CN1717113A - A Optical-Burst/Packet-Switching Hybrid Switch Structure - Google Patents
A Optical-Burst/Packet-Switching Hybrid Switch Structure Download PDFInfo
- Publication number
- CN1717113A CN1717113A CNA2004100401111A CN200410040111A CN1717113A CN 1717113 A CN1717113 A CN 1717113A CN A2004100401111 A CNA2004100401111 A CN A2004100401111A CN 200410040111 A CN200410040111 A CN 200410040111A CN 1717113 A CN1717113 A CN 1717113A
- Authority
- CN
- China
- Prior art keywords
- optical
- output
- wavelength
- input
- burst
- 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
Images
Landscapes
- Optical Communication System (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
Description
技术领域technical field
本发明属于通信系统中的交换控制技术领域,它特别涉及光突发/分组交换网络中核心节点的交换结构。The invention belongs to the technical field of switching control in communication systems, and particularly relates to the switching structure of core nodes in optical burst/packet switching networks.
背景技术Background technique
随着因特网的规模和因特网业务的指数级增长,人们对网络带宽的需求也急剧增加,因此对网络的传输能力以及路由器、交换机处理能力的要求也在不断增加。密集波分复用技术(Dense Wavelength Division Multiplexing简称WDM)能在一根光纤上复用多个波长信道,极大的提高了单根光纤的传输容量。利用它能够建立满足因特网业务增长的宽带基础网络,较好的解决了网络传输的问题。现在的技术能做到在一根光纤上复用80-120个波长,总共约400Gb/s的传输容量。With the scale of the Internet and the exponential growth of Internet services, people's demand for network bandwidth is also increasing sharply, so the requirements for the transmission capacity of the network and the processing capacity of routers and switches are also increasing. Dense Wavelength Division Multiplexing (WDM for short) can multiplex multiple wavelength channels on one optical fiber, greatly improving the transmission capacity of a single optical fiber. Utilizing it can establish a broadband basic network to meet the growth of Internet services, and better solve the problem of network transmission. The current technology can multiplex 80-120 wavelengths on one optical fiber, with a total transmission capacity of about 400Gb/s.
在此基础上开发出了四种在WDM网络上传输数据的技术:广播—选择接收,波长路由,光突发交换(optical burst switching,OBS),光分组交换(opticalpacket switching,OPS)。波长路由提供的交换粒度(基本交换单位)为波长,不适合传输突发性较强的分组业务,需要在上面覆盖组合因特网分组业务的层次结构,这样就增加了系统的复杂性,降低了传输效率。为降低复杂性,增加效率,就要求直接在WDM光网络上传输IP层数据,这是光突发交换和光分组交换技术出现的主要原因。其中,全光的分组交换技术被认为是克服传统电交换处理瓶颈的最根本的方法,但目前该技术还存在若干难点,如全光存储、光定时/同步等。为此,充分结合并发挥现有光、电技术的优势,构建光电混合的交换系统,是实现大容量交换系统的有效方法,如光突发交换(OBS)技术,它有效地结合了波长路由和光分组交换的优势。On this basis, four technologies for transmitting data on WDM networks have been developed: broadcast-selective reception, wavelength routing, optical burst switching (optical burst switching, OBS), and optical packet switching (optical packet switching, OPS). The switching granularity (basic switching unit) provided by wavelength routing is wavelength, which is not suitable for the transmission of bursty packet services. It needs to cover the hierarchical structure of combined Internet packet services on it, which increases the complexity of the system and reduces the transmission rate. efficiency. In order to reduce complexity and increase efficiency, it is required to transmit IP layer data directly on the WDM optical network, which is the main reason for the emergence of optical burst switching and optical packet switching technologies. Among them, the all-optical packet switching technology is considered to be the most fundamental method to overcome the bottleneck of traditional electrical switching processing, but there are still some difficulties in this technology, such as all-optical storage, optical timing/synchronization, etc. For this reason, it is an effective way to realize a large-capacity switching system by fully combining and giving full play to the advantages of existing optical and electrical technologies to construct an optical-electrical hybrid switching system, such as Optical Burst Switching (OBS) technology, which effectively combines wavelength routing and the advantages of optical packet switching.
在光突发交换网络中,承载用户业务的突发数据(Burst)可以看作是由大量数据分组构成的超长分组,而这个超长分组的分组头就是突发数据的控制分组(Burst Header Packet,BHP)。BHP中包含了突发数据的有关信息,如偏移时间(Offset Time,即源端发送BHP和相应突发数据之间的间隔时间)、突发长度、数据信道(波长)等。与传统分组交换不同的是,BHP与突发数据在物理信道上是分离的,即采用带外信令:在密集波分复用传输系统中,可以采用一个(或多个)专门的波长信道作为控制信道,用于传送BHP,而把其他的波长信道作为数据信道,BHP与突发数据间一一对应。In the optical burst switching network, the burst data (Burst) carrying user services can be regarded as an ultra-long packet composed of a large number of data packets, and the packet header of this ultra-long packet is the control packet (Burst Header) of the burst data. Packet, BHP). The BHP contains information about the burst data, such as offset time (Offset Time, that is, the interval between the source sending the BHP and the corresponding burst data), burst length, data channel (wavelength), etc. Different from traditional packet switching, BHP and burst data are separated on the physical channel, that is, out-of-band signaling is used: in a dense wavelength division multiplexing transmission system, one (or more) dedicated wavelength channels can be used As a control channel, it is used to transmit BHP, and other wavelength channels are used as data channels, and there is a one-to-one correspondence between BHP and burst data.
在光突发交换中,为了减小连接建立的延迟,提高了信道利用率,一般采用单向预约机制,即在源端真正发送突发数据前,先发送一个BHP,经过一定的偏移时间后再发送突发数据,使BHP先于突发数据到达光突发交换网络的中间交换节点进行资源预留。BHP在中间交换节点转换为电信号进行处理,包括出端口波长的预约以及交换矩阵的配置等,保证当突发数据到达时相应的全光数据信道已经配置好,从而实现数据在光域的透明传输。从完成的功能来看,BHP与传统电路交换网络中的信令非常相似,正是在这个意义上,BHP也被称为信令消息。但与传统信令不同的是,光突发交换信令是一种单向预约机制,而不必等待目的端的反馈确认。In optical burst switching, in order to reduce the delay of connection establishment and improve channel utilization, a one-way reservation mechanism is generally adopted, that is, before the source end actually sends burst data, it first sends a BHP, and after a certain offset time Then, the burst data is sent, so that the BHP arrives at the intermediate switching node of the optical burst switching network before the burst data for resource reservation. BHP converts electrical signals into electrical signals for processing at intermediate switching nodes, including reservation of outgoing port wavelengths and configuration of switching matrix, etc., to ensure that when burst data arrives, the corresponding all-optical data channel has been configured, so as to realize the transparency of data in the optical domain transmission. From the perspective of completed functions, BHP is very similar to signaling in traditional circuit switching networks. It is in this sense that BHP is also called signaling messages. But different from traditional signaling, optical burst switching signaling is a one-way reservation mechanism without waiting for the feedback confirmation from the destination end.
