19th USENIX Symposium on Networked Systems Design and Implementation (NSDI '22)
2022-04-04..2022-04-06
Renton, United States
| title | author | length |
|---|---|---|
| HeteroSketch: Coordinating Network-wide Monitoring in Heterogeneous and Dynamic Networks | Anup Agarwal, Zaoxing Liu, Srinivasan Seshan | 16m |
Framework places sketch-based telemetry across switches, SmartNICs, FPGAs and software switches by profiling devices and solving a placement and allocation… | ||
| Runtime Programmable Switches | Jiarong Xing, Kuo-Feng Hsu, Matty Kadosh, Alan Lo, Yonatan Piasetzky, Arvind Krishnamurthy, Ang Chen | 16m |
Prototype called FlexCore performs partial reconfiguration primitives on P4 pipelines by atomically switching PD-table pointers and installing flex edges… | ||
| Efficient Scheduling Policies for Microsecond-Scale Tasks | Sarah McClure, Amy Ousterhout, Scott Shenker, Sylvia Ratnasamy | 15m |
Simulations model 1μs tasks with measured overheads (100 ns to move a task, 5 μs to add a core) and show work stealing consistently yields the best… | ||
| Decentralized cloud wide-area network traffic engineering with BLASTSHIELD | Umesh Krishnaswamy, Rachee Singh, Nikolaj Bjørner, Himanshu Raj | 19m |
BLASTSHIELD decentralizes Microsoft’s WAN by slicing the network and running independent slice controllers that program only local path segments (via MPLS) and… | ||
| Cetus: Releasing P4 Programmers from the Chore of Trial and Error Compiling | Yifan Li, Jiaqi Gao, Ennan Zhai, Mengqi Liu, Kun Liu, Hongqiang Harry Liu | 16m |
NSDI '22 - Cetus: Releasing P4 Programmers from the Chore of Trial and Error Compiling
Yifan Li, Tsinghua University and Alibaba Group; Jiaqi Gao, Ennan Zhai… | ||
| SLearn - A Case for Task Sampling based Learning for Cluster Job Scheduling | Akshay Jajoo, Y. Charlie Hu, Xiaojun Lin, Nan Deng | 16m |
SLearn proposes learning in space by running a small, high-priority sample of tasks per job to estimate average task duration and total job size, then… | ||
| Automated Verification of Network Function Binaries | Solal Pirelli, Akvilė Valentukonytė, Katerina Argyraki, George Candea | 15m |
Claims an automated verifier, Klint, proves properties of real NF binaries (no source or debug info) by modeling arbitrary data structures as a universal ghost… | ||
| Performance Interfaces for Network Functions | Rishabh Iyer, Katerina Argyraki, George Candea | 17m |
Proposes concrete “performance interfaces” for network functions and implements them with PIX, a tool that extracts programmatic latency expressions and free… | ||
| Graham: Synchronizing Clocks by Leveraging Local Clock Properties | Ali Najafi, Michael Wei | 17m |
Graham is a system that characterizes each server's quartz oscillator as a temperature-to-frequency curve and uses commodity temperature sensors plus… | ||
| Starlight: Fast Container Provisioning on the Edge and over the WAN | Jun Lin Chen, Daniyal Liaqat, Moshe Gabel, Eyal de Lara | 19m |
Reports a push-based protocol that sends a per-file delta bundle computed on demand from a cloud proxy, replaces overlayfs with Starlight FS, and stores… | ||
| Privid: Practical, Privacy-Preserving Video Analytics Queries | Frank Cangialosi, Neil Agarwal, Venkat Arun, Junchen Jiang, Srinivas Narayana, Anand Sarwate, Ravi Netravali | 17m |
Privid defines ρ,K,ε-event-duration privacy and enforces it by splitting video into chunks, running analyst-provided black-box neural nets per chunk, then… | ||
| Swift: Adaptive Video Streaming with Layered Neural Codecs | Mallesham Dasari, Kumara Kahatapitiya, Samir R. Das, Aruna Balasubramanian, Dimitris Samaras | 16m |
Swift implements a layered neural codec that encodes frames as a base code plus residual layers, a single-shot decoder that fuses any subset of layers in one… | ||
