《電子技術(shù)應(yīng)用》
您所在的位置:首頁(yè) > 模拟设计 > 设计应用 > 基于Crossbar多端口缓存共享的交换架构设计
基于Crossbar多端口缓存共享的交换架构设计
电子技术应用
张子琪,方震,于宗光
中国电子科技集团公司第五十八研究所
摘要: 随着网络的不断发展,数据带宽不断提升,对网络设备中交换容量、转发延迟、内存利用率等关键指标提出了更高的要求。采用基于Crossbar多端口缓存共享的交换架构,实现报文的存储、调度、排序和转发。采用共享缓存架构并使用描述符定义数据包,实现了多端口下多个队列的共享队列缓存,每端口支持 8 个优先级队列,实现有序的调度,减少多队列的阻塞问题。在多SRAM的缓存管理中采用分页式链表的方法,实现了动态共享缓存,提高数据存取效率、降低转发延迟。基于Xilinx公司的XCZU3EG验证平台上完成了FPGA原型验证,完成了多端口结构的功能验证和性能测试,平均端口速率达6.08 Gb/s,实现交换延迟可控。
關(guān)鍵詞: 缓存共享 多端口 交换架构 QoS
中圖分類號(hào):TN915.05 文獻(xiàn)標(biāo)志碼:A DOI: 10.16157/j.issn.0258-7998.256651
中文引用格式: 張子琪,方震,于宗光. 基于Crossbar多端口緩存共享的交換架構(gòu)設(shè)計(jì)[J]. 電子技術(shù)應(yīng)用,2025,51(11):53-58.
英文引用格式: Zhang Ziqi,F(xiàn)ang Zhen,Yu Zongguang. Design of switching architecture with shared multi-port buffers based on crossbar[J]. Application of Electronic Technique,2025,51(11):53-58.
Design of switching architecture with shared multi-port buffers based on crossbar
Zhang Ziqi,Fang Zhen,Yu Zongguang
The 58th Research Institute of China Electronics Technology Group Corporation
Abstract: With the continuous development of the network and the continuous improvement of data bandwidth, higher requirements have been put forward for key indicators such as switching capacity, forwarding delay, and memory utilization in network devices. This design adopts a switching architecture based on crossbar multi-port cache sharing to achieve the storage, scheduling, sorting, and forwarding of messages. This design uses a shared cache architecture and defines data packets with descriptors, achieving shared queue cache for multiple queues under multiple ports. Each port supports 8 priority queues, realizing orderly scheduling and reducing the blocking problem of multiple queues. In the cache management of multiple SRAMs, a paged linked list method is adopted to achieve dynamic shared cache, improving data access efficiency and reducing forwarding delay. This design has completed FPGA prototype verification on the XCZU3EG verification platform of Xilinx Company, achieving functional verification and performance testing of the multi-port structure. The average port rate reaches 6.08 Gb/s, and the switching delay is controllable.
Key words : shared cache;multi-port;switching architecture;QoS

引言

交換結(jié)構(gòu)對(duì)交換機(jī)的性能有著巨大影響,交換機(jī)需要具備快速的數(shù)據(jù)處理和轉(zhuǎn)發(fā)能力[1],這對(duì)交換架構(gòu)中的緩存管理技術(shù)提出了更高的要求,需要其能夠高效地存儲(chǔ)和讀取數(shù)據(jù)包,減少數(shù)據(jù)的等待時(shí)間和處理延遲,因此,在交換中設(shè)計(jì)合理的緩存結(jié)構(gòu)是至關(guān)重要的[2]。

本文介紹基于Crossbar多端口緩存共享的交換架構(gòu),該架構(gòu)具有64 bit位寬16入16出的端口,支持?jǐn)?shù)據(jù)包長(zhǎng)度為64~1 024 B,數(shù)據(jù)包存儲(chǔ)采用32塊256 Kb的SRAM存儲(chǔ)器,將SRAM緩存設(shè)計(jì)為共享式緩存確保在交換中實(shí)現(xiàn)高速數(shù)據(jù)存儲(chǔ),并且SRAM控制器能夠通過(guò)分頁(yè)式鏈表建立和鏈接對(duì)緩存內(nèi)部的數(shù)據(jù)包進(jìn)行管理,避免數(shù)據(jù)包的長(zhǎng)度和存儲(chǔ)器數(shù)據(jù)通道的數(shù)量對(duì)存儲(chǔ)器資源的影響,實(shí)現(xiàn)內(nèi)存回收,動(dòng)態(tài)調(diào)整空間,在保證高速傳輸?shù)那疤嵯?,很大程度地?jié)約了存儲(chǔ)資源[3]。數(shù)據(jù)包交換采用描述符描述數(shù)據(jù)包,減少數(shù)據(jù)包多次存儲(chǔ)產(chǎn)生的時(shí)間,極大地降低了交換的延遲,解決了交換過(guò)程中緩存效率低的問(wèn)題[4]。為了保證交換的服務(wù)質(zhì)量,該架構(gòu)支持按 QoS調(diào)度,可實(shí)時(shí)切換嚴(yán)格優(yōu)先級(jí)、加權(quán)差額輪詢調(diào)度算法,支持多端口多隊(duì)列動(dòng)態(tài)共享緩存,解決了傳統(tǒng)Crossbar交換架構(gòu)中存在的隊(duì)頭阻塞等問(wèn)題[5]。經(jīng)驗(yàn)證該架構(gòu)具有存儲(chǔ)、排序、轉(zhuǎn)發(fā)功能,并能高效存儲(chǔ)數(shù)據(jù)包,實(shí)現(xiàn)高速、多端口、無(wú)沖突的交換。


本文詳細(xì)內(nèi)容請(qǐng)下載:

http://ihrv.cn/resource/share/2000006840


作者信息:

張子琪,方震,于宗光

(中國(guó)電子科技集團(tuán)公司第五十八研究所,江蘇 無(wú)錫 214026)


官方訂閱.jpg

此內(nèi)容為AET網(wǎng)站原創(chuàng),未經(jīng)授權(quán)禁止轉(zhuǎn)載。