众所周知,和点到点的传输线路不同,一个网络中各节点必需具有把输入数据在特定时间从特定输出通道输出的能力,这种能力称为交换数据的能力,它是通过具体的交换结构来实现的。而不同的交换结构实现交换的能力有比较大的差异,因此交换结构的好坏直接影响网络的性能。As we all know, unlike point-to-point transmission lines, each node in a network must have the ability to output input data from a specific output channel at a specific time. This ability is called the ability to exchange data, which is achieved through a specific switching structure Achieved. And different switching structures have relatively large differences in the ability to realize switching, so the quality of the switching structure directly affects the performance of the network.
一个光突发交换网络主要由边缘节点、核心节点和DWDM链路构成(见附图1)。其中,DWDM链路包括了控制信道(Control Channel)和数据信道(Data Channel)。核心节点的结构示意图见附图2。光交换模块上方的所有模块统称交换控制单元(Switch Control Unit,SCU),它控制光交换模块的运作使得突发数据正确通过。而光交换矩阵将根据交换控制单元的指令在指定时间把指定输入端口的指定波长信道上的突发数据交换到相应的输出波长信道。BHP和突发数据分离的结构可以避免在核心节点将所有数据转换到电域,也避免了大量存储器的使用。对于光突发交换和光分组交换网络的核心节点,其光交换模块具有通用性,下面仅以光突发交换中的光交换模块为例。同时,本发明侧重于光交换模块,因此在下面的介绍中略去了交换控制单元。An optical burst switching network is mainly composed of edge nodes, core nodes and DWDM links (see Figure 1). Among them, the DWDM link includes a control channel (Control Channel) and a data channel (Data Channel). The schematic diagram of the core node structure is shown in Figure 2. All the modules above the optical switching module are collectively referred to as Switch Control Unit (SCU), which controls the operation of the optical switching module so that burst data can pass through correctly. The optical switch matrix will switch the burst data on the specified wavelength channel of the specified input port to the corresponding output wavelength channel at a specified time according to the instruction of the switch control unit. The structure of separating BHP and burst data can avoid converting all data to the electric domain at the core node, and also avoids the use of a large amount of memory. For the core nodes of optical burst switching and optical packet switching networks, the optical switching modules are universal, and the optical switching modules in optical burst switching are used as examples below. Meanwhile, the present invention focuses on the optical switch module, so the switch control unit is omitted in the following introduction.
现有的光交换模块主要有以下几种:(见文献Lisong Xu,Harry G.Perros,andGeorge Rouskas,North Carolina State University,“Techniques for Optical PacketSwitching and Optical Burst Switching”,IEEE Communications Magazine January 2001,Volume:39,Issue:1,Jan.2001,Pages:136-142.)Existing optical switching modules mainly include the following types: (see literature Lisong Xu, Harry G. Perros, and George Rouskas, North Carolina State University, "Techniques for Optical Packet Switching and Optical Burst Switching", IEEE Communications Magazine January 2001, Volume: 39, Issue: 1, Jan. 2001, Pages: 136-142.)
1.输入端有光延迟线(Fiber Delay Line,FDL)的经典交换结构1. Classic switching structure with Fiber Delay Line (FDL) at the input
如附图3所示,此种结构为经典的交换结构,它由调度和交换两部分组成。其中,调度部分主要解决BHP的调度和冲突解决,而交换部分则把输入的突发数据交换到指定的输出端口。因为在同一时刻不能有两个突发数据从同一FDL的输出端输出,所以这种交换结构存在排头阻塞问题,也就是在每个输入波长信道上,前面的突发数据会对其后的突发数据造成阻塞。例如,假设每个突发数据都持续时间T,若输入波长信道i的两个顺序突发数据分别要从输出波长信道j和输出波长信道k送出。若第一个突发数据需要延迟时间T,则第二个突发数据也必须延迟T以上的时间,否则在时刻T有两个突发数据从同一个缓存波长信道输出,造成冲突。但是若第一个突发数据需要延迟nT(n>1),那么只要在端口k没有冲突第二个突发数据就可以不延时发送。由于排头阻塞的存在,FDL解决冲突的能力会受到很大影响,对丢失性能的改善也将是很有限的。同时,在这种结构中,FDL资源无法实现共享,造成代价较高。As shown in Figure 3, this structure is a classic switching structure, which consists of two parts: scheduling and switching. Among them, the scheduling part mainly solves the scheduling and conflict resolution of BHP, while the switching part switches the input burst data to the designated output port. Because two bursts of data cannot be output from the same FDL output at the same time, there is a head-of-line blocking problem in this switching structure, that is, on each input wavelength channel, the front burst data will interfere with the following burst. Sending data causes blockage. For example, assuming that each burst of data lasts time T, if two sequential bursts of input wavelength channel i are to be sent from output wavelength channel j and output wavelength channel k respectively. If the first burst of data needs to be delayed by T, the second burst of data must also be delayed for more than T, otherwise at time T, two bursts of data are output from the same buffered wavelength channel, causing conflicts. However, if the first burst of data needs to be delayed by nT (n>1), then the second burst of data can be sent without delay as long as there is no conflict at port k. Due to the existence of head-of-line blocking, the ability of FDL to resolve conflicts will be greatly affected, and the improvement of loss performance will be very limited. At the same time, in this structure, FDL resources cannot be shared, resulting in high cost.
2.广播—选择交换结构2.Broadcast—select switching structure
如附图4所示,此种结构有波长编码、缓存、波长选择三个部分组成,它在一定条件下可以进行N波长输入N波长输出K个时隙的交换。这种结构要求同一时刻在每个输出端仅有一个波长的信号输出。在编码部分通过N个固定波长变换器(fixed wavelength converter,FWC)将各输入波长变换成没有相同波长的信号,然后再组合在一起送入缓存部分。在缓存部分用分解器和FDL将波长编码部分送来的信号复制在各个时隙上,紧接着利用分解器、光开关、复用器构造出一个空分交换结构把信号交换到输出端。最后在波长选择部分利用解复用器、光开关、复用器把某一时刻在某各输出端上某一波长的信号选出来,这样就达到了光交换的目的。因为是在输出端选择波长,所以输出端的波长是可变的。As shown in Figure 4, this structure consists of three parts: wavelength encoding, buffering, and wavelength selection. Under certain conditions, it can perform the exchange of N wavelength input and N wavelength output K time slots. This structure requires only one wavelength of signal output at each output port at the same time. In the encoding part, N fixed wavelength converters (fixed wavelength converter, FWC) are used to convert each input wavelength into a signal without the same wavelength, and then combined and sent to the buffer part. In the buffer part, the splitter and FDL are used to copy the signal sent by the wavelength coding part to each time slot, and then the splitter, optical switch, and multiplexer are used to construct a space division switching structure to switch the signal to the output end. Finally, in the wavelength selection part, the demultiplexer, optical switch, and multiplexer are used to select the signal of a certain wavelength on a certain output end at a certain time, so that the purpose of optical switching is achieved. Because the wavelength is selected at the output, the wavelength at the output is variable.