| Detecting Ephemeral Optical Events with OpTel | Congcong Miao, Minggang Chen, Arpit Gupta, Zili Meng, Lianjin Ye, Jingyu Xiao, Jie Chen, Zekun He, Xulong Luo, Jilong Wang, Heng Yu | 16m |
A centralized vendor-agnostic controller standardizes optical device models, deploys push-based telemetry agents, and uses a cloud-backed collector cluster to… | ||
| DispersedLedger: High-Throughput Byzantine Consensus on Variable Bandwidth Networks | Lei Yang, Seo Jin Park, Mohammad Alizadeh, Sreeram Kannan, David Tse | 16m |
Uses erasure coding and a new verifiable information dispersal to separate data availability from block delivery so nodes agree on block hashes without… | ||
| IA-CCF: Individual Accountability for Permissioned Ledgers | Alex Shamis, Peter Pietzuch, Burcu Canakci, Miguel Castro, Cédric Fournet, Edward Ashton, Amaury Chamayou, Sylvan Clebsch, Antoine Delignat-Lavaud, Matthew Kerner, Julien Maffre, Olga Vrousgou, Christoph M. Wintersteiger, Manuel Costa, Mark Russinovich | 14m |
A permissioned ledger implementation called IA-CCF signs and logs BFT protocol messages into the ledger and issues Merkle-tree receipts that clients can audit… | ||
| Donar: Anonymous VoIP over Tor | Yérom-David Bromberg, Quentin Dufour, Davide Frey, Etienne Rivière | 15m |
DONAR is a user-side proxy that spreads traffic over several links by opening multiple Tor circuits, monitors per-link differential one-way delay, and… | ||
| YuZu: Neural-Enhanced Volumetric Video Streaming | Anlan Zhang, Chendong Wang, Bo Han, Feng Qian | 17m |
System applies optimized 3D super-resolution to point-cloud volumetric video, achieving 542× speedup on SR inference through patch-generation trimming, model… | ||
| Spectrum: High-bandwidth Anonymous Broadcast | Zachary Newman, Sacha Servan-Schreiber, Srinivas Devadas | 16m |
Spectrum is a deployed system that uses two or more non‑colluding servers, secret sharing, Carter‑Wegman MACs and distributed point functions to let few… | ||
| Ekya: Continuous Learning of Video Analytics Models on Edge Compute Servers | Romil Bhardwaj, Zhengxu Xia, Ganesh Ananthanarayanan, Junchen Jiang, Yuanchao Shu, Nikolaos Karianakis, Kevin Hsieh, Paramvir Bahl, Ion Stoica | 17m |
Edge servers run compressed video models that drift with live data; Ekya continuously retrains selected models and schedules GPU time to maximize mean… | ||
| Bluebird: High-performance SDN for Bare-metal Cloud Services | Manikandan Arumugam, Deepak Bansal, Navdeep Bhatia, James Boerner, Simon Capper, Changhoon Kim, Sarah McClure, Neeraj Motwani, Ranga Narasimhan, Urvish Panchal, Tommaso Pimpo, Ariff Premji, Pranjal Shrivastava, Rishabh Tewari | 19m |
Bluebird implements network virtualization on Azure using P4 programmable ASICs and a control plane that pushes JSON goal states to a zonal multitenant… | ||
| Differential Network Analysis | Peng Zhang, Aaron Gember-Jacobson, Yueshang Zuo, Yuhao Huang, Xu Liu, Hao Li | 17m |
Framework called differential network analysis incrementally processes differences in configs, routes, and links to output changed end-to-end behaviors instead… | ||
| Isolation Mechanisms for High-Speed Packet-Processing Pipelines | Tao Wang, Xiangrui Yang, Gianni Antichi, Anirudh Sivaraman, Aurojit Panda | 15m |
Menshen enforces isolation in RMT-style pipelines by attaching a per-tenant VLAN PID to packets, using per-program configuration overlays for parsers and key… | ||
| RDMA is Turing complete, we just did not know it yet! | Waleed Reda, Marco Canini, Dejan Kostić, Simon Peter | 18m |
Claims RDMA is Turing complete by running self-modifying RDMA chains on commodity NICs: pre-posted RDMA_RECEIVE/READ/WRITE operations get mutated at runtime… | ||
