技术领域technical field
本发明涉及一种数据通信技术领域的方法，具体地，涉及一种适用于动态自适应视频流媒体的分布式缓存放置方法。The invention relates to a method in the technical field of data communication, in particular to a method for placing distributed caches suitable for dynamic adaptive video streaming media.
背景技术Background technique
随着移动数据流量的迅速增长以及智能终端设备的日益普及，以移动视频服务为代表的无线视频流媒体技术在近几年得到了越来越广泛的应用。与此同时，移动用户在所使用的移动设备终端、点播内容以及网络连接性等方面呈现出更为复杂的异构特性。动态自适应流媒体技术能够为用户提供同一视频内容的不同版本，以提高异构网络中用户的视频观看满意度。其中，每个视频版本以给定的码率和/或分辨率进行编码，从而使得各用户能够根据自己的视频点播需求和网络状况来确定下载最合适的视频版本。With the rapid growth of mobile data traffic and the increasing popularity of smart terminal devices, wireless video streaming technology represented by mobile video services has been more and more widely used in recent years. At the same time, mobile users present more complex heterogeneous characteristics in terms of mobile device terminals, on-demand content, and network connectivity. Dynamic adaptive streaming media technology can provide users with different versions of the same video content, so as to improve the video viewing satisfaction of users in heterogeneous networks. Wherein, each video version is encoded with a given bit rate and/or resolution, so that each user can determine and download the most suitable video version according to their own video-on-demand requirements and network conditions.
另一方面，网络视频流量呈现出高度的时间变化特性，具体体现在高峰时间段的网络拥塞以及在低谷时间段的网络使用率不足。为了缓解高峰时间段视频流量拥塞的情况，在边缘服务器处的缓存操作能够利用边缘服务器的存储能力在非高峰时间段预先缓存特定的视频内容，从而起到平滑视频流量的时间变化特性以及降低网络拥塞和传输延迟的作用。相比于主服务器，边缘服务器更靠近移动用户，因此边缘服务器处的缓存还可以极大地缓解主服务器处的视频服务负载，并且通过边缘服务器到用户之间的本地高速链路将视频内容以更低的延迟传输至移动用户。On the other hand, network video traffic exhibits highly time-varying characteristics, specifically reflected in network congestion during peak hours and underutilized network during off-peak hours. In order to alleviate the video traffic congestion during peak hours, the cache operation at the edge server can utilize the storage capacity of the edge server to pre-cache specific video content during off-peak hours, thereby smoothing the time-varying characteristics of video traffic and reducing network traffic. The role of congestion and transmission delay. Compared with the main server, the edge server is closer to the mobile user, so the cache at the edge server can also greatly relieve the video service load at the main server, and the video content can be delivered at a faster rate through the local high-speed link between the edge server and the user. Low latency delivery to mobile users.
经过对现有技术的检索发现，Y.Jin等人在《IEEE Transactions on Circuits and Systemsfor Video Technology，Dec.2015，pp.1914-1925，(电气电子工程师协会用于视频技术的电路与系统学报，2015年12月，第1914-1925页)》上发表了题为“Optimal transcodingand caching for adaptive streaming in media cloud:An analytical approach(用于媒体云自适应流媒体的最优转码与缓存：一种解析方法)”的文章，该文章将缓存机制引入到动态自适应视频流媒体中，研究了媒体云中的最优转码和缓存资源分配问题，以最小化视频传输的总体操作代价。但是，该文章主要基于一个移动用户连接单个边缘服务器的场景假设，为各边缘服务器之间的进一步协作以提升缓存性能留下了提升空间。After searching the prior art, it was found that Y.Jin et al. published in "IEEE Transactions on Circuits and Systems for Video Technology, Dec.2015, pp.1914-1925, (Journal of Circuits and Systems for Video Technology of the Institute of Electrical and Electronics Engineers, In December 2015, pages 1914-1925) published a paper entitled "Optimal transcoding and caching for adaptive streaming in media cloud: An analytical approach (optimal transcoding and caching for media cloud adaptive streaming media: a Parsing method)", which introduces the caching mechanism into dynamic adaptive video streaming media, and studies the optimal transcoding and caching resource allocation in the media cloud to minimize the overall operation cost of video transmission. However, this article is mainly based on the scenario assumption that a mobile user connects to a single edge server, leaving room for further cooperation among edge servers to improve cache performance.
经检索还发现，K.Shanmugam等人在《IEEE Transactions on Information Theory，Dec.2013，pp.8402-8413，(电气电子工程师协会信息论学报，2013年12月，第8402-8413页)》上发表了题为“Femtocaching:Wireless content delivery through distributed cachinghelpers(Femto缓存：借助分布式缓存协助者的无线内容传递)”的文章，该文章针对单个移动用户能够同时连接多个边缘服务器的场景，研究了网络中分布式边缘服务器的视频内容缓存问题。通过各边缘服务器之间的协作，视频服务负载从主服务器段转移到各边缘服务器，从而实现了用户平均视频下载延迟的最小化。但是，该文章所涉及的视频内容仅具有单一码率版本，无法适应异构网络中用户的差异化带宽情况以及点播需求。After searching, it was also found that K. Shanmugam et al published in "IEEE Transactions on Information Theory, Dec.2013, pp.8402-8413, (Institute of Electrical and Electronics Engineers Information Theory Journal, December 2013, pp. 8402-8413)" Published an article entitled "Femtocaching: Wireless content delivery through distributed cachinghelpers (Femto caching: wireless content delivery through distributed caching helpers)", which studies the network Video content caching problem in distributed edge servers. Through the cooperation between each edge server, the video service load is shifted from the main server segment to each edge server, thereby realizing the minimization of user average video download delay. However, the video content involved in this article only has a single bit rate version, which cannot adapt to the differentiated bandwidth conditions and on-demand requirements of users in heterogeneous networks.