该结构的优点在于原理简单,控制算法易于实现。但由于广播的原因,在各信道上都充斥着该信道不需要的信息,所以资源利用率较低;而且各输出端的复用、光开关、解复用设备不能共享,需要的器件较多,代价高。而且经过分解器后光信号被分成N份,光信号的能量大大降低,信号误码的概率增加。The advantage of this structure is that the principle is simple and the control algorithm is easy to realize. However, due to broadcasting, each channel is full of unnecessary information on the channel, so the resource utilization rate is low; and the multiplexing, optical switch, and demultiplexing equipment at each output end cannot be shared, and many devices are required. The price is high. Moreover, after passing through the decomposer, the optical signal is divided into N parts, the energy of the optical signal is greatly reduced, and the probability of signal error is increased.
3.共享变换池结构3. Shared transformation pool structure
如附图5所示,此种结构由2N×2N的交换矩阵和N个TWC以及N个ODL组成,N个TWC用来将数据调整到需要输出的波长。而N个内部循环波长信道可以使得高优先级数据和低优先级数据发生冲突时将低优先级数据送入ODL再次延时,从而实现对优先级的支持。这一交换结构需要资源较多;有一半的波长并不能输出,波长利用效率不高,而且过多的波长使得AWG的代价过高;虽然能利用再次延时提供优先级,但是由于衰弱的原因光信号被延时的次数有限制,所以提供的优先级效果受到制约。As shown in Figure 5, this structure consists of a 2N×2N switch matrix, N TWCs and N ODLs, and the N TWCs are used to adjust the data to the output wavelength. The N internal cycle wavelength channels can make the low-priority data sent to the ODL to be delayed again when the high-priority data and low-priority data conflict, so as to realize the support for the priority. This switching structure requires more resources; half of the wavelengths cannot be output, the wavelength utilization efficiency is not high, and too many wavelengths make the AWG cost too high; although it can use the delay again to provide priority, but due to weak There is a limit to the number of times the optical signal can be delayed, so the priority effects provided are limited.
从以上阐述可知目前光突发交换研究中主要存在以下问题:From the above description, it can be known that the current optical burst switching research mainly has the following problems:
1)光突发交换提供的是一种尽力服务(Best effort Services),丢失率的理论下限为爱尔兰公式计算的全忙概率。相对于电域的分组交换而言,其丢失率过高。1) Optical burst switching provides Best Effort Services, and the theoretical lower limit of the loss rate is the full-busy probability calculated by the Irish formula. Compared with packet switching in the electrical domain, its loss rate is too high.
2)FDL不能提供随机访问,只能提供固定时延,设备体积大,有光功率损耗。而且其级数设计和深度设计较复杂,须考虑的因素较多。数据缓存和解决冲突的能力有限。2) FDL cannot provide random access, but can only provide fixed delay, the equipment is bulky and has optical power loss. Moreover, its series design and depth design are more complicated, and there are many factors to be considered. Data caching and limited ability to resolve conflicts.
发明内容Contents of the invention
本发明的目的是为光突发/分组交换网络中的核心节点提供一种支持光突发/分组交换的光电混合交换结构,采用本发明提出的交换结构,能有效降低丢失率,并能灵活提供对优先级的支持。The purpose of the present invention is to provide a kind of optical burst/packet switching optical hybrid switching structure for the core node in the optical burst/packet switching network, adopt the switching structure proposed by the present invention, can effectively reduce the loss rate, and can flexibly Provides support for prioritization.
为描述方便,先给出如下定义:For the convenience of description, the following definitions are first given:
边缘节点:位于OBS网络边缘的节点,在功能上主要实现突发数据的产生和光电转换。Edge node: A node located at the edge of the OBS network, which mainly realizes the generation of burst data and photoelectric conversion in terms of function.
核心节点:光突发交换网络中的非边缘节点,实现光突发数据的光交换。Core node: a non-edge node in the optical burst switching network, which realizes optical switching of optical burst data.
交换:把特定时间从特定输入端口的特定波长信道送入的突发数据在特定时间从特定输出端口的特定波长信道送出的过程。Switching: The process of sending burst data input from a specific wavelength channel of a specific input port at a specific time from a specific wavelength channel of a specific output port at a specific time.
丢失率:网络中数据包的丢失比率。Loss rate: The loss rate of data packets in the network.
冲突:多个突发数据包同时要到同一输出端口的某个特定波长信道上去的时候,就发生了“冲突”,在分组交换机中也称为“外部阻塞”。Collision: When multiple burst data packets go to a specific wavelength channel of the same output port at the same time, "collision" occurs, which is also called "external blocking" in packet switches.
时延:突发数据送入和输出某一设备的时间差。Latency: The time difference between burst data being sent into and out of a device.
传输错误:突发数据通过OBS网络从源端传输到了目的端,但是由于传输过程引入错误,使得源端发出的数据和目的端收到的数据不一致,这被成为传输错误。Transmission error: The burst data is transmitted from the source end to the destination end through the OBS network, but due to errors introduced during the transmission process, the data sent by the source end is inconsistent with the data received by the destination end, which is called a transmission error.