| Katra: Realtime Verification for Multilayer Networks | Ryan Beckett, Aarti Gupta | 16m |
Katra implements a new incremental verifier that models packets as a single header copy and uses partial equivalence classes plus a lazy repair of header… | ||
| Justitia: Software Multi-Tenancy in Hardware Kernel-Bypass Networks | Yiwen Zhang, Yue Tan, Brent E. Stephens, Mosharaf Chowdhury | 15m |
Justitia inserts a software layer before hardware kernel-bypass NICs, splitting large sends into chunks and issuing them under a token bucket so the NIC’s… | ||
| NetVRM: Virtual Register Memory for Programmable Networks | Hang Zhu, Tao Wang, Yi Hong, Dan R. K. Ports, Anirudh Sivaraman, Xin Jin | 16m |
NetVRM provides a virtual register memory abstraction plus a P4VRM compiler that merges multiple P4 programs so on-chip register regions are remapped at… | ||
| Check-N-Run: a Checkpointing System for Training Deep Learning Recommendation Models | Assaf Eisenman, Kiran Kumar Matam, Steven Ingram, Dheevatsa Mudigere, Raghuraman Krishnamoorthi, Krishnakumar Nair, Misha Smelyanskiy, Murali Annavaram | 16m |
System implements a decoupled, atomic checkpoint pipeline that copies GPU state to CPU and uploads with dedicated CPU workers, targets terabyte-scale… | ||
| MLaaS in the Wild: Workload Analysis and Scheduling in Large-Scale Heterogeneous GPU... | Qizhen Weng, Wencong Xiao, Yinghao Yu, Wei Wang, Cheng Wang, Jian He, Yong Li, Liping Zhang, Wei Lin, Yu Ding | 15m |
A two-month Alibaba MLaaS trace covers over 6,700 GPUs across ~1,800 nodes with 1.2M tasks, 7.5M instances and shows concrete pain points: 85% gang-scheduled… | ||
| Accelerating Collective Communication in Data Parallel Training across Deep Learning... | Joshua Romero, Junqi Yin, Nouamane Laanait, Bing Xie, M. Todd Young, Sean Treichler, Vitalii Starchenko, Albina Borisevich, Alex Sergeev, Michael Matheson | 17m |
Implements changes to Horovod: response caching that runs the control plane once and uses a global bitvector intersection for readiness, plus explicit… | ||
| Learning to Communicate Effectively Between Battery-free Devices | Kai Geissdoerfer, Marco Zimmerling | 13m |
Bonito builds pairwise connections between intermittently powered devices by having each node learn its charging-time distribution with online SGD and exchange… | ||
| Evolvable Network Telemetry at Facebook | Yang Zhou, Ying Zhang, Minlan Yu, Guangyu Wang, Dexter Cao, Eric Sung, Starsky Wong | 16m |
Production telemetry system PCAT introduces a uniform change cube representation, an intent-based layering that splits intent, data, collection and job models… | ||
| Optimizing Network Provisioning through Cooperation | Harsha Sharma, Parth Thakkar, Sagar Bharadwaj, Ranjita Bhagwan, Venkata N. Padmanabhan, Yogesh Bansal, Vijay Kumar, Kathleen Voelbel | 16m |
Cloud inter-DC provisioning uses application-declared deadlines and link failure time-series to cut capacity needs by about 30–45% on real substrate data. A… | ||
| Cocktail: A Multidimensional Optimization for Model Serving in Cloud | Jashwant Raj Gunasekaran, Cyan Subhra Mishra, Prashanth Thinakaran, Bikash Sharma, Mahmut Taylan Kandemir, Chita R. Das | 14m |
Cocktail is an ensemble-based model-serving framework that dynamically selects a small set of models per request using a class-wise accuracy dictionary and… | ||
| Orca: Server-assisted Multicast for Datacenter Networks | Khaled Diab, Parham Yassini, Mohamed Hefeeda | 18m |
Orca claims a practical multicast primitive for datacenters that splits forwarding work between switches and servers, implements a controller plus per-rack… | ||
