource 62 is a GPP platform.
所述环境调度单元6612，具体用于：在分配所述所需资源时，根据资源的使用情况及所述所需资源的使用时间片，分配所述所需资源。The environment scheduling unit 6612 is specifically configured to: when allocating the required resources, allocate the required resources according to the resource usage and the usage time slice of the required resources.
所述环境调度单元6612，还用于在执行所述业务的各子业务过程中，根据确定的维护机制，实时维护各处理资源的任务队列。The environment scheduling unit 6612 is further configured to maintain the task queue of each processing resource in real time according to the determined maintenance mechanism during the execution of each sub-service of the service.
所述环境调度单元6612，还用于在执行所述业务的各子业务过程中，释放未使用的处理资源。The environment scheduling unit 6612 is further configured to release unused processing resources during execution of each sub-service of the service.
所述模块资源65进一步包括算法模块651，所述业务调度单元6611，还用于在执行所述业务的各子业务过程中，所述业务调度单元与算法模块651进行接口适配，完成各子业务处理。The module resource 65 further includes an algorithm module 651, and the service scheduling unit 6611 is also used to perform interface adaptation between the service scheduling unit and the algorithm module 651 during the execution of each sub-service process of the service, and complete each sub-service business processing.
所述平台资源62、所述硬件资源63、以及所述底层资源64组成资源调度系统的资源，用于被所述环境调度单元6612控制。The platform resources 62 , the hardware resources 63 , and the underlying resources 64 constitute the resources of the resource scheduling system, and are used to be controlled by the environment scheduling unit 6612 .
以上所述，仅为本发明的较佳实施例而已，并非用于限定本发明的保护范围。The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310157089.8ACN104123185A (en) | 2013-04-28 | 2013-04-28 | Resource scheduling method, device and system |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310157089.8ACN104123185A (en) | 2013-04-28 | 2013-04-28 | Resource scheduling method, device and system |
| Publication Number | Publication Date |
|---|---|
| CN104123185Atrue CN104123185A (en) | 2014-10-29 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310157089.8APendingCN104123185A (en) | 2013-04-28 | 2013-04-28 | Resource scheduling method, device and system |
| Country | Link |
|---|---|
| CN (1) | CN104123185A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104680432A (en)* | 2015-03-24 | 2015-06-03 | 国家电网公司 | Method and device for determining service bearing condition of dispatching speciality |
| CN106776002A (en)* | 2016-11-15 | 2017-05-31 | 华为技术有限公司 | The communication means and device of the virtualization hardware framework of FPGA |
| CN106804054A (en)* | 2015-11-26 | 2017-06-06 | 中兴通讯股份有限公司 | A kind of method and device for virtualizing the shared transfer resource of base station access network network |
| CN106815061A (en)* | 2015-12-01 | 2017-06-09 | 阿里巴巴集团控股有限公司 | A kind of method for processing business and device |
| CN109428889A (en)* | 2017-08-30 | 2019-03-05 | 英特尔公司 | Technology for the machine learning scheme in the switching at runtime between adaptive connection and connection optimization |
| CN110838990A (en)* | 2018-08-17 | 2020-02-25 | 上海诺基亚贝尔股份有限公司 | A method and apparatus for accelerating layer 1 in C-RAN |
| CN112752304A (en)* | 2019-10-31 | 2021-05-04 | 上海华为技术有限公司 | Method and related device for processing uplink reference signal |
| CN115421917A (en)* | 2022-09-15 | 2022-12-02 | 天津津航计算技术研究所 | A load balancing method for embedded system multiprocessor modules |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102467415A (en)* | 2010-11-03 | 2012-05-23 | 大唐移动通信设备有限公司 | Business surface task processing method and device |
| CN102541640A (en)* | 2011-12-28 | 2012-07-04 | 厦门市美亚柏科信息股份有限公司 | Cluster GPU (graphic processing unit) resource scheduling system and method |
| CN102567086A (en)* | 2010-12-30 | 2012-07-11 | 中国移动通信集团公司 | Task scheduling method, equipment and system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102467415A (en)* | 2010-11-03 | 2012-05-23 | 大唐移动通信设备有限公司 | Business surface task processing method and device |
| CN102567086A (en)* | 2010-12-30 | 2012-07-11 | 中国移动通信集团公司 | Task scheduling method, equipment and system |