本发明提供的一种支持光突发/分组交换的光电混合交换结构(如图6所示),它包括:一个光交换矩阵,所述的光交换矩阵有(M+N×n)个输入端口和(M+N×n)个输出端口,M表示缓存信道的个数,N表示入端口或出端口的个数,N×n表示输入(输出)波长信道的个数,n表示每一个入端口或出端口包含的波长信道数;N个入端口、N个出端口,N个分波器、N个合波器,N×n个可调波长变换器(TWC);入端口1通过分波器分解出n个波长信道,n个波长信道利用光纤分别连接到光交换矩阵的任意n个输入端口;入端口2通过分波器分解出n个波长信道,n个波长信道利用光纤分别连接到光交换矩阵的任意n个输入端口;其余各入端口依此类推,构成N个入端口;所有入端口共使用光交换矩阵的N×n个输入端口;与入端口1对应,从光交换矩阵的n个输出端口输出的n个波长信道通过光纤连接到n个可调波长变换器,各可调波长变换器通过光纤连接到合波器形成出端口1;与入端口2对应,从光交换矩阵的n个输出端口输出的n个波长信道通过光纤连接到n个可调波长变换器,各可调波长变换器通过光纤连接到合波器形成出端口2;其余各出端口依此类推,构成N个出端口;所有出端口共使用光交换矩阵的N×n个输出端口;A kind of photoelectric hybrid switching structure (as shown in Figure 6) that supports optical burst/packet switching provided by the present invention, it comprises: an optical switching matrix, and described optical switching matrix has (M+N*n) input port and (M+N×n) output ports, M represents the number of buffered channels, N represents the number of incoming or outgoing ports, N×n represents the number of input (output) wavelength channels, and n represents each The number of wavelength channels included in the input port or output port; N input ports, N output ports, N wave splitters, N wave combiners, N×n tunable wavelength converters (TWC); the
其特征是它还包括:It is characterized in that it also includes:
由光电转换器、RAM缓存、电光转换器依次通过电域连接组成一个缓存信道,共有M个缓存信道;每一个缓存信道的输入端通过光纤连接在光交换矩阵的输出端口,每一个缓存信道的输出端通过光纤连接在光交换矩阵的输入端口。A buffer channel is composed of photoelectric converters, RAM caches, and electro-optical converters connected in turn through the electrical domain, and there are M buffer channels in total; the input end of each buffer channel is connected to the output port of the optical switch matrix through optical fibers, and the output port of each buffer channel The output end is connected to the input port of the optical switch matrix through an optical fiber.
需要说明的是,It should be noted,
上面所述的光交换矩阵的输入端口和输出端口,只能和一个信道相连接。The input port and output port of the optical switch matrix mentioned above can only be connected to one channel.
分波器的作用是将输入的多波长复合光信号分解成多个单波长光信号输出,各单波长光信号对应不同的波长。The role of the demultiplexer is to decompose the input multi-wavelength composite optical signal into multiple single-wavelength optical signals for output, and each single-wavelength optical signal corresponds to a different wavelength.
光交换矩阵的作用是将指定时段光交换矩阵某输入端口输入的光信号交换到光交换矩阵的指定输出端口输出。The function of the optical switch matrix is to switch the optical signal input from a certain input port of the optical switch matrix to the specified output port of the optical switch matrix for output during a specified time period.
光电转换器的作用是将输入的任意波长的光信号通过光电转换变成电信号输出;The function of the photoelectric converter is to convert the input optical signal of any wavelength into an electrical signal output through photoelectric conversion;
RAM缓存的作用是将输入的电信号存储一定时间再输出该电信号;电光转换器的作用是将由RAM缓存输入的电信号重新转换成某一波长的光信号。The function of the RAM buffer is to store the input electrical signal for a certain period of time and then output the electrical signal; the function of the electro-optic converter is to reconvert the electrical signal input by the RAM buffer into an optical signal of a certain wavelength.
可调波长变换器的作用是将输入的任意波长的光信号转换成指定输出波长的光信号输出。The function of the tunable wavelength converter is to convert the input optical signal of any wavelength into the output optical signal of the specified output wavelength.
合波器的作用是将输入的多个单波长光信号复合成一个多波长光信号输出,且各输入单波长光信号对应不同的波长。The function of the multiplexer is to combine multiple input single-wavelength optical signals into one multi-wavelength optical signal for output, and each input single-wavelength optical signal corresponds to a different wavelength.
本发明的工作过程:Working process of the present invention:
包含多个突发数据的多波长复合光信号通过光纤进入本发明交换结构的入端口,经过分波器后分解为不同波长的单波长光信号(每个波长上同时有且仅有一个突发数据在传送)并通过光纤连接到光交换矩阵的输入端口。突发数据到达光交换矩阵后,根据对先前收到的对应突发控制分组的处理结果将突发数据交换到光交换矩阵特定的输出端口输出。若没有冲突则交换到光交换矩阵与交换结构出端口对应的某个输出端口;若有冲突发生则交换到光交换矩阵与缓存信道对应的某个输出端口。对于发生冲突的的突发数据在从光交换矩阵输出端口输出后通过光纤送入缓存信道进行延时从而解决冲突。首先突发数据进入光电转换器转换为电信号,然后通过电连接送入RAM缓存进行延时,紧接着从RAM缓存中取出通过电连接送到电光转换器转换为光信号,最后从电光转换器通过光纤送回到光交换矩阵的输入端口。经过缓存信道延时后的光突发数据根据对应突发控制分组的处理结果重新无冲突的交换到出端口进行输出。对于没有冲突的突发数据以及经过缓存信道解决了冲突的突发数据在从光交换矩阵输出后经过可调波长变换器,变换到指定波长。多个已变换到指定波长的突发数据(变换后的突发数据使用的波长各不相同)通过光纤连接到合波器,经过合波器处理复合成包含多个突发数据的多波长复合光信号送到出端口输出。The multi-wavelength composite optical signal comprising a plurality of burst data enters the ingress port of the switching structure of the present invention through the optical fiber, and is decomposed into single-wavelength optical signals of different wavelengths after passing through the wave splitter (there is only one burst at the same time on each wavelength). The data is being transmitted) and connected to the input port of the optical switch matrix through optical fiber. After the burst data arrives at the optical switch matrix, the burst data is switched to a specific output port of the optical switch matrix for output according to the processing result of the previously received corresponding burst control packet. If there is no conflict, switch to an output port corresponding to the output port of the optical switch matrix and the switch structure; if there is a conflict, switch to a certain output port corresponding to the optical switch matrix and the buffer channel. For the conflicting burst data, after being output from the output port of the optical switch matrix, it is sent to the buffer channel through the optical fiber for delay so as to resolve the conflict. First, the burst data enters the photoelectric converter and converts it into an electrical signal, then sends it to the RAM cache through an electrical connection for delay, and then takes it out from the RAM cache and sends it to the electro-optical converter through an electrical connection to convert it into an optical signal, and finally from the electro-optic converter It is sent back to the input port of the optical switch matrix through the optical fiber. The optical burst data after the buffer channel delay is switched to the output port again without conflict according to the processing result of the corresponding burst control packet for output. For the burst data without conflict and the burst data whose conflict is resolved through the buffer channel, after being output from the optical switching matrix, it is converted to a designated wavelength through an adjustable wavelength converter. Multiple burst data that have been converted to specified wavelengths (the converted burst data use different wavelengths) are connected to the multiplexer through an optical fiber, and are processed by the multiplexer to form a multi-wavelength multiplexer that contains multiple burst data. The optical signal is sent to the output port for output.