| How to diagnose nanosecond network latencies in rich end-host stacks | Roni Haecki, Radhika Niranjan Mysore, Lalith Suresh, Gerd Zellweger, Bo Gan, Timothy Merrifield, Sujata Banerjee, Timothy Roscoe | 17m |
Tool called NSight reconstructs per-message lifetimes across kernel, NIC and user space by reconciling CPU and NIC traces into a single timeline with… | ||
| SCALE: Automatically Finding RFC Compliance Bugs in DNS Nameservers | Siva Kesava Reddy Kakarla, Ryan Beckett, Todd Millstein, George Varghese | 16m |
SCALE uses an executable RFC model to jointly generate zone files and queries and Ferret runs those tests against real servers. The pipeline uses the Zen… | ||
| SketchLib: Enabling Efficient Sketch-based Monitoring on Programmable Switches | Hun Namkung, Zaoxing Liu, Daehyeok Kim, Vyas Sekar, Peter Steenkiste | 15m |
SketchLib is a P4-targeted library that implements six correct-by-construction optimizations after identifying four hardware bottlenecks (hash units, SLUs… | ||
| Collie: Finding Performance Anomalies in RDMA Subsystems | Xinhao Kong, Yibo Zhu, Huaping Zhou, Zhuo Jiang, Jianxi Ye, Chuanxiong Guo, Danyang Zhuo | 15m |
Collie systematically searches RDMA workload space using simulated annealing guided by hardware performance and diagnostic counters to drive tests toward… | ||
| CloudCluster: Unearthing the Functional Structure of a Cloud Service | Weiwu Pang, Sourav Panda, Jehangir Amjad, Christophe Diot, Ramesh Govindan | 15m |
CloudCluster extracts a service’s functional groups by clustering the VM-to-VM traffic matrix, using feature scaling, dimensionality reduction (via singular… | ||
| Whisper: IoT in the TV White Space Spectrum | Tusher Chakraborty, Heping Shi, Zerina Kapetanovic, Bodhi Priyantha, Deepak Vasisht, Andrew Nelson, Parag Pandit, Prasad Pillai, Yaswant Chabria, Binh Vu, Ranveer Chandra, Michael Daum | 16m |
A complete IoT stack is built: a custom 150–960 MHz radio that avoids adjacent-channel leakage, a gateway with multiple base-station radios, and an FDMA-style… | ||
| NetHint: White-Box Networking for Multi-Tenant Data Centers | Jingrong Chen, Hong Zhang, Wei Zhang, Liang Luo, Jeffrey Chase, Ion Stoica, Danyang Zhuo | 17m |
NetHint exposes three lightweight hints — a hierarchical virtual topology, per-link residual bandwidth and number of competing flows, and a 100 ms update… | ||
| Dynamic Scheduling of Approximate Telemetry Queries | Chris Misa, Walt O'Connor, Ramakrishnan Durairajan, Reza Rejaie, Walter Willinger | 17m |
System reframes telemetry as a runtime resource scheduler and implements a time-division approximation with cluster sampling to estimate error from only… | ||
| Characterizing Physical-Layer Transmission Errors in Cable Broadband Networks | Jiyao Hu, Zhenyu Zhou, Xiaowei Yang | 17m |
Uses DOCSIS 3.0 codeword counters pulled from CMTS and cable modems across 77K+ devices on 394 HFC segments over 16 months, sampled roughly every 4 hours, to… | ||
| OrbWeaver: Using IDLE Cycles in Programmable Networks for Opportunistic Coordination | Liangcheng Yu, John Sonchack, Vincent Liu | 14m |
OrbWeaver claims to harvest sub-microsecond gaps on 100 Gbps links (example inter-packet gap ≈120 ns) to inject idle packets that carry coordination data while… | ||
| PowerTCP: Pushing the Performance Limits of Datacenter Networks | Vamsi Addanki, Oliver Michel, Stefan Schmid | 16m |
Introduces PowerTCP, a congestion control that drives send-window updates from the bandwidth–window product (called power) and measures it via in-band network… | ||
| Saiyan: Design and Implementation of a Low-power Demodulator for LoRa Backscatter Systems | Xiuzhen Guo, Longfei Shangguan, Yuan He, Nan Jing, Jiacheng Zhang, Haotian Jiang, Yunhao Liu | 15m |
