CN102821316A - Improved video on demand (VOD) transmission method based on peer-to-peer computing core algorithm - Google Patents
Improved video on demand (VOD) transmission method based on peer-to-peer computing core algorithm Download PDFInfo
- Publication number
- CN102821316A CN102821316A CN2012101479902A CN201210147990A CN102821316A CN 102821316 A CN102821316 A CN 102821316A CN 2012101479902 A CN2012101479902 A CN 2012101479902A CN 201210147990 A CN201210147990 A CN 201210147990A CN 102821316 A CN102821316 A CN 102821316A
- Authority
- CN
- China
- Prior art keywords
- node
- list
- server
- improved
- seed
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 230000005540 biological transmission Effects 0.000 title claims abstract description 20
- 230000007246 mechanism Effects 0.000 claims abstract description 11
- 238000009826 distribution Methods 0.000 claims abstract description 8
- 238000004519 manufacturing process Methods 0.000 claims abstract description 5
- 239000012634 fragment Substances 0.000 claims description 73
- 230000008569 process Effects 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 5
- 238000013467 fragmentation Methods 0.000 claims description 5
- 238000006062 fragmentation reaction Methods 0.000 claims description 5
- 230000006872 improvement Effects 0.000 claims description 5
- 238000009827 uniform distribution Methods 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 230000001174 ascending effect Effects 0.000 claims description 2
- 230000003044 adaptive effect Effects 0.000 abstract description 10
- 238000005516 engineering process Methods 0.000 description 17
- 238000004364 calculation method Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 101150012579 ADSL gene Proteins 0.000 description 1
- 102100020775 Adenylosuccinate lyase Human genes 0.000 description 1
- 108700040193 Adenylosuccinate lyases Proteins 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Images
Landscapes
- Information Transfer Between Computers (AREA)
Abstract
一种基于对等计算核心算法改进的视频点播传输方法拓扑结构包括种子文件制作服务器、电子节目单服务器(EPG)、改进后的索引服务器(Tracker)、流媒体服务器和改进后的BitTorrent客户端。本发明改进了BitTorrent协议中的片段选择算法和节点选择算法,引进了自适应滑动窗口机制,充分利用了BitTorrent协议高效的文件分发能力、同时又能够满足视频点播对实时性的要求。该方法既能够解决服务器单点失效和连接服务器瓶颈问题,同时又能减轻服务器端负载压力。
An improved video-on-demand transmission method based on the core algorithm of peer-to-peer computing. The topology structure includes a seed file production server, an electronic program guide server (EPG), an improved index server (Tracker), a streaming media server and an improved BitTorrent client. The invention improves the segment selection algorithm and node selection algorithm in the BitTorrent protocol, introduces an adaptive sliding window mechanism, fully utilizes the efficient file distribution capability of the BitTorrent protocol, and can meet the real-time requirements of video-on-demand. This method can not only solve the problem of server single point failure and connection server bottleneck, but also reduce the load pressure on the server side.
Description
技术领域 technical field
本发明是一种基于对等计算核心算法改进的视频点播系统传输方法,使用的对等计算协议为BitTorrent协议,它对BitTorrent中的片段选择算法和节点选择算法进行改进,引进自适应滑动窗口机制对分片选择请求进行限制,以满足视频点播对实时性的要求,属于对等网络应用领域。 The present invention is an improved video-on-demand system transmission method based on the peer-to-peer computing core algorithm. The peer-to-peer computing protocol used is the BitTorrent protocol. It improves the segment selection algorithm and node selection algorithm in BitTorrent, and introduces an adaptive sliding window mechanism. Restricting slice selection requests to meet the real-time requirements of video on demand belongs to the field of peer-to-peer network applications.
背景技术 Background technique
视频点播技术是分布式流媒体的一种重要的应用技术。视频点播也被称为VOD,全称是Video on Demand,即按需要的视频流播放。视频点播技术是当前流媒体通信应用的一项前沿技术,是面向未来的信息、通信等相关技术相结合的新技术。同时,视频点播又是一项业务,它泛指利用交互式网络将声音图像进行实时传输,以实现影视点播、信息查询、电视购物、远程教育等服务的一项交互式多媒体应用业务。 Video on demand technology is an important application technology of distributed streaming media. Video on Demand is also known as VOD, the full name is Video on Demand, that is, video streaming on demand. Video-on-demand technology is a cutting-edge technology in the current streaming media communication application, and it is a new technology that combines future-oriented information, communication and other related technologies. At the same time, video on demand is another business, which generally refers to an interactive multimedia application service that uses interactive networks to transmit audio and images in real time to realize video on demand, information query, TV shopping, distance education and other services.
因特网上的传统流媒体系统一般是基于C/S模式的,通常包括一台或多台服务器和若干客户端。在C/S模式下,流媒体系统容量(同时服务的客户端数量)主要由服务器端的网络输出带宽决定,有时服务器的处理能力、内存大小、I/O速率也影响到系统的容量。在C/S模式下,由于传输流媒体占用的带宽大,持续时间长,而服务器端可利用的网络带宽有限,所以即使是使用高档服务器,其系统容量也不过几百个客户,不具有经济规模性。另外,由于互联网的不确定性,如果客户端距离服务器较远,则流媒体传输过程中的延迟、抖动、带宽、丢包率等指标也将更加不确定。服务器为每个客户单独发送一次流媒体内容,故网络资源的消耗也十分巨大。所以在当前资源有限的条件下,如何满足不断增长的用户需求,并在确保服务质量的同时,降低流媒体的服务成本,称为流媒体技术研究中的重要课题。近几年兴起的P2P技术,能利用互联网中的各个节点进行对等计算,充分利用互联网上的空闲资源,允许两个客户端直接交换信息,因此受到了广泛关注。将P2P技术应用于流媒体业务,为解决服务器端网络带宽和服务器能力限制问题提供了一种新的思路。在此背景下,基于P2P的视频点播技术产生了。 Traditional streaming media systems on the Internet are generally based on the C/S model, usually including one or more servers and several clients. In the C/S mode, the capacity of the streaming media system (the number of clients served at the same time) is mainly determined by the network output bandwidth of the server, and sometimes the processing power, memory size, and I/O rate of the server also affect the system capacity. In the C/S mode, due to the large bandwidth and long duration of streaming media transmission, and the limited network bandwidth available on the server side, even if a high-end server is used, its system capacity is only a few hundred clients, which is not economical. Scale. In addition, due to the uncertainty of the Internet, if the client is far away from the server, indicators such as delay, jitter, bandwidth, and packet loss rate during streaming media transmission will also be more uncertain. The server sends streaming media content once for each client separately, so the consumption of network resources is also very huge. Therefore, under the current condition of limited resources, how to meet the ever-increasing user needs and reduce the service cost of streaming media while ensuring service quality is an important topic in the research of streaming media technology. The P2P technology that has emerged in recent years can use each node in the Internet to perform peer-to-peer computing, make full use of idle resources on the Internet, and allow two clients to directly exchange information, so it has received widespread attention. Applying P2P technology to streaming media business provides a new way of thinking to solve the problem of server-side network bandwidth and server capacity limitations. In this context, P2P-based video-on-demand technology emerged.