此外，上述工作仅考虑了边缘服务器缓存在操作代价或码率方面的性能，将所有视频视为同一数据文件看待，而忽略了不同视频所具有的不同内容信息(例如，不同视频内容具有不同的码率-失真性能)，从而在一定程度上导致了在缓存视频流媒体时系统整体性能的下降。In addition, the above work only considers the performance of edge server cache in terms of operation cost or bit rate, treats all videos as the same data file, and ignores the different content information of different videos (for example, different video contents have different rate-distortion performance), which to some extent leads to a decrease in the overall performance of the system when caching video streaming media.
发明内容Contents of the invention
针对现有技术中的缺陷，本发明的目的是提供一种适用于动态自适应视频流媒体的分布式缓存放置方法。Aiming at the defects in the prior art, the purpose of the present invention is to provide a distributed buffer placement method suitable for dynamic adaptive video streaming media.
为实现以上目的，本发明采用的技术方案是：结合主服务器处的动态自适应流媒体编码技术将各视频编码为多个不同码率的版本，同时兼顾了各不同视频内容之间码率-失真性能的差异、边缘服务器的缓存容量限制、不同用户的网络连接情况以及视频点播概率分布，采用分布式的缓存优化放置方法确定各边缘服务器所需缓存的视频版本子集，最终实现用户通过边缘服务器下载观看视频整体质量的最大化。本发明提高了边缘服务器缓存视频内容的利用率，减轻了主服务器处的视频流媒体服务负载，为用户提供更佳的视频服务质量。In order to achieve the above purpose, the technical solution adopted by the present invention is: each video is encoded into multiple versions with different bit rates in combination with the dynamic adaptive streaming media encoding technology at the main server, and at the same time, the bit rate between different video contents is taken into account. Distortion performance differences, cache capacity limitations of edge servers, network connections of different users, and video-on-demand probability distributions, use a distributed cache optimization placement method to determine the subset of video versions that each edge server needs to cache, and finally realize the user through the edge The server downloads and watches the video to maximize the overall quality. The invention improves the utilization rate of video content cached by the edge server, reduces the video stream media service load at the main server, and provides users with better video service quality.
本发明提供一种适用于动态自适应视频流媒体的分布式缓存放置方法，包括以下步骤：The present invention provides a distributed cache placement method suitable for dynamic adaptive video streaming media, comprising the following steps:
第一步，在主服务器处，使用动态自适应流媒体编码技术将各视频编码为多个不同码率的视频版本，并且得到各不同视频内容之间码率-失真性能的差异；In the first step, at the main server, each video is encoded into a plurality of video versions with different bit rates using dynamic adaptive streaming encoding technology, and the difference in bit rate-distortion performance between different video contents is obtained;
第二步，在边缘服务器处，根据边缘服务器缓存容量限制预先缓存具有特定码率的视频版本，以通过本地高速链路服务用户的视频点播请求；In the second step, at the edge server, the video version with a specific bit rate is pre-cached according to the limit of the edge server cache capacity, so as to serve the user's video-on-demand request through the local high-speed link;
第三步，在用户处，根据不同用户的点播需求和差异化网络条件，从与用户相邻的边缘服务器上选择下载最高码率的视频版本；The third step is to select and download the video version with the highest bit rate from the edge server adjacent to the user according to the on-demand requirements of different users and the differentiated network conditions at the user's place;
第四步，基于前三个步骤，采用参数：主服务器处编码得到的视频不同版本组成的全集、边缘服务器的缓存容量限制、以及用户的网络连接情况和视频点播概率分布，建立适用于动态自适应视频流媒体的分布式缓存放置的优化问题，采用快速高效的分布式缓存内容放置方法，得到第二步中各边缘服务器缓存的最优视频版本子集，所述最优视频版本子集约束了第三步中各用户能够下载的最高码率的视频版本的码率大小。The fourth step, based on the first three steps, using parameters: the full set of different versions of the video coded at the main server, the limit of the cache capacity of the edge server, and the user's network connection and video-on-demand probability distribution, to establish a dynamic self-supporting system. Adapt to the optimization problem of distributed cache placement of video streaming media, adopt a fast and efficient distributed cache content placement method, and obtain the optimal video version subset cached by each edge server in the second step, and the optimal video version subset constraint The code rate of the video version with the highest code rate that each user can download in the third step is determined.
优选地，第一步中，所述的主服务器能够使用动态自适应流媒体编码技术将任意一个视频文件编码为多个具有不同编码码率的视频版本。所述的各视频文件由于内容的不同呈现出不同的码率-失真性能。Preferably, in the first step, the main server can use dynamic adaptive streaming media encoding technology to encode any video file into multiple video versions with different encoding bit rates. The video files described above exhibit different rate-distortion performance due to different contents.
优选地，第二步中，所述的边缘服务器能够预先缓存具有特定码率的视频版本，所述边缘服务器缓存的视频版本总大小受到其物理缓存容量的限制。所述的边缘服务器比主服务器更靠近用户，因此可以通过对无线资源的高密度空间复用建立与用户之间的高速本地链路，从而更快速地响应和服务用户的视频点播请求。Preferably, in the second step, the edge server can pre-cache the video version with a specific bit rate, and the total size of the video version cached by the edge server is limited by its physical cache capacity. The edge server is closer to the user than the main server, so it can establish a high-speed local link with the user through high-density spatial multiplexing of wireless resources, thereby responding and serving the user's video-on-demand request more quickly.