本发明的创新:Innovation of the present invention:
本发明所提出的交换结构的主要思想是以光交换为主,采用由光电转换器、RAM缓存、电光转换器依次通过电域连接组成的缓存信道,以电域缓存代替光缓存FDL,并适当辅以电域交换。与光缓存FDL相比,它具有有效降低网络的丢失率、减少差错/丢失重传的时延、有效支持多优先级等特点。The main idea of the switching structure proposed by the present invention is based on optical switching, adopts a buffer channel composed of photoelectric converters, RAM caches, and electro-optical converters connected through electrical domains in turn, replaces optical buffer FDL with electrical domain buffers, and appropriately Supplemented by electric field exchange. Compared with the optical buffer FDL, it has the characteristics of effectively reducing the loss rate of the network, reducing the delay of error/loss retransmission, and effectively supporting multiple priorities.
本发明提供的一种支持光突发/分组交换的光电混合交换结构的优点是:The advantages of a photoelectric hybrid switching structure supporting optical burst/packet switching provided by the present invention are:
1)减小网络的丢失率,提高网络性能。光网络中引起数据丢失的最主要原因是冲突。本交换结构利用电域缓存能有效解决冲突,而且其解决冲突的能力大大优于光域解决冲突的方案。见附图7,ti1表示第i个突发数据的原始到达时间,ti2表示第i个突发数据的原始离开时间,ti1’表示第i个突发数据通过缓存后的到达时间,ti2’表示第i个突发数据通过缓存后的离开时间。如图所示三个突发数据的BHP依次到来,它们给出了相应突发数据的预计到达时间t11,t21,t31,以及离开时间t12,t22,t32。从图上可以看出突发数据3到来时突发数据1正在发送,必需把突发数据3缓存延时后再发送才能解决冲突。如果采用FDL来解决冲突,则延迟的时间必需是固定时间(设为T)的整数倍。这样就有可能造成即使有能够传送数据的空隙,但是因为延迟固定时间而使得数据不能在那段空隙中发送。在附图中可以发现,突发数据3通过光延时线延时T以后其发送过程又和突发数据2发生冲突,即使突发数据1和突发数据2之间的时间间隙可以发送突发数据3。从示意图可知突发数据3必需做4T的延时才能无冲突发送,这样就增加了突发数据从源到目的的发送时间,而且经过4T延时后的光信号衰弱增加,更容易引入错误。如果我们采用本专利提出的光电缓存结构则可以避免这些问题。如图所示,突发数据3在t31时到达和突发数据1发生冲突,对其进行光电转换在电域RAM中缓存,在经过Δt(Δt=t31’-t31)的延时后再进行电光转换,重新送回到光域传输。由于电域RAM的延时是任意的,经过小于T的延时Δt后就可以利用突发数据1和突发数据2之间的空隙传输突发数据3。因为Δt<<4T,所以大大缩短了从源到目的的传输时间;而且电光转换出的光信号是重新产生的光信号不存在因为延时而衰弱的问题,这也降低了传输错误的概率。1) Reduce the loss rate of the network and improve network performance. The most important cause of data loss in optical networks is collision. This switching structure can effectively resolve conflicts by utilizing the electric domain cache, and its ability to resolve conflicts is much better than that of optical domain conflict resolution schemes. As shown in Figure 7, ti1 represents the original arrival time of the i-th burst data, ti2 represents the original departure time of the i-th burst data, ti1' represents the arrival time of the i-th burst data after passing through the cache, and ti2' represents The departure time after the i-th burst data passes through the cache. As shown in the figure, the BHPs of the three bursts arrive sequentially, and they give the expected arrival times t11, t21, t31, and departure times t12, t22, t32 of the corresponding bursts. It can be seen from the figure that when the burst data 3 arrives, the
2)减少差错/丢失重传的时延,降低因重传而造成的网络负荷,从而一定程度上增加了有效带宽。OBS只提供尽力服务,也就是说它不保证服务的质量,因此对于冲突丢包和传输错误的处理是由上层协议完成的。当出现冲突丢包时,上层协议只能采用重传的方式解决;而对于传输错误上层协议可以通过高效的编码方案来纠正,或者要求源端重传。高效的编码方案的确可以纠正一定的错误,但是当错误较多时用于纠错的冗余代价太高,反而不如重传来得有效。可见重传这种方式是提供可靠传输的重要手段,但是重传会使得网络负荷加重,降低网络的有效传输性能。从1)的分析中可以看出本专利提出的交换结构能减小丢失率,降低出现传输错误的概率。所以这也就减小了重传的概率,也就是间接降低了网络的负荷,增加了有效带宽。2) Reduce the time delay of error/loss retransmission and reduce the network load caused by retransmission, thereby increasing the effective bandwidth to a certain extent. OBS only provides best-effort service, that is to say, it does not guarantee the quality of service, so the handling of conflict packet loss and transmission error is completed by the upper layer protocol. When there is a conflict and packet loss, the upper layer protocol can only solve it by retransmission; for transmission errors, the upper layer protocol can correct it through an efficient coding scheme, or require the source to retransmit. Efficient coding schemes can indeed correct certain errors, but when there are many errors, the redundancy cost for error correction is too high, and it is not as effective as retransmission. It can be seen that retransmission is an important means to provide reliable transmission, but retransmission will increase the network load and reduce the effective transmission performance of the network. From the analysis of 1), it can be seen that the switching structure proposed in this patent can reduce the loss rate and the probability of transmission errors. So this also reduces the probability of retransmission, that is, indirectly reduces the load on the network and increases the effective bandwidth.
3)可更加灵活方便地为不同业务提供不同的优先级。目前在光突发交换中能提供QoS的方法不多,大多数方法都是基于偏移时间或突发汇聚的。基于偏移时间的方法会引入较大的时延,而且能区分的优先级非常有限。基于突发汇聚的方法是在边缘节点实现的,对于核心网络QoS的提供帮助不大。其它方法,如用FDL提供有限缓存等,也由于FDL自身的缺陷,提供的QoS有限,且不灵活。在本专利提出的交换结构中,可利用电域成熟的技术,轻松得到几乎没有限制的存储时间和存储容量,同时还可随机访问。因此我们可以对不同优先级的业务分配出满足它们优先级要求的传输时隙,也就是说我们可以获得和纯电域传输网络几乎相同的优先级支持。3) Different priorities can be provided for different services more flexibly and conveniently. At present, there are not many methods that can provide QoS in optical burst switching, most of which are based on offset time or burst aggregation. The method based on offset time will introduce a large delay, and the priority that can be distinguished is very limited. The method based on burst aggregation is implemented at the edge nodes, which is not very helpful for providing core network QoS. Other methods, such as using FDL to provide limited cache, also provide limited QoS and are inflexible due to the defects of FDL itself. In the switching structure proposed in this patent, the mature technology of the electric field can be used to easily obtain almost unlimited storage time and storage capacity, and at the same time, random access is also possible. Therefore, we can allocate transmission time slots that meet their priority requirements for services with different priorities, that is to say, we can obtain almost the same priority support as the pure electric domain transmission network.