BitTorrent是一种高效的P2P文件共享下载工具,通过一个中心的Tracker服务器,BitTorrent客户端获取当前下载任务的节点列表,然后各节点之间完全对等的协同完成各文件分片的下载。文件分片共享提高了下载效率,同时无需任何其他服务器的参与。 BitTorrent is an efficient P2P file sharing download tool. Through a central Tracker server, the BitTorrent client obtains the node list of the current download task, and then complete peer-to-peer collaboration between each node completes the download of each file segment. File slice sharing improves download efficiency without any other server involvement.
BitTorrent协议中采用的片段选择算法为:局部最少分片优先,即每个对等节点都优先选择邻居节点中最少的那些片断去下载,而那些在系统中相对较多的片断,放在后面下载。 The fragment selection algorithm used in the BitTorrent protocol is: the local minimum fragment first, that is, each peer node first selects the fragments with the fewest neighbor nodes to download, and those relatively more fragments in the system are downloaded later .
BitTorrent协议中采用的节点选择算法为:在保持种子所占比例接近全局的种子的比例的情况下,随机选取节点信息返回给请求节点。 The node selection algorithm adopted in the BitTorrent protocol is: in the case of keeping the proportion of seeds close to the proportion of global seeds, randomly select node information and return it to the requesting node.
发明内容 Contents of the invention
技术问题:本发明的目的是通过对BitTorrent协议中的片段选择算法和节点选择算法进行改进,同时引进一种自适应滑动窗口机制满足视频点播对实时性的要求。较之传统视频点播系统,它不仅能够高效的进行数据传送,确保视频点播的流畅性,并能够降低系统负载,降低运营成本。 Technical problem: the purpose of the present invention is to improve the segment selection algorithm and node selection algorithm in the BitTorrent protocol, and introduce an adaptive sliding window mechanism to meet the real-time requirements of video on demand. Compared with the traditional VOD system, it can not only transmit data efficiently, ensure the smoothness of VOD, but also reduce system load and operation cost.
技术方案:本发明的一种基于对等计算核心算法改进的视频点播传输方法采用改进的BitTorrent客户端和改进的索引服务器Tracker提升流媒体资源文件传输速度,算法涉及种子文件制作服务器、电子节目单服务器EPG、改进的索引服务器Tracker、流媒体服务器和改进的BitTorrent客户端模块;整个视频点播的传输过程如下: Technical scheme: a kind of improved video-on-demand transmission method based on peer-to-peer computing core algorithm of the present invention adopts improved BitTorrent client and improved index server Tracker to improve the transmission speed of streaming media resource files, and the algorithm involves seed file production server, electronic program list Server EPG, improved index server Tracker, streaming server and improved BitTorrent client module; the transmission process of the entire video on demand is as follows:
步骤.11:种子文件制作服务器将要发布的流媒体文件制作成种子文件,然 Step.11: The seed file production server makes the streaming media file to be released into a seed file, and then
后将种子文件发送到电子节目单服务器EPG; Afterwards, the seed file is sent to the electronic program list server EPG;
步骤12:电子节目单服务器EPG下载种子文件,同时根据种子信息制作电 Step 12: The electronic program guide server EPG downloads the seed file, and at the same time makes a TV program according to the seed information.
子节目单; sub program list;
步骤13:视频点播客户端在启动后会自动向电子节目单服务器EPG请求电 Step 13: The video-on-demand client will automatically request the electronic program list server EPG after starting.
子节目单,电子节目单服务器EPG根据该用户的信息验证是否属于注册用户,如果用户通过验证电子节目单服务器EPG就会向该用户发送电子节目单,然后用户的客户端上就会出现点播节目菜单; Sub program list, the electronic program list server EPG verifies whether the user belongs to the registered user according to the user's information, if the user passes the verification, the electronic program list server EPG will send the electronic program list to the user, and then the on-demand program will appear on the user's client menu;
步骤14:根据点播菜单用户点播想看的节目,之后向电子节目单服务器EPG Step 14: According to the on-demand menu, the user orders the program he wants to watch, and then sends it to the electronic program list server EPG
请求种子文件,接着电子节目单服务器EPG向用户发送种子文件; Request the seed file, and then the electronic program list server EPG sends the seed file to the user;
步骤15:当用户收到种子文件后马上进入下载过程,首先通过改进的 Step 15: When the user receives the torrent file, enter the download process immediately, first through the improved
BitTorrent客户端与改进的索引服务器Tracker建立连接,得到邻居节点信息; The BitTorrent client establishes a connection with the improved index server Tracker to obtain neighbor node information;
步骤16:改进的BitTorrent客户端通过这些邻居节点的信息,采用滑动窗口 Step 16: The improved BitTorrent client passes the information of these neighbor nodes, using a sliding window
机制部分有序的下载流媒体文件,边下载边播放,同时也为其他邻居节点提供上载服务;客户端下载完成后变成种子节点,种子节点通过改进的片段选择算法为其它节点提供上传。 The mechanism part downloads streaming media files in an orderly manner, plays them while downloading, and also provides uploading services for other neighbor nodes; after the client downloads, it becomes a seed node, and the seed node provides uploading for other nodes through an improved segment selection algorithm.