优选地，第三步中，所述的用户可以向相邻的边缘服务器提出视频点播请求，并且从邻近的边缘服务器选择下载具有高码率的视频版本。确定下载最高码率的视频版本的准则为：首先查询所请求视频的最高码率版本是否预先缓存在与用户相邻的边缘服务器集合中，如果存在，则从所有缓存该最高码率的视频版本的边缘服务器中选择具有最高下载带宽的边缘服务器处下载该视频版本；如果不存在，则对该视频第二高码率的版本进行查询；以此类推直到在邻近边缘服务器集合中发现用户所请求的视频的某一码率版本为止；如果用户在任何邻近的边缘服务器都无法找到缓存的该视频的任何码率版本，则将选择从主服务器下载该视频。Preferably, in the third step, the user can make a video-on-demand request to an adjacent edge server, and choose to download a video version with a high bit rate from the adjacent edge server. The criterion for determining the video version with the highest bit rate to download is: first check whether the highest bit rate version of the requested video is pre-cached in the set of edge servers adjacent to the user, and if so, cache the video version with the highest bit rate from all Select the edge server with the highest download bandwidth to download the video version; if it does not exist, query the version with the second highest bit rate of the video; and so on until the user's request is found in the set of adjacent edge servers If the user cannot find any cached version of the video at any nearby edge server, the video will be downloaded from the main server.
优选地，第四步中，所述的适用于动态自适应视频流媒体的分布式缓存放置的优化问题结合了主服务器处编码得到的视频不同版本组成的全集、边缘服务器的缓存容量限制、以及用户的网络连接情况和视频点播概率分布，并使用网络效用最大化建模方法得到。Preferably, in the fourth step, the optimization problem of distributed cache placement suitable for dynamic adaptive video streaming media combines the full set of different versions of video encoded at the main server, the cache capacity limit of the edge server, and The user's network connection status and video-on-demand probability distribution are obtained by using the network utility maximization modeling method.
优选地，第四步中，所述的分布式缓存内容放置方法在为各边缘服务器确定具体的缓存视频版本子集时，采用具有多项式时间复杂度和高近似优化性能的高成本效益贪心算法，最终快速高效地实现各边缘服务器分布式缓存内容的优化放置。Preferably, in the fourth step, when the described distributed cache content placement method determines a specific subset of cached video versions for each edge server, it adopts a highly cost-effective greedy algorithm with polynomial time complexity and high approximate optimization performance, Finally, the optimized placement of the distributed cache content of each edge server can be realized quickly and efficiently.
更优选地，第四步中，所述的分布式缓存内容放置方法，具体执行步骤为：More preferably, in the fourth step, the specific execution steps of the described distributed cache content placement method are:
(a)初始化：设置初始局部最优解集合为空集、初始搜索集合为视频不同版本组成的全集，以及初始步数为1；(a) Initialization: set the initial local optimal solution set to an empty set, the initial search set to a full set composed of different versions of the video, and the initial step number to 1;
(b)迭代搜索步骤：根据已有的局部最优解集合，在剩余搜索集合中寻找使得边际增量与码率代价的比值最大的元素，所述剩余搜索集合为局部最优解对于搜索集合的补集，所述剩余搜索集合中的一个元素对应于在某一个边缘服务器上缓存某一视频的某一码率版本；(b) Iterative search step: according to the existing local optimal solution set, find the element that makes the ratio of the marginal increment to the code rate cost maximum in the remaining search set, and the remaining search set is the local optimal solution for the search set Complementary set, an element in the remaining search set corresponds to a certain code rate version of a certain video cached on a certain edge server;
(c)更新步骤：如果添加步骤(b)中搜索得到的元素仍能满足各边缘服务器的缓存容量约束，则将该元素添加至局部最优解集合，并且搜索集合保持不变；如果添加该元素不能满足各边缘服务器的缓存容量约束，则局部最优解集合保持不变，并且将该元素从搜索集合中移除；(c) Update step: If the element searched in adding step (b) still satisfies the cache capacity constraints of each edge server, then add the element to the local optimal solution set, and the search set remains unchanged; if adding the If the element cannot meet the cache capacity constraints of each edge server, the local optimal solution set remains unchanged, and the element is removed from the search set;
(d)判定步骤：如果剩余搜索集合不是空集，则令搜索步数加一并返回迭代搜索步骤；否则，停止迭代并且将当前的局部最优解集合输出为最优结果。(d) Judgment step: If the remaining search set is not an empty set, add one to the number of search steps and return to the iterative search step; otherwise, stop the iteration and output the current local optimal solution set as the optimal result.
与现有技术相比，本发明具有如下的有益效果：Compared with the prior art, the present invention has the following beneficial effects:
本发明为适应动态自适应流媒体技术的需要，提供了一种完全分布式的缓存放置方法，提高了边缘服务器缓存视频内容的利用率，减轻了主服务器处的视频流媒体服务负载，为用户提供更佳的视频服务质量。In order to meet the needs of dynamic self-adaptive streaming media technology, the present invention provides a fully distributed cache placement method, which improves the utilization rate of video content cached by edge servers, reduces the video streaming service load at the main server, and provides users with Provide better video service quality.