Demodulator design claims 3.5–5x demodulation-range gain and a prototype two‑layer PCB that reaches up to 180 m while consuming about 370 μW (ASIC sim ~93.2… | ||
| Buffer-based End-to-end Request Event Monitoring in the Cloud | Kaihui Gao, Chen Sun, Shuai Wang, Dan Li, Yu Zhou, Hongqiang Harry Liu, Lingjun Zhu, Ming Zhang | 16m |
BufScope models the request datapath as a buffer chain and injects per-request IDs plus a remaining-size field into packet headers so SmartNICs and… | ||
| Data-Parallel Actors: A Programming Model for Scalable Query Serving Systems | Peter Kraft, Fiodar Kazhamiaka, Peter Bailis, Peter Ballis, Matei Zaharia | 14m |
Data-parallel actors (DPA) is a programming model and runtime (Uniserv) that composes single-node components into distributed query-serving systems using… | ||
| Modular Switch Programming Under Resource Constraints | Mary Hogan, Shir Landau Feibish, Mina Tahmasbi Arashloo, Jennifer Rexford, David Walker | 14m |
P4All extends P4 with elastic data structures (symbolic sizes + objective functions) and compile-time loops and uses dependency analysis plus an ILP-based… | ||
| Enabling IoT Self-Localization Using Ambient 5G Signals | Suraj Jog, Junfeng Guan, Sohrab Madani, Ruochen Lu, Songbin Gong, Deepak Vasisht, Haitham Hassanieh | 19m |
Claim: ISLA lets battery‑constrained IoT nodes self‑localize by overhearing ambient 5G and without any base‑station changes. It uses a fabricated MEMS… | ||
| SwiSh: Distributed Shared State Abstractions for Programmable Switches | Lior Zeno, Dan R. K. Ports, Jacob Nelson, Daehyeok Kim, Shir Landau Feibish, Idit Keidar, Arik Rinberg, Alon Rashelbach, Igor De-Paula, Mark Silberstein | 17m |
SwiSh provides a data‑plane shared‑state layer that runs on P4/Tofino switches and implements three replication protocols (strong, eventual, and Strong… | ||
| An edge-queued datagram service for all datacenter traffic | Vladimir Olteanu, Haggai Eran, Dragos Dumitrescu, Adrian Popa, Cristi Baciu, Georgios Nikolaidis, Mark Silberstein, Mark Handley, Costin Raiciu | 16m |
EQDS claims to move queuing to senders and enforce delivery via receiver-driven grants, implemented with either a packet trimming backend (Intel Tofino… | ||
| Backdraft: a Lossless Virtual Switch that Prevents the Slow Receiver Problem | Alireza Sanaee, Farbod Shahinfar, Gianni Antichi, Brent E. Stephens | 17m |
Lossless virtual switching implemented on BESS combines Dynamic Per-Flow Queuing to avoid head‑of‑line blocking, Doorbell queues to cut CPU overhead, and an… | ||
| Yeti: Stateless and Generalized Multicast Forwarding | Khaled Diab, Mohamed Hefeeda | 16m |
Yeti implements generalized multicast by moving forwarding state into packets: a centralized controller generates controller-generated labels that encode… | ||
| PLatter: On the Feasibility of Building-scale Power Line Backscatter | Junbo Zhang, Elahe Soltanaghaei, Artur Balanuta, Reese Grimsley, Swarun Kumar, Anthony Rowe | 16m |
PLatter demonstrates that injecting a 13 MHz carrier into building power lines lets wall-mounted battery-assisted tags modulate parasitic impedance amplitude… | ||
| Gearbox: A Hierarchical Packet Scheduler for Approximate Weighted Fair Queuing | Peixuan Gao, Anthony Dalleggio, Yang Xu, H. Jonathan Chao | 16m |
Proposes a hierarchical scheduler that approximates WFQ with O(1) enqueue/dequeue operations and a provable normalized discrepancy ≤ 1, using a few logical… | ||
| IMap: Fast and Scalable In-Network Scanning with Programmable Switches | Guanyu Li, Menghao Zhang, Cheng Guo, Han Bao, Mingwei Xu, Hongxin Hu, Fenghua Li | 16m |
IMap turns P4-capable programmable switches into a high-speed in-network scanner that generates randomized probes from template packets, encodes per-probe… | ||