改进的BitTorrent客户端,采用改进的片段选择算法,能够解决文件分片的均匀分发,改进步骤如下: The improved BitTorrent client uses an improved fragment selection algorithm, which can solve the even distribution of file fragments. The improvement steps are as follows:
步骤21:为了分片在网络中的均匀分布,废除种子的上传策略,不再根据 Step 21: For the uniform distribution of fragments in the network, abolish the seed upload strategy, no longer according to
下载速度来提供上传; Download speed to provide upload;
步骤22:在下载完成节点Seed中设置一个列表,初始化时所有片段号都放 Step 22: Set a list in the download completion node Seed, and put all the fragment numbers at initialization
在列表中,当片段被客户端节点上传后,则把该片段号从列表删除; In the list, when the fragment is uploaded by the client node, delete the fragment number from the list;
步骤23:当下载节点使用最少的分片优先下载的方式向Seed请求分片时, Step 23: When the download node uses the method of downloading the least fragments first to request fragments from the Seed,
如果下载分片在列表中,Seed发送分片文件给下载节点,并将分片号从列表中删除;如果不在,Seed通过下载节点的位域发送给下载节点没有且片段号在列表中的分片,然后将该分片从列表中删除,如果以上两种情况都不符合,则阻塞该节点; If the download fragment is in the list, Seed sends the fragment file to the download node, and deletes the fragment number from the list; slice, and then delete the slice from the list, if the above two conditions do not meet, then block the node;
步骤24:当列表中的片段号都被删除后,即列表为空时,重新进行初始化, Step 24: When the fragment numbers in the list are all deleted, that is, when the list is empty, re-initialize,
把所有片段号重新加入列表供节点下载。 Add all fragment numbers back to the list for node download.
改进的索引服务器Tracker,采用改进的节点选择算法,能够充分利用节点的上传效率,改进步骤如下: The improved index server Tracker adopts an improved node selection algorithm, which can make full use of the upload efficiency of nodes. The improvement steps are as follows:
步骤31.扩展节点与追踪服务器之间的通信,让Tracker能够知道节点的下 Step 31. Expand the communication between the node and the tracking server, so that the Tracker can know the node's next
载进度; loading progress;
步骤32.在Tracker服务器方对每个种子文件建立一个根据下载进度排列 Step 32. On the Tracker server side, establish a sequence for each torrent file according to the download progress
的升序列表; ascending list of
步骤33.当有种子文件发布时,将节点Seed首先加入到列表中,并将它的 Step 33. When a seed file is published, the node Seed is first added to the list, and its
下载进度置为1; The download progress is set to 1;
步骤34.当有对等节点与Tracker通信时,如果该节点为新加入节点则根据下载进度将它插入到列表相应的位置,如果对等节点已经在列表中,则删除以前列表所在位置,并将根据节点进度重新插入到列表中去; Step 34. When a peer node communicates with the Tracker, if the node is a newly added node, it will be inserted into the corresponding position of the list according to the download progress. If the peer node is already in the list, the previous position of the list will be deleted, and it will be The progress is reinserted into the list;
步骤35.根据节点位置查找40个跟节点进度差距最小的节点; Step 35. Find 40 nodes with the smallest progress gap with the node according to the node position;
步骤36.返回邻居节点列表给请求节点。 Step 36. Return the list of neighbor nodes to the requesting node.
步骤16中采用滑动窗口机制部分有序的下载流媒体文件,使得播放点附近的分片能够优先得到下载;在运行过程中,每次有新的文件分片要请求时,先检查紧急窗口中的文件分片是否被全部请求或下载完,若没有,则先请求滑动窗口中的分片,若请求完则从非滑动窗口中取文件分片进行请求,随着播放点的移动,滑动窗口也做出相应的滑动;并且这个滑动窗口的大小在播放过程中能够根据解码速度和播放延迟自适应的做出调整,既满足了视频点播对实时性的要求,又充分利用了BitTorrent高效的传输效率。 In step 16, the sliding window mechanism is used to partially and orderly download streaming media files, so that the fragments near the playback point can be downloaded first; during the operation, each time a new file fragment needs to be requested, first check the Whether the file fragments are all requested or downloaded, if not, first request the fragments in the sliding window, if the request is completed, take the file fragments from the non-sliding window to request, with the movement of the playback point, the sliding window Also make a corresponding slide; and the size of the sliding window can be adaptively adjusted according to the decoding speed and playback delay during playback, which not only meets the real-time requirements of video on demand, but also makes full use of BitTorrent's efficient transmission efficiency.
有益效果:本发明方法提出了基于BitTorrent核心算法改进的视频点播传输方法,旨在结合BitTorrent高效的文件分发机制,来解决视频点播系统中流媒体服务器负载压力过重,效率不高的问题。该发明提出的技术方法并不是现有技术的简单罗列,而是根据现有技术的特点,通过不断实践提炼出的,通过将各种机制有效的结合,来实现高效的P2P视频点播传输系统。下面给出具体的说明。 Beneficial effects: the method of the present invention proposes an improved video-on-demand transmission method based on the BitTorrent core algorithm, aiming to solve the problems of heavy load pressure and low efficiency of the streaming media server in the video-on-demand system by combining the efficient file distribution mechanism of BitTorrent. The technical method proposed in this invention is not a simple list of existing technologies, but is refined through continuous practice according to the characteristics of existing technologies, and realizes an efficient P2P video-on-demand transmission system by effectively combining various mechanisms. A specific description is given below.
数据孤岛:如果网络中只有服务器A拥有资源S,那么当服务器A由于某些原因宕机或是损坏时,网络中的所有节点都无法访问或是获取资源S,直到服务器A重新正常工作,服务器A的这种状态就称为“数据孤岛”。数据孤岛严重的影响了系统的稳定性和灵活性,网络中的少数节点就能影响到全局的性能。数据孤岛在传统的音乐点播系统中极为常见,点播用户的所有资源均来自媒体服务器,一旦媒体服务器宕掉,所有点播用户就无法获取媒体资源,一种解决方法是设置多个备用的媒体服务器,但这会带来极大的成本开销。而在我们的基于BitTorrent的视频点播传输系统中,点播用户的资源并不一定来自于媒体服务,而且在很多时候可以完全脱离媒体服务器,这样就摆脱了数据孤岛的问题了。 Data island : If only server A owns resource S in the network, then when server A goes down or is damaged for some reason, all nodes in the network cannot access or obtain resource S until server A works normally again. This state of A is called "data island". Data islands seriously affect the stability and flexibility of the system, and a few nodes in the network can affect the overall performance. Data islands are very common in traditional music on-demand systems. All resources of on-demand users come from the media server. Once the media server goes down, all on-demand users cannot obtain media resources. One solution is to set up multiple backup media servers. But this will bring great cost overhead. In our BitTorrent-based VOD transmission system, the resources of VOD users do not necessarily come from media services, and in many cases can be completely separated from the media server, thus getting rid of the problem of data islands.
负载均衡:是指系统的各个处理单元均衡的分担服务请求。即将网络中的负载分散到多个服务器或是节点上,从而有效的避免了网络中部分服务器或是节点负载过重,使得所有节点平均分担网络中的负载,保证网络更有效的运行。传统的多媒体点播系统中的负载均衡技术,通常只是对系统中的所有媒体服务器进行负载均衡,让所有的媒体服务器有接近的负载量,避免少数媒体服务器不堪重负,而有些却没有被有效的利用,如共享式多媒体点播系统中的负载均衡技术。在P2P音乐点播系统中,负载均衡技术是利用客户节点分担媒体服务器的负载,来减轻媒体服务器的负载,这种负载均衡技术相对于传统多媒体点播系统中的负载均衡技术来说,其作用更广泛,也更加彻底和有效。 Load balancing : refers to the balanced sharing of service requests among each processing unit of the system. That is, the load in the network is distributed to multiple servers or nodes, thereby effectively avoiding the overload of some servers or nodes in the network, so that all nodes can share the load in the network evenly, and ensure more efficient operation of the network. The load balancing technology in the traditional multimedia on-demand system usually only performs load balancing on all media servers in the system, so that all media servers have a similar load, so as to prevent a few media servers from being overwhelmed, while some are not effectively utilized , such as the load balancing technology in the shared multimedia on-demand system. In the P2P music on-demand system, the load balancing technology uses client nodes to share the load of the media server to reduce the load of the media server. Compared with the load balancing technology in the traditional multimedia on-demand system, this load balancing technology has a wider role , is also more thorough and effective.
改进的BitTorrent核心算法:BitTorrent协议中的片段选择算法采用的是分片副本数量局部最少优先策略,这种策略不能使分片数量均匀度接近于最优,本发明采用改进的种子节点上传策略对这种策略进行改进,实验证明分片数量均匀度、下载速度、平均下载时间比传统的策略和改进前的策略有了较大的改进。Tracker节点选择算法近似于一种随机选择策略,算法存在不确定性、随机性比较强,可能需要很长的搜寻时间才能找到合适的服务提供者。针对这个问题,本发明采用改进的节点选择算法,帮助节点在最短的时间内找到合适的节点。经过实验证明改进后算法下载速度有了明显提高,并且在最后阶段模式下载速度也没有明显的降低。 Improved BitTorrent core algorithm: What the segment selection algorithm in the BitTorrent agreement adopted is the local minimum priority strategy of the number of fragment copies, this strategy can not make the uniformity of the number of fragments close to optimal, the present invention adopts the improved seed node upload strategy to This strategy is improved, and the experiment proves that the uniformity of the number of fragments, the download speed, and the average download time have been greatly improved compared with the traditional strategy and the strategy before the improvement. The Tracker node selection algorithm is similar to a random selection strategy. The algorithm has uncertainty and strong randomness, and it may take a long search time to find a suitable service provider. To solve this problem, the present invention adopts an improved node selection algorithm to help nodes find a suitable node in the shortest time. Experiments have proved that the download speed of the improved algorithm has been significantly improved, and the download speed of the final stage mode has not been significantly reduced.
附图说明 Description of drawings
图1是改进后种子节点上传流程图,, Figure 1 is the flow chart of the improved seed node upload,
图2是改进后的节点选择算法流程图, Figure 2 is a flow chart of the improved node selection algorithm,
图3 是基于BitTorrent的视频点播传输的流程图, Fig. 3 is a flow chart of BitTorrent-based VOD transmission,
图4 是系统在运行过程中的具体流程图。 Figure 4 is the specific flow chart of the system during operation.
具体实施方式 Detailed ways
本发明的方法强调对BitTorrent中的核心算法进行改进,并引进自适应滑动窗口机制以满足视频点播对实时性的要求。 The method of the invention emphasizes improving the core algorithm in BitTorrent, and introduces an adaptive sliding window mechanism to meet the real-time requirements of video on demand.
.改进的片段选择算法. Improved Fragment Selection Algorithm
在BitTorrent系统中采用的是局部文件片优先策略(Local Rarest First),也就是节点只需知道共享文件的文件片在其邻居节点(一般是50个节点)中的分布,在其和其所有邻居节点中执行最少文件块优先策略。这样可以降低系统负载,并在局部保持了文件块的均匀分布。但BitTorrent在片段选择策略方面存在一些问题:由于局部最少优先仅仅基于局部信息,局部最少不代表全局最少,因此文件块的选择在一定程度上存在盲目性。 In the BitTorrent system, the Local Rarest First strategy is adopted, that is, the node only needs to know the distribution of the file slices of the shared file among its neighbor nodes (usually 50 nodes). The least file block priority strategy is implemented in the node. This reduces system load and maintains an even distribution of file blocks locally. But there are some problems in BitTorrent's fragment selection strategy: because the local least priority is only based on local information, the local least does not mean the global least, so the selection of file blocks is blind to a certain extent.
改进后的片段选择算法为:通过改进种子上传策略,由上传策略间接的影响文件分片分布的均衡度。 The improved fragment selection algorithm is as follows: by improving the seed upload strategy, the upload strategy indirectly affects the balance of file fragment distribution.
传统的种子上传策略为:当客户端节点的所有分片都下载完成后,如果它没有马上离开网络它将转变为种子节点,由于种子节点不会再从其他节点下载文件分片,它将不再使用节点选择策略选择节点进行上传,而是选择从本节点得到最快下载速度的几个节点提供上传,这样可以尽可能的利用上传带宽。 The traditional seed upload strategy is: when all the fragments of the client node are downloaded, if it does not leave the network immediately, it will become a seed node. Since the seed node will not download file fragments from other nodes, it will not Then use the node selection strategy to select nodes for uploading, but choose several nodes that get the fastest download speed from this node to provide uploading, so that the uploading bandwidth can be utilized as much as possible.
本发明提出一种改进后的片段选择算法,算法思想如图1所示,其流程为: The present invention proposes an improved fragment selection algorithm, the algorithm idea is as shown in Figure 1, and its process is:
1. 为了分片在网络中的均匀分布,废除种子的上传策略,不再根据下载速 1. For the uniform distribution of fragments in the network, the upload strategy of seeds is abolished, and no longer based on the download speed
度来提供上传。 degrees to provide uploads.
2. 下载节点的片段选择算法不做改变,即还保留以前的随机的第一个分片、 2. The fragment selection algorithm of the download node remains unchanged, that is, the previous random first fragment,
最少的优先、严格的优先级、最后阶段模式。 Least-first, strict-priority, last-stage modes.
3. 在Seed里设置一个列表,初始化时所有片段号都放在列表中,当片段被 3. Set a list in Seed, all fragment numbers are placed in the list at initialization, when the fragment is
Seed节点上传后,则把该片段号从列表删除。 After the Seed node is uploaded, the segment number is deleted from the list.
4. 当下载节点使用最少优先向Seed请求分片时,如果下载分片在列表中, 4. When the download node uses the least priority to request fragments from Seed, if the download fragment is in the list,
Seed发送分片文件给下载节点,并将分片号从列表中删除。如果不在,Seed可以通过下载节点的bitfield域发送给下载节点下载节点没有且片段号在列表中的分片,然后将该分片从列表中删除,如果以上两种情况都不符合,则阻塞该节点。 Seed sends the fragment file to the download node and deletes the fragment number from the list. If not, the Seed can be sent to the download node through the bitfield field of the download node to the fragment that the download node does not have and the fragment number is in the list, and then delete the fragment from the list. If the above two conditions do not meet, block the fragment node.
5. 当列表中的片段号都被删除后,即列表为空时,重新进行初始化,把所 5. When the fragment numbers in the list are all deleted, that is, when the list is empty, re-initialize and set all
有片段号重新加入列表供节点下载。 There are fragment numbers to re-join the list for nodes to download.
.改进的节点选择算法. Improved Node Selection Algorithm
为了充分利用邻居节点的上传带宽,节点的邻居节点最好包含节点所需的文件分片。而传统的Tracker算法采用一种随机选取邻居节点的办法,这在一定程度上影响了节点的下载效率,不利于文件分片高效的分发。这种情况在节点下载的最后阶段尤为明显,由于在下载的最后节点所需的分片数量比较少,如果选取拥有分片比较少的节点作为邻居节点,往往得到分片的概率会很低,也就造成了在最后阶段节点的下载速度会有明显的降低。 In order to make full use of the upload bandwidth of neighbor nodes, the neighbor nodes of the node should preferably contain the file fragments required by the node. The traditional Tracker algorithm uses a method of randomly selecting neighbor nodes, which affects the download efficiency of nodes to a certain extent, and is not conducive to the efficient distribution of file fragments. This situation is especially obvious in the final stage of node downloading. Since the number of fragments required for the final node to be downloaded is relatively small, if a node with fewer fragments is selected as a neighbor node, the probability of obtaining fragments is often very low. It also causes the download speed of the node in the final stage to be significantly reduced.
本发明提出一种改进的节点选择算法,算法思想如图2所示,其流程为: The present invention proposes an improved node selection algorithm, the algorithm idea is shown in Figure 2, and its flow is:
1.扩展对等节点与Tracker之间的通信,让Tracker能够知道节点的下载进度。 1. Extend the communication between peer nodes and Tracker, so that Tracker can know the download progress of the node.
2.在Tracker服务器方对每个torrent文件建立一个根据下载进度排列的升 2. On the Tracker server side, create an upgrade list for each torrent file according to the download progress.
序列表。 sequence listing.
3.当有种子文件发布时,将Seed节点首先加入到列表中,并将它的下载进 3. When a seed file is released, the Seed node is first added to the list, and its download is entered into
度置为1。 degree is set to 1.
4.当对等节点与Tracker通信时,如果对等节点为新加入节点则根据下载进度将它插入到列表相应的位置,如果对等节点已经在列表中,则删除以前列表所在位置,并将根据节点进度重新插入到列表中去。 4. When the peer node communicates with the Tracker, if the peer node is a newly added node, it will be inserted into the corresponding position of the list according to the download progress. If the peer node is already in the list, the previous position of the list will be deleted, and will be The progress is reinserted into the list.
5.根据节点位置查找40个跟节点进度差距最小的节点。 5. Find 40 nodes with the smallest progress gap with the node according to the node position.
6.返回邻居节点列表给请求节点。 6. Return the list of neighbor nodes to the requesting node.
.自适应滑动窗口机制. Adaptive sliding window mechanism
滑动窗口维护一个关于媒体数据接收缓冲的窗口,这个窗口将根据播放器消耗数据和对等节点接收数据的状态变化向前移动,一旦位于滑动窗口内的数据被播放器消耗掉,则滑动窗口将向前移动,以获取播放器后续播放需要的媒体数据。位于滑动窗口内的数据将根据紧急程度按照优先级顺序地从服务器或其它伙伴节点请求。 The sliding window maintains a window about the media data receiving buffer. This window will move forward according to the state changes of the data consumed by the player and the data received by the peer node. Once the data within the sliding window is consumed by the player, the sliding window will Move forward to get the media data needed by the player for subsequent playback. Data within the sliding window will be requested from the server or other partner nodes in priority order according to urgency.
从前面的讨论我们可以知道,滑动窗口越大,则获取紧急数据的速度越快,同时由于滑动窗口内每个分片都向多个对等节点请求(以达到和保证迅速获取分片),无疑将增加系统中分片请求和响应的冗余,而滑动窗口越小,相对地,位于滑动窗口外的数据请求比率将会增大。对于那些位于滑动窗口外的数据,除了后面我们所采取的设置“锚点”的部分,我们仍然采用的是BitTorrent的“最少者优先”策略,系统的冗余将相对变小,能够有效地降低系统负载,但是这样可能不能够及时地满足播放器对即将播放数据的需求。这主要取决于播放器缓冲数据接收和消耗的速度的变化,当接收速度大于消耗速度的时候,显然缓冲能够满足播放器消耗的速度,这时候应该将滑动窗口减小,以降低系统负载,而滑动窗口将用于保证“最少者优先”策略没有下载下来的播放器所需的紧急数据的请求(我们知道“最少者优先”为了平衡系统负载,总是最先下载群集中复制较少的分片,这种无序下载的方式不能够满足播放器顺序播放的需要),当接收速度小于消耗速度的时候,因为播放器缓冲数据消耗较快,为了减少用户播放等待的时间,我们应该增大滑动窗口,以尽快获取播放器所需要的数据,因此,为了动态地反映滑动窗口内数据的变化,以达到更好的平衡系统负载,我们提出一种自适应大小的滑动窗口方案。 From the previous discussion, we can know that the larger the sliding window, the faster the urgent data can be obtained. At the same time, since each fragment in the sliding window requests multiple peer nodes (to achieve and ensure the rapid acquisition of fragments), Undoubtedly, it will increase the redundancy of fragmentation requests and responses in the system, and the smaller the sliding window is, correspondingly, the ratio of data requests outside the sliding window will increase. For the data outside the sliding window, we still adopt BitTorrent's "least first" strategy except for the part of setting "anchor point" later, so the redundancy of the system will be relatively smaller, which can effectively reduce the System load, but this may not be able to meet the player's demand for the data to be played in time. This mainly depends on the changes in the receiving and consuming speed of the player's buffered data. When the receiving speed is greater than the consumption speed, obviously the buffer can meet the consumption speed of the player. At this time, the sliding window should be reduced to reduce the system load. The sliding window will be used to ensure that the "least first" strategy does not have urgent data requests for downloaded players (we know that "least first" always downloads the less replicated parts in the cluster first in order to balance the system load) slices, this method of out-of-order download cannot meet the needs of the player to play sequentially), when the receiving speed is lower than the consumption speed, because the buffer data consumption of the player is faster, in order to reduce the waiting time for the user to play, we should increase the The sliding window is used to obtain the data required by the player as soon as possible. Therefore, in order to dynamically reflect the data changes in the sliding window and achieve a better balance of system load, we propose an adaptive sliding window scheme.
从上述论述中我们可以清楚地认识到,滑动窗口的大小是一个影响系统性能的至关重要的参数。我们认为滑动窗口的大小应该与播放启动等待时间 相对应,客户端应该在这个播放等待时间内填充满滑动窗口缓冲,而后开始播放媒体流,在播放的同时,不断地继续填充这个缓冲,以保证播放的连续性。这样关于滑动窗口的最佳大小,根据试验统计,滑动窗口的大小应该静态的设置为能够缓冲300s的媒体数据。根据(其中表示播放器的播放延迟,即从客户端请求节目到播放器开始播放之间的时间间隔;表示分片的大小,用于保存在对等节点的媒体数据的最小单位,大小为256K字节;表示媒体文件的解码速度)的滑动窗口大小计算公式,前提都在于假设分片获取的速度总是至少大于或等于播放器流媒体的解码速度,这在实际应用中,特别对于广大的ADSL用户和具有高质量QoS保证的流媒体数据,具有极大的不现实性。我们滑动窗口的初始大小同样采取的计算方案,同时鉴于BitTorrent分片获取的波动性,以及前述所讨论的滑动窗口对系统整体性能的影响,我们的自适应大小滑动窗口将根据分片获取的速度动态地自我调整大小。 From the above discussion, we can clearly realize that the size of the sliding window is a crucial parameter that affects the performance of the system. We believe that the size of the sliding window should be related to the playback start wait time Correspondingly, the client should fill the sliding window buffer within the playback waiting time, and then start playing the media stream, and continue to fill the buffer while playing, so as to ensure the continuity of the playback. In this way, regarding the optimal size of the sliding window, according to experimental statistics, the size of the sliding window should be statically set to be able to buffer 300s of media data. according to (in Indicates the playback delay of the player, that is, the time interval between the client requesting a program and the player starting to play; Indicates the size of the fragment, which is the smallest unit of media data stored in the peer node, with a size of 256K bytes; Indicates the decoding speed of the media file), the premise of the sliding window size calculation formula is that the speed of fragmentation acquisition is always at least greater than or equal to the decoding speed of the player's streaming media. In practical applications, especially for the majority of ADSL users and Streaming media data with high-quality QoS guarantee is extremely unrealistic. The initial size of our sliding window also adopts the same calculation scheme, and in view of the volatility of BitTorrent fragmentation acquisition and the impact of the sliding window on the overall performance of the system discussed above, our adaptive sliding window will be based on the speed of fragmentation acquisition Dynamically resizes itself.
一、体系结构1. Architecture
系统设计的目的是为了实现新的系统,但是新的系统如果与原有系统差异过大,就会造成系统升级的代价巨大,特别是对于大规模的网络系统,系统的重大升级会带来相当的成本和时间代价。因此,在设计基于BitTorrent的视频点播系统时要充分考虑原有系统构架,以实现两种系统之间的简单升级。从传统的视频点播系统中可以看出,它的问题主要来源于媒体服务器的瓶颈,而对于网络中的其他部分没有严重的问题。按照现在通用的共享式视频点播系统来简单的在各媒体服务器间均衡负载,无法彻底解决媒体服务器的瓶颈,而以P2P的方式将服务器的负载均衡到各节点上,才能彻底实现媒体服务器瓶颈问题的彻底解决。 The purpose of system design is to implement a new system, but if the new system is too different from the original system, the cost of system upgrade will be huge, especially for large-scale network systems, a major upgrade of the system will bring considerable cost and time cost. Therefore, when designing a BitTorrent-based video-on-demand system, the original system architecture should be fully considered in order to realize a simple upgrade between the two systems. It can be seen from the traditional video-on-demand system that its problems mainly come from the bottleneck of the media server, but there are no serious problems for other parts of the network. According to the common shared video-on-demand system to simply balance the load among the media servers, the bottleneck of the media server cannot be completely solved, but the bottleneck problem of the media server can be completely realized by balancing the load of the server to each node in a P2P manner completely resolved.