附图说明Description of drawings
通过阅读参照以下附图对非限制性实施例所作的详细描述，本发明的其它特征、目的和优点将会变得更明显：Other characteristics, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:
图1为本发明一实施例的方法流程图；Fig. 1 is a method flowchart of an embodiment of the present invention;
图2为本发明一实施例分布式缓存网络的示意图；FIG. 2 is a schematic diagram of a distributed cache network according to an embodiment of the present invention;
图3为本发明一实施例用户从邻近边缘服务器确定下载最大码率视频版本的流程图；Fig. 3 is a flow chart of the user determining to download the maximum code rate video version from the adjacent edge server according to an embodiment of the present invention;
图4为本发明一实施例分布式缓存放置方法的流程图；FIG. 4 is a flowchart of a method for placing a distributed cache according to an embodiment of the present invention;
图5为本发明一实施例网络设置的示意图；FIG. 5 is a schematic diagram of network settings according to an embodiment of the present invention;
图6为本发明一实施例视频码率失真性能的示意图；FIG. 6 is a schematic diagram of video rate-distortion performance according to an embodiment of the present invention;
图7为本发明一实施例分布式缓存放置性能的示意图。FIG. 7 is a schematic diagram of distributed cache placement performance according to an embodiment of the present invention.
具体实施方式detailed description
下面结合具体实施例对本发明进行详细说明。以下实施例将有助于本领域的技术人员进一步理解本发明，但不以任何形式限制本发明。应当指出的是，对本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进。这些都属于本发明的保护范围。The present invention will be described in detail below in conjunction with specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention. These all belong to the protection scope of the present invention.
参照图1所示，一种适用于动态自适应视频流媒体的分布式缓存放置方法的流程，具体实现包括如下步骤：Referring to Fig. 1, a process of a distributed cache placement method suitable for dynamic adaptive video streaming media, the specific implementation includes the following steps:
1、主服务器处的动态自适应流媒体编码1. Dynamic adaptive streaming media encoding at the main server
如图2所示，对分布式缓存网络进行实例分析，假设在主服务器处存放了F个视频文件(F为任意大于2的正整数)，记为视频文件集合每个视频文件的播放时间长度均为T。使用动态自适应流媒体编码技术，将任意一个视频文件编码为M个不同码率的视频版本(M为任意大于2的正整数)，记为视频版本集合其中第m个视频版本fm的编码码率记为Rfm，并且该集合按照编码码率呈降序排列，即从而，包含所有F个视频文件的全部版本的完整集合可以记为另一方面，主服务器掌握所有视频文件的码率-失真性能信息，即Dmax-Df(R)表示视频文件f在编码码率为R时所对应的视频失真，其中Dmax和Df(R)分别表示当视频无法解码时的最大失真常量以及在成功解码视频文件f的码率为R的版本之后的视频失真减少量。As shown in Figure 2, an example analysis of the distributed cache network is performed, assuming that F video files are stored at the main server (F is any positive integer greater than 2), which is recorded as a video file set The playing time length of each video file is T. Use dynamic adaptive streaming media coding technology to convert any video file Coded into M video versions with different bit rates (M is any positive integer greater than 2), recorded as a video version set The coding bit rate of the mth video version fm is denoted as Rfm , and the set is arranged in descending order according to the coding bit rate, namely Thus, the complete set containing all versions of all F video files can be written as On the other hand, the main server grasps the rate-distortion performance information of all video files, that is, Dmax -Df (R) represents the video distortion corresponding to video file f when the encoding bit rate is R, where Dmax and Df (R) respectively denote the maximum distortion constant when the video cannot be decoded and the reduction in video distortion after successfully decoding a version of the video file f with bitrate R.
2、边缘服务器处的分布式缓存2. Distributed cache at the edge server
图2还示出在网络中分布的S个边缘服务器，记为边缘服务器集合这些边缘服务器能够预先缓存具有特定码率的视频版本，由于它们比主服务器更靠近用户，因此可以通过对无线资源的高密度空间复用建立与用户之间的高速本地链路，从而更快速地响应和服务用户的视频点播请求。对于每个边缘服务器而言，它能够从主服务器处预先获取并缓存的视频版本数量受到其物理存储容量Bs的限制。定义基集合表示所有可能在边缘服务器s处缓存的视频版本的全集，其中元素表示边缘服务器s缓存视频文件f的第m个码率版本。Figure 2 also shows S edge servers distributed in the network, denoted as edge server set These edge servers can pre-cache the video version with a specific bit rate. Since they are closer to the user than the main server, they can establish a high-speed local link with the user through high-density spatial multiplexing of wireless resources, thereby enabling faster Respond to and serve video-on-demand requests from users. For each edge server In terms of , the number of video versions it can pre-fetch and cache from the main server is limited by its physical storage capacity Bs . Define the base collection represents the full set of all possible video versions cached at the edge server s, where the element Indicates that the edge server s caches the m-th bit rate version of the video file f.
3、用户处的视频请求与下载3. Video requests and downloads from users
图2还示出网络中随机分布的U个用户，记为用户集合u＝{1，2，...，U}。用户可以向相邻的边缘服务器提出视频点播请求，并且从邻近的边缘服务器选择下载具有最大码率(即最佳质量)的视频版本。对于每个用户而言，将通过无线链路与其相连的边缘服务器集合记为并且将集合中的各边缘服务器按照与用户u之间无线链路的下载带宽降序排列，从而使得表示与用户u之间无线链路的下载带大小宽排在第i位的边缘服务器。FIG. 2 also shows U users randomly distributed in the network, denoted as user set u={1, 2, . . . , U}. A user can make a video-on-demand request to an adjacent edge server, and choose to download a video version with the largest bit rate (ie, the best quality) from the adjacent edge server. for each user In terms of , the set of edge servers connected to it through wireless links is denoted as and will assemble Each edge server in is arranged in descending order according to the download bandwidth of the wireless link with user u, so that Indicates the edge server that ranks i in the download bandwidth of the wireless link with user u.
