Similar articles

Abstract: To address the issues of head-of-line (HOL) blocking at the virtual output queue (VOQ) level, packet loss, and congestion spreading caused by buffer overflow in the shared-buffer-based combined input and output queued (CIOQ) switching architecture, while enhancing its performance and stability, we propose a de-blocking adaptive feedback control (AFC) design in this study. The introduction of the credit timeout detection mechanism (CTDM) enables the CIOQ to achieve theoretical 100% non-blocking state, effectively eliminating the impact of HOL blocking. With the combined effect of the proposed VOQ dynamic regulation algorithm (VDRA) and threshold dynamic adaptive algorithm (TDAA), it can reduce the risk of congestion spreading caused by buffer overflow and consequently improve the overall performance of the system. Both theoretical analysis and experimental results demonstrate that, under typical traffic conditions, the proposed design achieves a maximum throughput of 1499.66 Gb/s and a minimum latency of 83 ns. Additionally, the effective throughput ratio reaches 96.94%, with a data link layer packet (DLLP) loss ratio of merely 0.61% and a packet loss rate as low as 0.6%. In comparison with traditional CIOQ and input queued (IQ) switch architectures, the proposed design demonstrates improvements in throughput by 15.12% and 20.55%, and forwarding latency is reduced by 26.9% and 54.7%, respectively, and the system stability is stronger, which can fully satisfy the demand for data exchange in complex situations.

面向共享缓存CIOQ交换架构的去阻塞自适应反馈调节设计

[1]Bandara S, Sanaullah A, Tahir Z, et al., 2024. Performance evaluation of VirtIO device drivers for host-FPGA PCIe communication. IEEE Int Parallel and Distributed Processing Symp Workshops, p.169-176.

[2]Chen BW, Zhou WB, 2025. A review of input-queued scheduling algorithms in high-speed switching systems. Microelectr Comput, 42(5):1-8 (in Chinese).

[3]Chen C, Li H, Zhou L, 2024. A design of radar high-speed data storage module based on VPX. 4^th Int Conf on Computer Science Electronic Information Engineering and Intelligent Control Technology, p.234-239.

[4]Dagli I, Belviranli ME, 2024. Shared memory-contention-aware concurrent DNN execution for diversely heterogeneous system-on-chips. Proc 29^th ACM SIGPLAN Annual Symp on Principles and Practice of Parallel Programming, p.243-256.

[5]de la Rosa MS, Gomez-Lopez G, Andújar FJ, et al., 2025. Quality-of-service provision for BXIv3-based interconnection networks. J Supercomput, 81(4):601.

[6]Dong CL, Shen JL, Li PJ, et al., 2024. Multi-protocol switching circuit for software defined interconnection system. J Commun, 45(5):44-53 (in Chinese).

[7]Firoozshahian A, Manshadi V, Goel A, et al., 2007. Efficient, fully local algorithms for CIOQ switches. 26^th IEEE Int Conf on Computer Communications, p.2491-2495.

[8]Hou WT, Zhang J, Wang ZK, et al., 2024. Understanding routable PCIe performance for composable infrastructures. 21^st USENIX Symp on Networked Systems Design and Implementation, p.297-312.

[9]Hu B, Fan FJ, Yeung KL, et al., 2018. Highest rank first: a new class of single-iteration scheduling algorithms for input-queued switches. IEEE Access, 6:11046-11062.

[10]Hu JB, Huang JW, Wang JX, et al., 2023. A transmission control mechanism for lossless datacenter network based on direct congestion notification. Acta Electr Sin, 51(9):2355-2365 (in Chinese).

[11]Kaltenhauser A, Stefanidi E, Schöning J, 2024. Playing with perspectives and unveiling the autoethnographic kaleidoscope in HCI – a literature review of autoethnographies. Proc CHI Conf on Human Factors in Computing Systems, Article 819.

[12]Kim H, Ryu J, Lee J, 2024. TCCL: discovering better communication paths for PCIe GPU clusters. Proc 29^th ACM Int Conf on Architectural Support for Programming Languages and Operating Systems, p.999-1015.

[13]Luo JF, Yu F, Li WJ, et al., 2024. A novel switch architecture for multi-die optimization with efficient connections. Electronics, 13(16):3205.

[14]Mohtavipour SM, Mollajafari M, Naseri A, 2020. A novel packet exchanging strategy for preventing HoL-blocking in fat-trees. Clust Comput, 23(2):461-482.

[15]Nag SN, 2023. Technical analysis of PCIe to PCIe 6: a next-generation interface evolution. World J Eng Technol, 11(3):504-525.

[16]Ouyang YM, Yang JF, Xing K, et al., 2018. An improved communication scheme for non-HOL-blocking wireless NoC. Integration, 60:240-247.

[17]Palnitkar SS, Kanade S, 2024. Q-memory task routing to prevent deadlocks in Ethernet control with memory crossbar switching. Opt Mem Neur Netw, 33(1):72-85.

[18]Ran C, Su H, Sun Y, et al., 2023. Research on fair scheduling algorithm of high-performance input queuing switch. Foreign Electr Meas Technol, 42(2):114-119 (in Chinese).

[19]Ran C, Su H, Sun Y, et al., 2024. Research on improved iSLIP scheduling algorithm based on CIOQ architecture. Comput Simul, 41(4):325-329, 417 (in Chinese).

[20]Shen ZJ, Gao J, Wu RG, 2018. Feedback and reverse transmission mechanism based two-stage switch architecture. J Electr Inform Technol, 40(3):697-704 (in Chinese).

[21]Shen ZJ, Tao DH, Gao J, 2019. Multichannel-feedback-based two-stage switch architecture. J Front Comput Sci Technol, 13(9):1516-1523 (in Chinese).

[22]Wu WX, Zhang T, Li Z, et al., 2025. Dynamic per-flow queues in shared buffer TSN switches. ACM Trans Des Autom Electr Syst, 30(3):38.

[23]Xu W, Dainoff MJ, Ge LZ, et al., 2023. Transitioning to human interaction with AI systems: new challenges and opportunities for HCI professionals to enable human-centered AI. Int J Hum–Comput Int, 39(3):494-518.

[24]Yébenes P, Escudero-Sahuquillo J, García PJ, et al., 2019. Head-of-line blocking avoidance in slim fly networks using deadlock-free non-minimal and adaptive routing. Concurr Comput, 31(2):e4441.

[25]Zhang YR, Wang SG, Ren FY, 2025. A review of lossless network traffic management. J Comput Res Dev, 62(5):1290-1306 (in Chinese).

[26]Zheng CM, Yao XX, Zhou F, et al., 2022. Adaption and implementation of server chipsets for the Loongson CPU. Chin J Eng, 44(7):1244-1254 (in Chinese). https:/doi.org/10.13374/j.issn2095-9389.2021.10.08.003

[27]Zheng R, Shen JL, Lv P, et al., 2026. Optimized design of non-transparent bridge for heterogeneous interconnects in hyper-converged infrastructure. J Electr Inform Technol, 48(2):567-582 (in Chinese).

[28]Zhou WW, Sheng WX, Yan BY, 2024. A single-chip wafer-level packaged SR-crossbar RF MEMS switch matrix. IEEE Electr Dev Lett, 45(7):1309-1312.

[29]Zyla K, Liess M, Wild T, et al., 2024. FlexCross: high-speed and flexible packet processing via a crosspoint-queued crossbar. 27^th Euromicro Conf on Digital System Design, p.98-105.