图3给出了基于BitTorrent视频点播系统的总体架构,该结构以现有成熟的大规模视频点播系统为蓝本,同时考虑到P2P的特性。不仅实现了基于BitTorrent视频点播系统,同时保证了系统的性能和可扩展性。 Figure 3 shows the overall architecture of the BitTorrent-based video-on-demand system, which is based on the existing mature large-scale video-on-demand system, while taking into account the characteristics of P2P. It not only realizes the video-on-demand system based on BitTorrent, but also guarantees the performance and scalability of the system.
从图3中可以看出,在总体架构上,它与传统的视频点播系统很相似,这也确保了在对系统升级时的改动会比较小。而整个系统的主要改变在与客户节点端,同时服务器的功能也会有所改变,它不再只是简单的返回媒体服务器的信息,还要返回邻居节点的信息。对于客户节点的改变,主要是针对于客户节点上运行的客户端,在升级过程中也相对简单的多。 As can be seen from Figure 3, in terms of overall architecture, it is very similar to the traditional video-on-demand system, which also ensures that the changes in the system upgrade will be relatively small. The main change of the whole system is on the client side, and at the same time, the function of the server will also change. It no longer simply returns the information of the media server, but also returns the information of the neighbor nodes. The change of the client node is mainly aimed at the client running on the client node, and the upgrade process is relatively simple.
二、实施例Two, the embodiment
根据图4中所示,系统在运行过程中的具体流程为: As shown in Figure 4, the specific process of the system during operation is as follows:
1、 种子文件制作服务器制作将要发布的流媒体文件,然后将种子文件发送 1. The seed file making server makes the streaming media file to be released, and then sends the seed file
EPG服务器。 EPG server.
2、 EPG服务器下载种子文件,同时根据种子信息制作电子节目单。 2. The EPG server downloads the seed file, and at the same time makes an electronic program list based on the seed information.
3、 视频点播客户端在启动后会自动向EPG服务器请求电子节目单,EPG 3. After the VOD client is started, it will automatically request the electronic program list from the EPG server, and the EPG
服务器根据该用户的信息验证是否属于注册用户,如果用户通过验证EPG服务器就会向该用户发送电子节目单,然后用户的客户端上就会出现点播节目菜单。 The server verifies whether the user belongs to the registered user according to the information of the user. If the user passes the verification, the EPG server will send the electronic program list to the user, and then the program menu on demand will appear on the user's client.
4、 根据点播菜单用户可以点播想看的节目,之后向EPG服务器请求种子文 4. According to the on-demand menu, users can order the programs they want to watch, and then request the seed file from the EPG server
件,接着EPG服务器向用户发送种子文件。 file, and then the EPG server sends the seed file to the user.
5、 当用户收到种子文件后马上进入下载过程,首先在客户端上经过改进的 5. When the user receives the torrent file, he immediately enters the download process. First, the improved
BitTorrent客户端与Tracker服务器建立连接,根据改进后的Tracker算法得到邻居节点信息。 The BitTorrent client establishes a connection with the Tracker server, and obtains neighbor node information according to the improved Tracker algorithm.
6、客户端通过这些邻居节点的信息,根据改进后的BitTorrent客户端部分 6. The client passes the information of these neighbor nodes, according to the improved BitTorrent client part
有序的下载流媒体文件,边下载边播放,同时也为其他邻居节点提供上载服务,这样用户马上就可以看到点播节目直至文件播放完毕。 Download streaming media files in an orderly manner, play while downloading, and provide upload services for other neighbor nodes at the same time, so that users can watch on-demand programs immediately until the files are played.
下面介绍一下自适应滑动窗口的实现。 The following describes the implementation of the adaptive sliding window.
首先给出如下定义: First give the following definition:
:节点下载分片的速度之和,是实时动态的反映,根据我们上面的讨论,可得知的大小有如下计算公式: :node The sum of the speed of downloading fragments is a reflection of real-time dynamics. According to our discussion above, we can know The size of is calculated by the following formula:
:初始滑动窗口大小(理想滑动窗口大小),大小可由如下公式计算 : Initial sliding window size (ideal sliding window size), the size can be calculated by the following formula
:时刻的自适应滑动窗口大小 : Adaptive sliding window size for moments
则我们定义的计算公式如下: then we define The calculation formula is as follows:
其中N是节点所请求文件的分片数,,是调整滑动窗口大小的频度。 Where N is the number of fragments of the file requested by the node, , is how often to resize the sliding window.
当的时候,即起始的时候,。 when when, at the beginning, .
因此自适应大小的滑动窗口算法可以很容易做如下实现: Therefore, the adaptive size sliding window algorithm can be easily implemented as follows:
由前面我们知道为分片数,假设一个512MB的视频文件,则,而为滑动窗口的最小值,理论上,按顺序下载的话,当分块获取速度大于缓冲区消耗速度时,可以趋近于0,即是不需要滑动窗口的存在,但是正如我们前面说讨论的,由于BT无序下载的特性,我们必须保留滑动窗口为一定的大小,通过试验,我们发现当解码速度为4Mbps,播放延迟为5s时,取值为10~20为最佳,而取值15%~20%为最佳。 From the previous we know is the number of fragments, assuming a 512MB video file, then ,and It is the minimum value of the sliding window. In theory, if downloading in sequence, when the block acquisition speed is greater than the buffer consumption speed, It can be close to 0, that is, there is no need for the existence of the sliding window, but as we discussed earlier, due to the characteristics of BT out-of-order downloading, we must keep the sliding window at a certain size. Through experiments, we found that when the decoding speed is 4Mbps, when the playback delay is 5s, The value of 10~20 is the best, and The value of 15%~20% is the best.
Claims (4)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012101479902A CN102821316A (en) | 2012-05-14 | 2012-05-14 | Improved video on demand (VOD) transmission method based on peer-to-peer computing core algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012101479902A CN102821316A (en) | 2012-05-14 | 2012-05-14 | Improved video on demand (VOD) transmission method based on peer-to-peer computing core algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102821316A true CN102821316A (en) | 2012-12-12 |
Family
ID=47305132
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2012101479902A Pending CN102821316A (en) | 2012-05-14 | 2012-05-14 | Improved video on demand (VOD) transmission method based on peer-to-peer computing core algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102821316A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103442256A (en) * | 2013-08-30 | 2013-12-11 | Tcl集团股份有限公司 | Electronic program menu implementation method and system based on HTML5 |
CN110430285A (en) * | 2019-08-27 | 2019-11-08 | 佛山职业技术学院 | A kind of playback method of Baffle Box of Bluetooth and its online audio file |
CN111190915A (en) * | 2020-01-02 | 2020-05-22 | 腾讯科技(深圳)有限公司 | Method for determining road tool identifier or role identifier, server and storage medium |
CN116566973A (en) * | 2023-06-20 | 2023-08-08 | 北京中宏立达科技发展有限公司 | File transmission system based on peer-to-peer network |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101478564A (en) * | 2008-12-31 | 2009-07-08 | 西安交通大学 | Adaptive hierarchical transmission structure design method for P2P stream media network |
CN101534204A (en) * | 2008-03-10 | 2009-09-16 | 中国网通集团宽带业务应用国家工程实验室有限公司 | Streaming media information distribution system and method thereof and user end |
CN101867777A (en) * | 2010-04-20 | 2010-10-20 | 南京邮电大学 | A video-on-demand transmission method based on a peer-to-peer computer top box |
-
2012
- 2012-05-14 CN CN2012101479902A patent/CN102821316A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101534204A (en) * | 2008-03-10 | 2009-09-16 | 中国网通集团宽带业务应用国家工程实验室有限公司 | Streaming media information distribution system and method thereof and user end |
CN101478564A (en) * | 2008-12-31 | 2009-07-08 | 西安交通大学 | Adaptive hierarchical transmission structure design method for P2P stream media network |
CN101867777A (en) * | 2010-04-20 | 2010-10-20 | 南京邮电大学 | A video-on-demand transmission method based on a peer-to-peer computer top box |
Non-Patent Citations (1)
Title |
---|
裴双迎: "基于Bit Torrent的核心算法分析与改进", 《中国优秀硕士学位论文全文数据库》, 7 December 2011 (2011-12-07) * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103442256A (en) * | 2013-08-30 | 2013-12-11 | Tcl集团股份有限公司 | Electronic program menu implementation method and system based on HTML5 |
CN103442256B (en) * | 2013-08-30 | 2017-08-04 | Tcl集团股份有限公司 | A kind of method and system that electric program menu is realized based on HTML5 |
CN110430285A (en) * | 2019-08-27 | 2019-11-08 | 佛山职业技术学院 | A kind of playback method of Baffle Box of Bluetooth and its online audio file |
CN111190915A (en) * | 2020-01-02 | 2020-05-22 | 腾讯科技(深圳)有限公司 | Method for determining road tool identifier or role identifier, server and storage medium |
CN111190915B (en) * | 2020-01-02 | 2023-05-16 | 腾讯科技(深圳)有限公司 | Prop identification or role identification determining method, server and storage medium |
CN116566973A (en) * | 2023-06-20 | 2023-08-08 | 北京中宏立达科技发展有限公司 | File transmission system based on peer-to-peer network |
CN116566973B (en) * | 2023-06-20 | 2023-11-07 | 北京中宏立达科技发展有限公司 | File transmission system based on peer-to-peer network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101478564B (en) | Adaptive hierarchical transmission structure design method for P2P stream media network | |
CN101480050B (en) | Performance aware peer-to-peer video-on-demand | |
CN101867777B (en) | Video ordering transmission method based on peer-to-peer computing set-top box | |
CN101473651B (en) | Admission control for performance aware peer-to-peer video-on-demand | |
CN102546711B (en) | Storage adjustment method, device and system for contents in streaming media system | |
CN113453038B (en) | Effectiveness optimal collaborative cache management method under CDN-P2P hybrid architecture | |
CN101742238A (en) | Streaming media on-demand method and device combining RTS with P2P | |
CN104158904B (en) | A kind of cloud auxiliary mobile P 2 P network collaboration method for down loading | |
CN101841691A (en) | Method and device for exchanging streaming media data | |
CN101150506A (en) | Content acquisition method, device and content transmission system | |
CN102158767B (en) | Scalable-coding-based peer to peer live media streaming system | |
CN102821316A (en) | Improved video on demand (VOD) transmission method based on peer-to-peer computing core algorithm | |
CN104822084A (en) | Concurrent-stream-based rapid channel switching method of P2P real-time playing system | |
Kim et al. | Efficient neighbor selection through connection switching for P2P live streaming | |
CN101459678B (en) | A fusion realization method of digital media on demand and digital resource download | |
CN101448139B (en) | A digital media on-demand method based on P2P network | |
Liang et al. | ipass: Incentivized peer-assisted system for asynchronous streaming | |
CN108833554A (en) | A large-scale network-oriented real-time highly reliable message distribution system and its method | |
CN101626400B (en) | A Compensation Method for P2P Music On Demand System | |
US9386056B1 (en) | System, method and computer readable medium for providing media stream fragments | |
Riad et al. | A framework for cloud P2P VoD system based on user's behavior analysis | |
CN109450815B (en) | Peer-to-peer network streaming media system data scheduling method based on utility value | |
CN102638704B (en) | Performance perceived peering content on-demand | |
Erman | Extending bittorrent for streaming applications | |
CN104506897B (en) | Weighted random data block selection method in P2P live broadcast stream media systems |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20121212 |