图3示出了用户从邻近边缘服务器确定下载视频版本的流程图。如图3所示，当用户u提出视频文件f的点播请求之后，首先查询该视频的最高码率版本f1是否预先缓存在与其相邻的边缘服务器集合中，如果存在，则从所有缓存该视频版本的边缘服务器中选择具有最高下载带宽的边缘服务器处下载该视频版本；如果不存在，则对该视频第二高码率的版本f2进行查询；以此类推直到在中发现该视频的某一码率版本为止。如果用户u在中的任何边缘服务器都无法找到缓存的该视频的任何码率版本，则将选择从主服务器下载该视频。FIG. 3 shows a flow chart for a user to determine a video version to download from a neighboring edge server. As shown in Figure 3, when user u makes an on-demand request for video file f,first check whether the highest bit rate version f1 of the video is pre-cached in the set of edge servers adjacent to it , if it exists, then select the edge server with the highest download bandwidth from all the edge servers that cache the video version to download the video version; if it does not exist, query thesecond highest bit rate version f2 of the video; and so on until the Until a certain bit rate version of the video is found in the If user u is in If none of the edge servers in the .
4、建立适用于动态自适应视频流媒体的分布式缓存放置的优化问题，提出快速高效的分布式缓存内容放置方法4. Establish the optimization problem of distributed cache placement suitable for dynamic adaptive video streaming media, and propose a fast and efficient distributed cache content placement method
建立适用于动态自适应视频流媒体的分布式缓存放置的优化问题如下(其中每个参数的含义可在上下文中对应获取)：The optimization problem of establishing a distributed cache placement suitable for dynamic adaptive video streaming media is as follows (the meaning of each parameter can be obtained correspondingly in the context):
目标优化问题：Objective optimization problem:
约束条件：Restrictions:
其中，优化变量为：表示在各个边缘服务器上缓存的视频版本组成的集合。具体地，某一个元素表示视频文件f的第m个码率版本缓存在边缘服务器s上。Among them, the optimization variables are: Represents a collection of video versions cached on each edge server. Specifically, an element Indicates that the m-th bit rate version of the video file f is cached on the edge server s.
优化目标为：最大化所有用户的预期视频失真减少量之和其中，是所述优化问题的基集合，则表示已知用户u请求视频文件f的概率为Pu，f，基于图3所示的下载流程，通过边缘服务器集合响应用户u的所有视频点播请求并且将所需视频版本传输给用户u以实现的预期视频失真减少量，即The optimization objective is: to maximize the sum of the expected video distortion reductions for all users in, is the basis set of the optimization problem, It means that the probability that the known user u requests video file f is Pu, f , based on the download process shown in Figure 3, through the collection of edge servers The expected reduction in video distortion achieved by responding to all video-on-demand requests from user u and delivering the desired video version to user u, namely
上式中的指示函数的值在的时候为1，在的时候为0，其中表示与用户u之间无线链路的下载带大小宽排在第j位的边缘服务器缓存了视频文件f的第m个码率版本。The indicator function in the above formula The value of When it is 1, in When is 0, where Indicates that the edge server ranked j in the download bandwidth of the wireless link with user u has cached the m bit rate version of the video file f.
约束条件为：边缘服务器物理缓存限制条件，即要求任意一个边缘服务器上缓存的所有视频版本的大小之和不超过其物理存储容量Bs。The constraints are: edge server physical cache constraints, that is, any edge server is required The sum of the sizes of all video versions cached on the Internet does not exceed its physical storage capacity Bs .
如图4所示，给出具有多项式时间复杂度和高近似优化性能的高成本效益贪心算法，最终快速高效地实现各边缘服务器分布式缓存内容的优化放置。所述分布式缓存放置方法的执行过程如下(其中每个参数的含义可在上下文中对应获取)：As shown in Figure 4, a cost-effective greedy algorithm with polynomial time complexity and high approximate optimization performance is given, and finally the optimal placement of the distributed cache content of each edge server can be realized quickly and efficiently. The execution process of the distributed cache placement method is as follows (the meaning of each parameter can be obtained correspondingly in the context):
(a)初始化：设置初始局部最优解集合初始搜索集合v0＝v，初始步数t＝1。(a) Initialization: set the initial local optimal solution set The initial search set v0 =v, the initial step number t=1.
(b)迭代搜索步骤(t＝1，2，3，...)：(b) Iterative search steps (t=1, 2, 3, . . . ):
根据已有的局部最优解集合在剩余搜索集合(即对于Vt-1的补集)中寻找使得边际增量与码率代价的比值最大的元素即According to the existing local optimal solution set In the rest of the search collection (which is For the complement of Vt-1 ) to find such that the marginal increment and bit rate cost The element with the largest ratio which is
上式中，元素表示在边缘服务器s上缓存视频文件f的第m个码率版本，元素表示在第t个搜索步骤中搜索得到的边际增量与码率代价的比值最大的元素，对应于在边缘服务器st上缓存视频文件ft的第mt个码率版本，表示视频文件f的第m个码率版本的编码码率，T表示视频文件的时间长度。In the above formula, the element Indicates that the m-th bit rate version of the video file f is cached on the edge server s, the element Indicates the element with the largest ratio of marginal increment to bit rate cost obtained in the tth search step, corresponding to the mt bit rate version of the video file ft cached on the edge serverst , Indicates the encoding bit rate of the mth bit rate version of the video file f, and T indicates the time length of the video file.