| CN102541640A (en)* | 2011-12-28 | 2012-07-04 | 厦门市美亚柏科信息股份有限公司 | Cluster GPU (graphic processing unit) resource scheduling system and method |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104680432A (en)* | 2015-03-24 | 2015-06-03 | 国家电网公司 | Method and device for determining service bearing condition of dispatching speciality |
| CN106804054A (en)* | 2015-11-26 | 2017-06-06 | 中兴通讯股份有限公司 | A kind of method and device for virtualizing the shared transfer resource of base station access network network |
| CN106815061A (en)* | 2015-12-01 | 2017-06-09 | 阿里巴巴集团控股有限公司 | A kind of method for processing business and device |
| CN106776002A (en)* | 2016-11-15 | 2017-05-31 | 华为技术有限公司 | The communication means and device of the virtualization hardware framework of FPGA |
| CN106776002B (en)* | 2016-11-15 | 2020-09-25 | 华为技术有限公司 | Communication method and device for virtualized hardware architecture of FPGA |
| CN109428889A (en)* | 2017-08-30 | 2019-03-05 | 英特尔公司 | Technology for the machine learning scheme in the switching at runtime between adaptive connection and connection optimization |
| CN110838990A (en)* | 2018-08-17 | 2020-02-25 | 上海诺基亚贝尔股份有限公司 | A method and apparatus for accelerating layer 1 in C-RAN |
| CN112752304A (en)* | 2019-10-31 | 2021-05-04 | 上海华为技术有限公司 | Method and related device for processing uplink reference signal |
| CN112752304B (en)* | 2019-10-31 | 2022-08-26 | 上海华为技术有限公司 | Method and related device for processing uplink reference signal |
| US12068995B2 (en) | 2019-10-31 | 2024-08-20 | Huawei Technologies Co., Ltd. | Method for processing uplink reference signal and related apparatus |
| CN115421917A (en)* | 2022-09-15 | 2022-12-02 | 天津津航计算技术研究所 | A load balancing method for embedded system multiprocessor modules |
| Publication | Publication Date | Title |
|---|---|---|
| CN104123185A (en) | Resource scheduling method, device and system | |
| EP3962032B1 (en) | Method and apparatus for balancing server load in cloud ran systems | |
| Kan et al. | Task offloading and resource allocation in mobile-edge computing system | |
| CN110913409B (en) | Network configuration method, device and system | |
| US11894992B2 (en) | Network component management method and network device | |
| CN103906257B (en) | LTE wide-band communication system computing resource schedulers and its dispatching method based on GPP | |
| CN103916909B (en) | Baseband Pool System | |
| JP7607557B2 (en) | Organizer and interconnection fabric mapper for virtual wireless base stations - Patents.com | |
| KR20180123379A (en) | Method and apparatus for allocating resource in software defined network | |
| WO2015165095A1 (en) | Method for creating virtual base station, and base station cloud device | |
| CN102438338B (en) | Base station based on multicore general processor for broadband mobile communication system | |
| CN101330711A (en) | Baseband Resource Pool Management Method and System for Global System for Mobile Communications | |
| CN102668613A (en) | Method and apparatus for realizing base band resource pool in lte base station | |
| CN102740326B (en) | Managing method and device for processing resources in BBU | |
| CN110708708B (en) | Wireless resource optimization method and device | |
| CN103501498B (en) | A kind of baseband processing resource distribution method and its device | |
| CN113905449B (en) | Computing resource scheduling method, system and equipment | |
| CN107426109A (en) | A kind of traffic scheduling method, VNF modules and flow scheduling server | |
| CN102870482A (en) | Method and apparatus for traffic channel assignment | |
| CN104954417B (en) | Method and apparatus for handling data | |
| CN105517176B (en) | The method of dynamic dispatching virtualization base station resource | |
| CN111538564A (en) | Wireless device sharing method and device | |
| CN114675972B (en) | Cloud network resource flexible scheduling method and system based on integral algorithm | |
| CN109348511A (en) | A method and device for allocating baseband resources | |
| CN115884161A (en) | Resource scheduling method and related communication device |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication | Application publication date:20141029 |