| SwarmMap: Scaling Up Real-time Collaborative Visual SLAM at the Edge | Jingao Xu, Hao Cao, Zheng Yang, Longfei Shangguan, Jialin Zhang, Xiaowu He, Yunhao Liu | 17m |
SwarmMap is an edge framework that logs and replays SLAM operations to avoid sending full maps, uses a change-log synchronization layer, a priority-aware STS… | ||
| In-Network Velocity Control of Industrial Robot Arms | Sándor Laki, Csaba Györgyi, József Pető, Péter Vörös, Géza Szabó | 14m |
Offloads real-time velocity control into a P4-enabled Tofino switch that stores trajectories as linked-list entries in match-action tables, receives UDP… | ||
| Aquila: A unified, low-latency fabric for datacenter networks | Dan Gibson, Hema Hariharan, Eric Lance, Moray McLaren, Behnam Montazeri, Arjun Singh, Stephen Wang, Hassan M. G. Wassel, Zhehua Wu, Sunghwan Yoo, Raghuraman Balasubramanian, Prashant Chandra, Michael Cutforth, Peter Cuy, David Decotigny, Rakesh Gautam, Alex Iriza, Milo M. K. Martin, Rick Roy, Zuowei Shen, Ming Tan, Ye Tang, Monica Wong-Chan, Joe Zbiciak, Amin Vahdat | 15m |
Aquila builds a unified datacenter fabric using a GNet cell-based L2 protocol, a custom Tin/TORonic ASIC and integrated switch, and Dragonfly topology with 40… | ||
| cISP: A Speed-of-Light Internet Service Provider | Debopam Bhattacherjee, Waqar Aqeel, Sangeetha Abdu Jyothi, Ilker Nadi Bozkurt, William Sentosa, Muhammad Tirmazi, Anthony Aguirre, Balakrishnan Chandrasekaran, P. Brighten Godfrey, Gregory Laughlin, Bruce M. Maggs, Ankit Singla | 17m |
Design augments fiber with free-space microwave hops chosen from FCC and American Tower databases and terrain and canopy maps, solving an ILP to pick ~3,000… | ||
| CurvingLoRa to Boost LoRa Network Throughput via Concurrent Transmission | Chenning Li, Xiuzhen Guo, Longfei Shangguan, Zhichao Cao, Kyle Jamieson | 14m |
Replaces LoRa’s linear chirp with non-linear chirps so misaligned concurrent transmissions scatter energy instead of producing a single overpowering peak… | ||
| Passive DSSS: Empowering the Downlink Communication for Backscatter Systems | Songfan Li, Hui Zheng, Chong Zhang, Yihang Song, Minghua Chen, Shen Yang, Li Lu, Mo Li | 15m |
Proposes shifting DSSS spreading-code generation and synchronization to the gateway and sending a baseband channel plus a synchronized spreading-code channel… | ||
| Elixir: A High-performance and Low-cost Approach to Managing Hardware/Software Hybrid | Yanshu Wang, Dan Li, Yuanwei Lu, Jianping Wu, Hua Shao, Yutian Wang | 15m |
Elixir: A High-performance and Low-cost Approach to Managing Hardware/Software Hybrid Flow Tables Considering Flow Burstiness
Yanshu Wang and Dan Li, Tsinghua… | ||
| Enabling In-situ Programmability in Network Data Plane: From Architecture to Language | Yong Feng, Zhikang Chen, Haoyu Song, Wenquan Xu, Jiahao Li, Zijian Zhang, Tong Yun, Ying Wan, Bin Liu | 15m |
A new switch architecture and language extension enable changing protocol parsing, table layouts, and per-packet processing at runtime without stopping… | ||
| Scaling Open vSwitch with a Computational Cache | Alon Rashelbach, Ori Rottenstreich, Mark Silberstein | 14m |
Open vSwitch packet classification uses neural inference via NuevoMatch-APP, a reworked NuevoMatch that speeds training by ~1,000× through bucketizing ranges… | ||
| Zeta: A Scalable and Robust East-West Communication Framework in Large-Scale Clouds | Qianyu Zhang, Gongming Zhao, Hongli Xu, Zhuolong Yu, Liguang Xie, Yangming Zhao, Chunming Qiao, Ying Xiong, Liusheng Huang | 15m |
Zeta builds a gateway cluster layer with XDP/eBPF-based gateways and a multi‑IP migration trick that reassigns virtual IPs instead of rewriting switch rules… | ||