(c)更新步骤：(c) Update steps:
如果添加元素仍能满足各边缘服务器的缓存容量约束，即If adding element can still meet the cache capacity constraints of each edge server, that is,
则将该元素添加至局部最优解集合，并且搜索集合保持不变，即令以及vt＝vt-1；如果添加该元素不能满足各边缘服务器的缓存容量约束，则局部最优解集合保持不变，并且将该元素从搜索集合中移除，即令以及then the element Added to the local optimal solution set, and the search set remains unchanged, that is, And vt ＝vt-1 ; if adding this element cannot meet the cache capacity constraints of each edge server, the local optimal solution set remains unchanged, and this element is removed from the search set, that is, as well as
上式中，表示边缘服务器st的物理存储容量，指代在边缘服务器st上缓存的某一个视频版本，具体的表示视频文件f的第m个码率版本，表示所有可能在边缘服务器st处缓存的视频版本的全集。In the above formula, Indicates the physical storage capacity of the edge server st , Refers to a certain video version cached on the edge server st , and specifically represents the mth bit rate version of the video file f, represents the corpus of all possible video versions cached at the edge serverst .
(d)判定步骤：(d) Judgment steps:
如果剩余搜索集合不是空集，则令t＝t+1并返回迭代搜索步骤；否则，停止迭代并且将当前的局部最优解集合输出为最优结果。If the remaining search collection is not an empty set, set t=t+1 and return to the iterative search step; otherwise, stop the iteration and output the current local optimal solution set as the optimal result.
图5示出了一个具体的分布式缓存网络示例，包括三个边缘服务器S1、S2和S3，以及20个移动用户。图中的连线表示边缘服务器和用户之间存在无线链路，并且无线链路的带宽大小与该连线的长度成反比。Fig. 5 shows a specific example of a distributed cache network, including three edge servers S1 , S2 and S3 , and 20 mobile users. The connection in the figure indicates that there is a wireless link between the edge server and the user, and the bandwidth of the wireless link is inversely proportional to the length of the connection.
图6示出了主服务器处三个具体视频文件(Crowd Run、Tractor和Sunflower)的码率-失真性能曲线，其中三个视频的空间分辨率均为1080p(1920×1080)，编码帧率均为30帧每秒。Figure 6 shows the rate-distortion performance curves of three specific video files (Crowd Run, Tractor, and Sunflower) at the main server, where the spatial resolutions of the three videos are all 1080p (1920×1080), and the encoding frame rates are all 30 frames per second.
图7示出了在不同的边缘服务器数量设置下，本发明所述的分布式缓存内容放置方法的缓存性能要优于在背景技术中所提到的Femto缓存方法，其中最大视频失真为Dmax＝500。Figure 7 shows that under different edge server quantity settings, the cache performance of the distributed cache content placement method of the present invention is better than the Femto cache method mentioned in the background technology, wherein the maximum video distortion isDmax =500.
本发明为适应动态自适应视频流媒体多码率版本传输以及异构网络用户差异化带宽条件和点播需求的需要，建立基于边缘服务器分布式缓存的优化放置问题，并且相应地提供了一种高效快速的分布式缓存优化放置方法，实现了用户通过边缘服务器下载观看视频整体质量的最大化。本发明提高了边缘服务器缓存视频内容的利用率，减轻了主服务器处的视频流媒体服务负载，为用户提供更佳的视频服务质量。In order to meet the needs of dynamic self-adaptive video streaming media multi-bit rate version transmission and heterogeneous network users' differentiated bandwidth conditions and on-demand requirements, the present invention establishes an optimal placement problem based on edge server distributed cache, and accordingly provides an efficient The fast distributed cache optimization placement method maximizes the overall quality of videos downloaded and watched by users through edge servers. The invention improves the utilization rate of video content cached by the edge server, reduces the video stream media service load at the main server, and provides users with better video service quality.