4)可在一定程度上支持波长粒度的交换。在目前的交换结构中,突发数据的丢失率已经比电域的分组交换高很多,当需要较长时间的电路交换(即有波长交换)时,将引入更高的丢失率。而在本结构中,在保证突发丢失率稳定的情况下,可使用端口中的一些波长信道提供一定波长交换。4) It can support the exchange of wavelength granularity to a certain extent. In the current switching structure, the loss rate of burst data is much higher than that of packet switching in the electrical domain. When circuit switching (that is, wavelength switching) is required for a long time, a higher loss rate will be introduced. However, in this structure, under the condition that the burst loss rate is guaranteed to be stable, some wavelength channels in the port can be used to provide certain wavelength switching.
综上所述,本专利提出的交换结构能有效降低网络的丢失率,减少差错/丢失重传的时延,有效支持多优先级,提供波长粒度的交换,是一种性能良好的交换结构。To sum up, the switching structure proposed in this patent can effectively reduce the loss rate of the network, reduce the error/loss retransmission delay, effectively support multiple priorities, and provide wavelength granular switching. It is a switching structure with good performance.
附图说明Description of drawings
图1是现有的OBS网络结构示意图Figure 1 is a schematic diagram of the existing OBS network structure
图2是核心节点结构示意图Figure 2 is a schematic diagram of the core node structure
其中,IDC表示输入数据信道,ODC表示输出数据信道,ICC表示输入控制信道,OCC表示输出控制信道,FDL表示光纤延迟线,BCP表示突发控制分组。Among them, IDC represents the input data channel, ODC represents the output data channel, ICC represents the input control channel, OCC represents the output control channel, FDL represents the fiber delay line, and BCP represents the burst control packet.
图3输入缓存式交换结构示意图Figure 3 Schematic diagram of input buffer switching structure
其中, 表示可调波长变换器, 表示光延迟线。in, Indicates a tunable wavelength converter, represents an optical delay line.
图4广播选择式交换结构示意图Figure 4 Schematic Diagram of Broadcast Selective Switching Structure
其中,K表示光延迟线的最大延时段数;N表示该交换结构的输入波长信道的个数;分解器的作用是将一个信号复制出多个相同的信号;组合器的作用是将多个信道组合为一个信道;光开关的作用是允许或者禁止光信号的通过。表示固定波长变换器, 表示光延迟线, 表示分解器,表示光开关, 表示组合器。Among them, K represents the maximum number of delay segments of the optical delay line; N represents the number of input wavelength channels of the switch structure; the function of the resolver is to copy a signal into multiple identical signals; the function of the combiner is to combine multiple The channels are combined into one channel; the role of the optical switch is to allow or prohibit the passage of optical signals. denotes a fixed wavelength converter, represents an optical delay line, Indicates the resolver, Indicates the optical switch, Represents a combiner.
图5共享变换池结构示意图Figure 5 Schematic diagram of shared transformation pool structure
其中,N表示该交换结构的输入波长信道的个数;P表示光缓存信道的个数;其中, 表示可调波长变换器, 表示光延迟线。Wherein, N represents the number of the input wavelength channels of the switch structure; P represents the number of optical buffer channels; wherein, Indicates a tunable wavelength converter, represents an optical delay line.
图6本发明能够降低丢失率的突发交换结构示意图其中, ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄表示光域连接, ---------------表示电域连接Figure 6. Schematic diagram of the burst switching structure capable of reducing the loss rate of the present invention. Among them,  ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄indicates optical domain connection, ---------------indicates electrical domain connection
图7电域解决冲突示意图Figure 7 Schematic diagram of conflict resolution in the electrical domain
其中,T表示光延迟线(FDL)的单位固定时延;ti1表示第i个突发数据的原始到达时间;ti2表示第i个突发数据的原始离开时间;ti1’表示第i个突发数据通过缓存后的到达时间;ti2’表示第i个突发数据通过缓存后的离开时间;Δt(Δt=t31’-t31)表示使得突发数据无冲突发送的最小电域延时。Among them, T represents the unit fixed delay of the optical delay line (FDL); ti1 represents the original arrival time of the i-th burst data; ti2 represents the original departure time of the i-th burst data; ti1' represents the i-th burst The arrival time of data after passing through the buffer; ti2' represents the departure time of the i-th burst data after passing through the buffer; Δt(Δt=t31'-t31) represents the minimum electrical domain delay for burst data to be sent without conflict.
表示突发控制分组到达时间,_________________表示光延迟线固定时延,■表示发生冲突的部分, 表示突发数据占用时间,表示光延时后能无冲突发送 Indicates the arrival time of the burst control packet, _________________ indicates the fixed delay of the optical delay line, ■ indicates the part where the conflict occurs, Indicates the burst data occupation time, Indicates that it can be sent without conflict after optical delay
图8冲突后100%不丢包实施例示意图Figure 8 is a schematic diagram of an embodiment of 100% no packet loss after the conflict
图9冲突后部分不丢包实施例示意图Figure 9 is a schematic diagram of an embodiment of partial non-packet loss after a conflict
具体实施方式Detailed ways
实施例一、冲突后100%不丢包实施例
见附图8,我们采用32×32的无内部阻塞光交换矩阵;入端有两个输入端口各8个波长信道,共16个输入波长信道作为信号的输入;出端同样也有两个输出端口各8个波长信道,共16个输出波长信道作为信号的输出,和入端不同的是16个波长变换器可以将输出的信号调整到需要的波长上;入端和出端剩下的16个波长信道和16个光电转换器,16个RAM缓存,以及16个电光转换器合起来组成电缓存部分提供16个电缓存信道起到解决冲突的作用。As shown in Figure 8, we use a 32×32 non-internal blocking optical switch matrix; the input port has two input ports each with 8 wavelength channels, and a total of 16 input wavelength channels are used as signal input; the output port also has two
当该节点开始工作时,电缓存部分没有存储数据,所有的数据都从16个输入波长信道输入。此时发生冲突的最坏情况是16个波长信道同时各有一个突发数据需要发送,而且所有16个数据都需要送到同一个输出波长信道上。这种情况下解决冲突的办法是把16个数据中的一个从其需要输出的波长信道送出,而将剩下的15个突发数据送入15个不同的电缓存信道,然后再选择合适的时间无冲突的送出。所以16个输入输出波长信道16个电缓存信道可以保证在初始工作时100%的冲突后不丢包。但这还不是发生冲突的最坏情况,最坏情况是在交换矩阵的入端16个输入波长信道和16个电缓存信道同时有32个突发数据需要发送。根据对光突发交换系统的介绍,我们知道在突发数据到来之前我们处理了它们相应的BHP,也就是说我们是预先知道冲突情况的。因此16个电缓存信道发出的突发数据是彼此不冲突的,也就是说16个电缓存信道发出的突发数据是分别送往16个输出波长信道即16个输出波长信道上都有突发数据需要送出。那么冲突只可能发生在16个输入波长信道送入的突发数据之间以及输入波长信道送入的突发数据和电缓存送入的突发数据之间。从各输出波长信道上来看,当n个突发数据发生冲突时我们会选择其中一个送出,而将其余n-1个送入缓存延时解决冲突。所以在16个输出波长信道均有数据的情况下,16个突发数据被输出波长信道送出,而32-16=16个突发数据被分别送入16个电缓存信道延时。这样在最坏的情况下我们也能提供100%的冲突后不丢包。当然还需要考虑的问题是,缓存的容量能不能无限大,以至能够装下所有冲突后送入的突发数据;上层协议对缓存的时间是否有限制,如果上层协议对数据的整个传输时间有限制那么缓存时间过长也会让上层协议认为是丢包而重传。综上所述,在理想情况下,若电缓存信道数M等于输入输出总波长信道数N×n,则本专利提出的交换结构可以提供100%的冲突后不丢包保护。When the node starts to work, there is no data stored in the electrical cache, and all data is input from 16 input wavelength channels. The worst case of conflict at this time is that each of the 16 wavelength channels needs to send a burst of data at the same time, and all 16 data need to be sent to the same output wavelength channel. In this case, the way to resolve the conflict is to send one of the 16 data from the wavelength channel that needs to be output, and send the remaining 15 burst data to 15 different electrical buffer channels, and then select the appropriate one. The time is sent without conflict. Therefore, 16 input and output wavelength channels and 16 electrical buffer channels can guarantee no packet loss after 100% conflicts in the initial operation. But this is not the worst case of conflicts. The worst case is that there are 32 bursts of data to be sent at the same time in the 16 input wavelength channels and 16 electrical buffer channels at the input end of the switch matrix. According to the introduction of the optical burst switching system, we know that we have processed their corresponding BHP before the burst data arrives, that is to say, we know the conflict situation in advance. Therefore, the burst data sent by the 16 electrical buffer channels do not conflict with each other, that is to say, the burst data sent by the 16 electrical buffer channels are respectively sent to the 16 output wavelength channels, that is, there are bursts on all 16 output wavelength channels. Data needs to be sent. Then conflicts can only occur between the burst data sent by the 16 input wavelength channels and between the burst data sent by the input wavelength channel and the burst data sent by the electrical buffer. From the perspective of each output wavelength channel, when n bursts of data collide, we will select one of them to send, and send the remaining n-1 bursts to the buffer to delay the resolution of the conflict. Therefore, in the case that all 16 output wavelength channels have data, 16 burst data are sent out by the output wavelength channels, and 32-16=16 burst data are respectively sent to 16 electrical buffer channels for delay. In this way we can also provide 100% no packet loss after collision in the worst case. Of course, the question that still needs to be considered is whether the capacity of the cache can be infinitely large enough to hold all the burst data sent after the conflict; If the limit is too long, the upper layer protocol will think that the packet is lost and retransmit it if the cache time is too long. To sum up, under ideal conditions, if the number of electrical buffer channels M is equal to the total number of input and output wavelength channels N×n, the switching structure proposed in this patent can provide 100% protection against packet loss after collision.
实施例二、冲突后存在一定的丢失率的实施例
在实际应用中如果允许一定程度的数据丢失率,那么附图9的配置也能取得较好的性价比。如图所示,依然采用32×32的无阻塞光交换矩阵;入端有三个输入端口各8个波长信道,共24个输入波长信道作为信号的输入;出端有三个输出端口各8个波长信道,共24个输出波长信道作为信号的输出,24个波长变换器可以将输出的信号调整到需要的波长上;入端和出端剩下的8个波长信道和8个光电转换器,8个RAM缓存,以及8个电光转换器合起来组成电缓存部分提供8个电缓存信道起到解决冲突的作用。增加的8个输入输出波长信道可以增加输入输出的能力,支持更多的业务通过。出现最坏冲突情况的概率非常小,而8个电缓存信道能提供对冲突的较好支持。也就是说在不改变交换矩阵大小的情况下,通过牺牲一点数据丢失率来取得更大的业务传输能力。因为高速光交换矩阵随着端口的增加其造价增加非常大,所以这种结构既控制了成本,又保证丢失性能。In practical applications, if a certain degree of data loss rate is allowed, then the configuration in FIG. 9 can also achieve better cost performance. As shown in the figure, a 32×32 non-blocking optical switch matrix is still used; the input end has three input ports with 8 wavelength channels each, and a total of 24 input wavelength channels are used as signal input; the output end has three output ports with 8 wavelength channels each channels, a total of 24 output wavelength channels are used as signal output, and 24 wavelength converters can adjust the output signal to the required wavelength; the remaining 8 wavelength channels and 8 photoelectric converters at the input and output ends, 8 A RAM cache and 8 electrical-optical converters are combined to form the electrical cache part to provide 8 electrical cache channels to resolve conflicts. The additional 8 input and output wavelength channels can increase the input and output capabilities and support more business passing. The probability of the worst case of collision is very small, and 8 electrical buffer channels can provide better support for collision. That is to say, without changing the size of the switching matrix, a greater service transmission capacity can be obtained by sacrificing a little data loss rate. Because the cost of the high-speed optical switch matrix increases greatly with the increase of ports, this structure not only controls the cost, but also ensures the loss of performance.
Claims (3)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2004100401111A CN1717113A (en) | 2004-07-02 | 2004-07-02 | A Optical-Burst/Packet-Switching Hybrid Switch Structure |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2004100401111A CN1717113A (en) | 2004-07-02 | 2004-07-02 | A Optical-Burst/Packet-Switching Hybrid Switch Structure |
Publications (1)
Publication Number | Publication Date |
---|---|
CN1717113A true CN1717113A (en) | 2006-01-04 |
Family
ID=35822455
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNA2004100401111A Pending CN1717113A (en) | 2004-07-02 | 2004-07-02 | A Optical-Burst/Packet-Switching Hybrid Switch Structure |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN1717113A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008077322A1 (en) * | 2006-12-26 | 2008-07-03 | Huawei Technologies Co., Ltd. | A network node and a buffer device and a scheduling method |
CN1996799B (en) * | 2006-11-30 | 2010-04-14 | 上海交通大学 | Optical Packet Access Method Using Photoelectric Hybrid Buffer Structure |
CN101924961A (en) * | 2010-07-19 | 2010-12-22 | 浙江工业大学 | Multi-granularity optical cross-connect device for core node of optical burst switching network |
CN101370312B (en) * | 2007-08-17 | 2010-12-29 | 电子科技大学 | Construction method for hybrid switching optical network based on loop circuit |
CN101426152B (en) * | 2008-12-08 | 2011-03-16 | 电子科技大学 | Data transmission method for integrated mixed optical network based on burst exchange |
CN101621719B (en) * | 2009-07-29 | 2012-04-25 | 电子科技大学 | Switch processor matched with core node of hybrid optical switching network |
CN101656900B (en) * | 2009-09-11 | 2012-09-26 | 重庆邮电大学 | OPS node optical buffering device and method based on load selection |
CN101778047B (en) * | 2010-02-09 | 2013-01-23 | 重庆邮电大学 | Method for resolving optical switching grouping competition and photoelectric hybrid cache device |
CN102118305B (en) * | 2010-01-06 | 2014-04-30 | 上海贝尔股份有限公司 | Service board of communication equipment and communication equipment employing same |
WO2015109606A1 (en) * | 2014-01-27 | 2015-07-30 | 华为技术有限公司 | Data exchange device and system |
CN105515993A (en) * | 2015-11-26 | 2016-04-20 | 西安空间无线电技术研究所 | Photoelectric hybrid exchange internal path mapping method |
CN103873833B (en) * | 2012-12-10 | 2017-06-20 | 常州华龙通信科技有限公司 | Digital video light exchanges transmission platform |
CN107333190A (en) * | 2017-06-01 | 2017-11-07 | 清华大学 | A kind of photoelectricity hybrid switching structure of variable granularity |
-
2004
- 2004-07-02 CN CNA2004100401111A patent/CN1717113A/en active Pending
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1996799B (en) * | 2006-11-30 | 2010-04-14 | 上海交通大学 | Optical Packet Access Method Using Photoelectric Hybrid Buffer Structure |
US8249449B2 (en) | 2006-12-26 | 2012-08-21 | Huawei Technologies Co., Ltd. | Network node, buffer device, and scheduling method |
CN101212818B (en) * | 2006-12-26 | 2010-12-08 | 华为技术有限公司 | Network node, cache device and scheduling method |
WO2008077322A1 (en) * | 2006-12-26 | 2008-07-03 | Huawei Technologies Co., Ltd. | A network node and a buffer device and a scheduling method |
CN101370312B (en) * | 2007-08-17 | 2010-12-29 | 电子科技大学 | Construction method for hybrid switching optical network based on loop circuit |
CN101426152B (en) * | 2008-12-08 | 2011-03-16 | 电子科技大学 | Data transmission method for integrated mixed optical network based on burst exchange |
CN101621719B (en) * | 2009-07-29 | 2012-04-25 | 电子科技大学 | Switch processor matched with core node of hybrid optical switching network |
CN101656900B (en) * | 2009-09-11 | 2012-09-26 | 重庆邮电大学 | OPS node optical buffering device and method based on load selection |
CN102118305B (en) * | 2010-01-06 | 2014-04-30 | 上海贝尔股份有限公司 | Service board of communication equipment and communication equipment employing same |
CN101778047B (en) * | 2010-02-09 | 2013-01-23 | 重庆邮电大学 | Method for resolving optical switching grouping competition and photoelectric hybrid cache device |
CN101924961A (en) * | 2010-07-19 | 2010-12-22 | 浙江工业大学 | Multi-granularity optical cross-connect device for core node of optical burst switching network |
CN101924961B (en) * | 2010-07-19 | 2013-04-17 | 浙江工业大学 | Multi-granular optical cross connection device for core nodes in optical burst switching network |
CN103873833B (en) * | 2012-12-10 | 2017-06-20 | 常州华龙通信科技有限公司 | Digital video light exchanges transmission platform |
WO2015109606A1 (en) * | 2014-01-27 | 2015-07-30 | 华为技术有限公司 | Data exchange device and system |
US9806909B2 (en) | 2014-01-27 | 2017-10-31 | Huawei Technologies Co., Ltd. | Data switching apparatus and system |
CN105515993A (en) * | 2015-11-26 | 2016-04-20 | 西安空间无线电技术研究所 | Photoelectric hybrid exchange internal path mapping method |
CN105515993B (en) * | 2015-11-26 | 2019-02-15 | 西安空间无线电技术研究所 | An internal path mapping method for optoelectronic hybrid switching |
CN107333190A (en) * | 2017-06-01 | 2017-11-07 | 清华大学 | A kind of photoelectricity hybrid switching structure of variable granularity |
CN107333190B (en) * | 2017-06-01 | 2020-04-28 | 清华大学 | An optoelectronic hybrid exchange structure with variable particle size |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Kim et al. | Contention resolution for optical burst switching networks using alternative routing | |
JP2009094962A (en) | Passive optical network system and station side optical transmission line termination device | |
EP2083528B1 (en) | A network node and a buffer device and a scheduling method | |
CN1717113A (en) | A Optical-Burst/Packet-Switching Hybrid Switch Structure | |
Iizuka et al. | A scheduling algorithm minimizing voids generated by arriving bursts in optical burst switched WDM network | |
CN1520196A (en) | Large Capacity Optical Router Using Electrical Buffer | |
Takahashi et al. | Recent progress on the hybrid optoelectronic router | |
CN101778047B (en) | Method for resolving optical switching grouping competition and photoelectric hybrid cache device | |
CN101052000A (en) | Method for realizing separate service in light burst exchange network | |
Xue et al. | High-capacity multiservice optical label switching for the next-generation Internet | |
US7499650B2 (en) | Optical burst switching node with internal speedup | |
KR100715520B1 (en) | Optical data transfer device in optical switching system and method | |
CN1275413C (en) | Burst counting and multiplexing method for avoiding optical signal conflict in optical network | |
CN102427567B (en) | Asynchronous multi-wavelength mesh network adaptive node system based on optical packet switching | |
JP4926193B2 (en) | Passive optical network system and station side optical transmission line termination device | |
CN202385242U (en) | Asynchronous multi-wavelength mesh network self-adaptive node device based on optical packet switching | |
CN1625150A (en) | A method of supporting real-time services in optical burst switching | |
CN1320789C (en) | Method for resoving light burst resource conflict | |
Kazovsky et al. | UltraFlow access networks: A dual-mode solution for the access bottleneck | |
CN100349470C (en) | Elastic optical group exchanging method and node equipment based on dicyclic all-optical buffer storage | |
Samoud et al. | Performance improvements of hybrid opto-electronic packet switch using SDM in addition to WDM | |
Nakahara et al. | Demonstration of hybrid optoelectronic buffer for 10-Gbit/s arbitrary-length asynchronous optical packets | |
Orphanoudakis et al. | A hybrid optical switch architecture with shared electronic buffers | |
Sowailem et al. | Contention resolution strategy in optical burst switched datacenters | |
CN1372400A (en) | Integrated intelligent photoelectric mixed router |
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 |