| Closed-loop Network Performance Monitoring and Diagnosis with SpiderMon | Weitao Wang, Xinyu Crystal Wu, Praveen Tammana, Ang Chen, T. S. Eugene Ng | 16m |
SpiderMon detects performance problems by adding an accumulated queuing delay header on data‑plane packets and triggering when that sum exceeds a threshold… | ||
| Unlocking the Power of Inline Floating-Point Operations on Programmable Switches | Yifan Yuan, Omar Alama, Jiawei Fei, Jacob Nelson, Dan R. K. Ports, Amedeo Sapio, Marco Canini, Nam Sung Kim | 16m |
Proposes FPISA, a compact floating-point representation and pipeline methods, implements it on Intel Tofino then identifies throughput and accuracy limits, and… | ||
| RDC: Energy-Efficient Data Center Network Congestion Relief with Topological | Weitao Wang, Dingming Wu, Sushovan Das, Afsaneh Rahbar, Ang Chen, T. S. Eugene Ng | 14m |
RDC proposes inserting circuit switches at the edge between ToR switches and servers to regroup servers across racks by reconfiguring circuits based on… | ||
| Tiara: A Scalable and Efficient Hardware Acceleration Architecture for Stateful Layer-4 | Chaoliang Zeng, Layong Luo, Teng Zhang, Zilong Wang, Luyang Li, Wenchen Han, Nan Chen, Lebing Wan, Lichao Liu, Zhipeng Ding, Xiongfei Geng, Tao Feng, Feng Ning, Kai Chen, Chuanxiong Guo | 18m |
Tiara splits stateful L4 load balancing into memory-heavy real-server selection on FPGAs with HBM and throughput-heavy encapsulation on a programmable switch… | ||
| Exploiting Digital Micro-Mirror Devices for Ambient Light Communication | Talia Xu, Marco Zúñiga Zamalloa, Miguel Chávez Tapia, Marco Zúñiga | 17m |
Describes an optical model and a working prototype, the PhotoLink platform, that swaps LCD shutters for digital micro-mirror devices (DMDs) and an FPGA… | ||
| Re-architecting Traffic Analysis with Neural Network Interface Cards | Giuseppe Siracusano, Salvator Galea, Davide Sanvito, Mohammad Malekzadeh, Gianni Antichi, Paolo Costa, Hamed Haddadi, Roberto Bifulco | 14m |
The paper argues for replacing traditional supervised ML models for online network traffic analysis with binary neural networks, and introduces N3IC, a system… | ||
| C2DN: How to Harness Erasure Codes at the Edge for Efficient Content Delivery | Juncheng Yang, Anirudh Sabnis, Daniel S. Berger, K.V. Rashmi, Ramesh K. Sitaraman | 15m |
Analyzes month-long logs from over 2000 clusters and implements C2DN on Apache Traffic Server, replacing replication with erasure coding plus targeted parity… | ||
| Configanator: A Data-driven Approach to Improving CDN Performance. | Usama Naseer, Theophilus A. Benson | 18m |
Configanator is a CDN-side system that groups connections into network classes and learns per-class TCP/HTTP settings using a contextual multi-armed bandit… | ||
| Backpressure Flow Control | Prateesh Goyal, Preey Shah, Kevin Zhao, Georgios Nikolaidis, Mohammad Alizadeh, Thomas E. Anderson | 16m |
Claims a practical per‑hop flow control design that runs in hardware, implemented in P4 and implemented on Tofino2 and evaluated in packet‑level ns3… | ||
| FlexTOE: Flexible TCP Offload with Fine-Grained Parallelism | Rajath Shashidhara, Timothy Stamler, Antoine Kaufmann, Simon Peter | 14m |
FlexTOE is a software TCP offload engine that runs on SmartNICs (Agilio‑CX40, Mellanox BlueField and x86 prototypes) and eliminates host TCP data-path by… | ||
| Packet Order Matters! Improving Application Performance by Deliberately Delaying Packets | Hamid Ghasemirahni, Tom Barbette, Georgios P. Katsikas, Alireza Farshin, Amir Roozbeh, Massimo Girondi, Marco Chiesa, Gerald Q. Maguire Jr., Dejan Kostić | 16m |
Builds Reframer, a user-space reordering layer that classifies packets into per-flow queues using a hash table plus a flush linked list, briefly buffers them… | ||
| nothing matches | ||