以上对本发明的具体实施例进行了描述。需要理解的是，本发明并不局限于上述特定实施方式，本领域技术人员可以在权利要求的范围内做出各种变形或修改，这并不影响本发明的实质内容。Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610298248.XACN105979274B (en) | 2016-05-06 | 2016-05-06 | A Distributed Cache Placement Method for Dynamic Adaptive Video Streaming Media |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610298248.XACN105979274B (en) | 2016-05-06 | 2016-05-06 | A Distributed Cache Placement Method for Dynamic Adaptive Video Streaming Media |
| Publication Number | Publication Date |
|---|---|
| CN105979274Atrue CN105979274A (en) | 2016-09-28 |
| CN105979274B CN105979274B (en) | 2019-02-22 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610298248.XAActiveCN105979274B (en) | 2016-05-06 | 2016-05-06 | A Distributed Cache Placement Method for Dynamic Adaptive Video Streaming Media |
| Country | Link |
|---|---|
| CN (1) | CN105979274B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106713956A (en)* | 2016-11-16 | 2017-05-24 | 上海交通大学 | Rate control and version selection method and system for dynamic adaptive video streaming media |
| CN107241577A (en)* | 2017-07-03 | 2017-10-10 | 华中科技大学 | A kind of processing system for video based on collaborative group mechanism |
| CN107295070A (en)* | 2017-05-31 | 2017-10-24 | 上海交通大学 | The distributed coding caching laying method and system of file size isomery |
| CN108777809A (en)* | 2018-04-11 | 2018-11-09 | 中国科学院信息工程研究所 | A kind of panoramic video fragment mobile network caching method and system, panoramic video method for down loading |
| CN108810468A (en)* | 2018-06-11 | 2018-11-13 | 哈尔滨工程大学 | A kind of video frequency transmitter and method of optimization display effect |
| CN108833996A (en)* | 2018-07-03 | 2018-11-16 | 湖北大学 | Service Node Selection, Update and Code Rate Adaptation Method in Distributed DASH System |
| CN109451517A (en)* | 2018-12-27 | 2019-03-08 | 同济大学 | A kind of caching placement optimization method based on mobile edge cache network |
| CN110248210A (en)* | 2019-05-29 | 2019-09-17 | 上海交通大学 | Video frequency transmission optimizing method |
| CN111447506A (en)* | 2020-03-25 | 2020-07-24 | 武汉理工大学 | Streaming media content placement method based on delay and cost balance in cloud edge environment |
| CN111629217A (en)* | 2020-04-20 | 2020-09-04 | 南京邮电大学 | Optimization method of VOD service cache based on XGBoost algorithm in edge network environment |
| CN114630183A (en)* | 2022-03-17 | 2022-06-14 | 东南大学 | Edge device caching method and evaluation method based on scalable coding |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102186072A (en)* | 2011-04-20 | 2011-09-14 | 上海交通大学 | Optimized transmission method of multi-rate multicast communication for scalable video stream |
| CN103069492A (en)* | 2010-06-18 | 2013-04-24 | 三星电子株式会社 | Storage file format for multimedia streaming file, and storage method and client apparatus using the same |
| CN103428107A (en)* | 2012-05-14 | 2013-12-04 | 中国科学院声学研究所 | Self-adaptive bitstream switching method and system based on cache underflow probability estimation |
| CN103733589A (en)* | 2011-08-09 | 2014-04-16 | 阿尔卡特朗讯公司 | Method for streaming video content, edge node and client entity realizing such a method |
| US20140280679A1 (en)* | 2013-03-14 | 2014-09-18 | The Regents Of The University Of California | System and method for video caching in wireless networks |
| US20140308023A1 (en)* | 2011-06-09 | 2014-10-16 | Xiaoping Zhang | System and method for video summarization and navigation based on statistical models |
| US20140355624A1 (en)* | 2013-05-31 | 2014-12-04 | Broadcom Corporation | Transmitting multiple adaptive bit rate (abr) segment streams on a shared frequency |
| US20150012928A1 (en)* | 2013-07-08 | 2015-01-08 | The Trustees Of Princeton University | Quota aware video adaptation |
| CN105007541A (en)* | 2015-07-29 | 2015-10-28 | 上海交通大学 | Scalable video stream dynamic multi-rate multicast optimal transmission method |
| CN105068755A (en)* | 2015-07-08 | 2015-11-18 | 南京邮电大学 | Data duplicate storage method facing cloud computing content distribution network |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103069492A (en)* | 2010-06-18 | 2013-04-24 | 三星电子株式会社 | Storage file format for multimedia streaming file, and storage method and client apparatus using the same |
| CN102186072A (en)* | 2011-04-20 | 2011-09-14 | 上海交通大学 | Optimized transmission method of multi-rate multicast communication for scalable video stream |
| US20140308023A1 (en)* | 2011-06-09 | 2014-10-16 | Xiaoping Zhang | System and method for video summarization and navigation based on statistical models |
| CN103733589A (en)* | 2011-08-09 | 2014-04-16 | 阿尔卡特朗讯公司 | Method for streaming video content, edge node and client entity realizing such a method |
| CN103428107A (en)* | 2012-05-14 | 2013-12-04 | 中国科学院声学研究所 | Self-adaptive bitstream switching method and system based on cache underflow probability estimation |
| US20140280679A1 (en)* | 2013-03-14 | 2014-09-18 | The Regents Of The University Of California | System and method for video caching in wireless networks |
| US20140355624A1 (en)* | 2013-05-31 | 2014-12-04 | Broadcom Corporation | Transmitting multiple adaptive bit rate (abr) segment streams on a shared frequency |
| US20150012928A1 (en)* | 2013-07-08 | 2015-01-08 | The Trustees Of Princeton University | Quota aware video adaptation |
| CN105068755A (en)* | 2015-07-08 | 2015-11-18 | 南京邮电大学 | Data duplicate storage method facing cloud computing content distribution network |
| CN105007541A (en)* | 2015-07-29 | 2015-10-28 | 上海交通大学 | Scalable video stream dynamic multi-rate multicast optimal transmission method |
| Title |
|---|
| CHENGLIN LI，, HONGKAI XIONG: "Distributed Robust Optimization for Scalable Video Multirate Multicast Over Wireless Networks", 《IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY》* |
| YICHAO JIN ET AL: "Optimal Transcoding and Caching for Adaptive Streaming in Media Cloud:an Analytical Approach", 《IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY》* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10778982B2 (en) | 2016-11-16 | 2020-09-15 | Shanghai Jiao Tong University | Method and system for bit rate control and version selection for dynamic adaptive video streaming media |
| WO2018090774A1 (en)* | 2016-11-16 | 2018-05-24 | 上海交通大学 | Method and system for bit rate control and version selection for dynamic adaptive video streaming media |
| CN106713956A (en)* | 2016-11-16 | 2017-05-24 | 上海交通大学 | Rate control and version selection method and system for dynamic adaptive video streaming media |
| WO2018218850A1 (en)* | 2017-05-31 | 2018-12-06 | 上海交通大学 | Decentralized coded caching placement method and system for files with heterogeneous sizes |
| CN107295070A (en)* | 2017-05-31 | 2017-10-24 | 上海交通大学 | The distributed coding caching laying method and system of file size isomery |
| CN107295070B (en)* | 2017-05-31 | 2019-10-29 | 上海交通大学 | The distributed coding caching laying method and system of file size isomery |
| CN107241577B (en)* | 2017-07-03 | 2019-08-13 | 华中科技大学 | A kind of processing system for video based on collaborative group mechanism |
| CN107241577A (en)* | 2017-07-03 | 2017-10-10 | 华中科技大学 | A kind of processing system for video based on collaborative group mechanism |
| CN108777809B (en)* | 2018-04-11 | 2020-08-21 | 中国科学院信息工程研究所 | A panorama video fragmentation mobile network caching method and system, and panorama video downloading method |
| CN108777809A (en)* | 2018-04-11 | 2018-11-09 | 中国科学院信息工程研究所 | A kind of panoramic video fragment mobile network caching method and system, panoramic video method for down loading |
| CN108810468A (en)* | 2018-06-11 | 2018-11-13 | 哈尔滨工程大学 | A kind of video frequency transmitter and method of optimization display effect |
| CN108810468B (en)* | 2018-06-11 | 2020-06-16 | 哈尔滨工程大学 | Video transmission device and method for optimizing display effect |
| CN108833996A (en)* | 2018-07-03 | 2018-11-16 | 湖北大学 | Service Node Selection, Update and Code Rate Adaptation Method in Distributed DASH System |
| CN108833996B (en)* | 2018-07-03 | 2020-07-10 | 湖北大学 | Service node selection, update and code rate self-adaption method in distributed DASH system |
| CN109451517A (en)* | 2018-12-27 | 2019-03-08 | 同济大学 | A kind of caching placement optimization method based on mobile edge cache network |
| CN110248210A (en)* | 2019-05-29 | 2019-09-17 | 上海交通大学 | Video frequency transmission optimizing method |
| CN110248210B (en)* | 2019-05-29 | 2020-06-30 | 上海交通大学 | Video Transmission Optimization Methods |
| CN111447506A (en)* | 2020-03-25 | 2020-07-24 | 武汉理工大学 | Streaming media content placement method based on delay and cost balance in cloud edge environment |
| CN111447506B (en)* | 2020-03-25 | 2021-10-15 | 武汉理工大学 | A streaming media content placement method based on delay and cost balance in cloud-side environment |
| CN111629217A (en)* | 2020-04-20 | 2020-09-04 | 南京邮电大学 | Optimization method of VOD service cache based on XGBoost algorithm in edge network environment |
| CN111629217B (en)* | 2020-04-20 | 2021-11-26 | 南京邮电大学 | XGboost algorithm-based VOD (video on demand) service cache optimization method in edge network environment |
| CN114630183A (en)* | 2022-03-17 | 2022-06-14 | 东南大学 | Edge device caching method and evaluation method based on scalable coding |
| CN114630183B (en)* | 2022-03-17 | 2024-03-26 | 东南大学 | An edge device caching method and evaluation method based on scalability coding |
| Publication number | Publication date |
|---|---|
| CN105979274B (en) | 2019-02-22 |
| Publication | Publication Date | Title |
|---|---|---|
| CN105979274B (en) | A Distributed Cache Placement Method for Dynamic Adaptive Video Streaming Media | |
| US10778982B2 (en) | Method and system for bit rate control and version selection for dynamic adaptive video streaming media | |
| CN107295070B (en) | The distributed coding caching laying method and system of file size isomery | |
| US7636363B2 (en) | Adaptive QoS system and method | |
| CN105512053B (en) | The mirror cache method of mobile transparent computing system server end multi-user access | |
| US20140365613A1 (en) | Defragmentation of adaptive streaming segment files in a content delivery network | |
| CN112543357B (en) | Stream media data transmission method based on DASH protocol | |
| CN113282786B (en) | An edge-cooperative cache replacement method for panoramic video based on deep reinforcement learning | |
| RU2571732C2 (en) | Control device and method of controlling network streaming of video data to network user device | |
| CN109451517B (en) | Cache placement optimization method based on mobile edge cache network | |
| CN110913239B (en) | Video cache updating method for refined mobile edge calculation | |
| Shi et al. | LEAP: learning-based smart edge with caching and prefetching for adaptive video streaming | |
| CN108769729B (en) | Cache arrangement system and cache method based on genetic algorithm | |
| CN111314349B (en) | Code caching method based on joint maximum distance code division and cluster cooperation in fog wireless access network | |
| CN109040771A (en) | Based on the video cache method and system to cooperate between more cache servers | |
| KR20130057232A (en) | Contents delivery network system having edge node and method for caching and managing contents | |
| CN111447506A (en) | Streaming media content placement method based on delay and cost balance in cloud edge environment | |
| US9992300B2 (en) | Method of adaptively deploying cache positioned at subscriber network, and system therefor | |
| KR101973724B1 (en) | Method and apparatus for providing adaptively video service | |
| CN116320008A (en) | Transfer method for device-to-device encoding caching problem based on random requester transfer | |
| CN108429919B (en) | Cache and transmission optimization method of multi-rate video in wireless network | |
| CN116319999B (en) | Content heat decay method and device in CDN | |
| CN116112708B (en) | Self-adaptive streaming media-oriented combined content storage, code rate conversion and power allocation resource optimization method | |
| CN110597464A (en) | Method and device for hierarchical management of streaming media | |
| CN113992653B (en) | CDN-P2P network content downloading, pre-storing and replacing method based on edge cache |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |