A High-Speed Stateful Packet Processing Approach for Tbps Programmable Switches
Innovative approach for high-speed packet processing using programmable switches, external memory, and processing devices. Enables complex stateful operations at 300G throughput with a single server.
Doing More with Less - Orchestrating Serverless Applications without an Orchestrator
Explore application-level orchestration for serverless apps using Unum, offering flexibility, cost-efficiency, and performance gains over standalone orchestrators while maintaining programming interfaces and execution guarantees.
μMote - Enabling Passive Chirp De-spreading and μW-level Long-Range Downlink for Backscatter Devices
Innovative long-range receiver for backscatter devices, enabling passive chirp de-spreading and μW-level power consumption. Achieves 400m range at 2kbps with low BER, overcoming previous limitations.
Transparent GPU Sharing in Container Clouds for Deep Learning Workloads
Innovative system for efficient GPU sharing in container clouds, enabling higher utilization and faster job completion for deep learning workloads while maintaining performance isolation.
RF-Chord - Towards Deployable RFID Localization System for Logistic Networks
Innovative RFID localization system for efficient inventory tracking in logistics. Offers real-time processing, extended range, and reduced errors, enhancing automation in warehouses and delivery stores.
Practical Intent-Driven Routing Configuration Synthesis
Introduces Aura, a system for synthesizing datacenter routing policies, featuring RPL language and a compiler for automatic switch configuration generation, supporting frequent policy changes.
Enhancing Global Network Monitoring with Magnifier
Innovative approach to enhance global network monitoring by combining sampling with mirroring, improving coverage without increasing overhead. Introduces "negative mirroring" for efficient traffic validation.
Flattened Clos - Designing High-performance Deadlock-free Expander Data Center Networks Using Graph Contraction
Innovative approach to designing high-performance, deadlock-free expander data center networks using graph contraction, offering improved efficiency and throughput compared to traditional Clos networks.
Scalable Tail Latency Estimation for Data Center Networks
Innovative techniques for fast and accurate estimation of flow-level tail latency in large-scale data center networks, overcoming limitations of traditional simulators and recent machine learning approaches.
Electrode - Accelerating Distributed Protocols with eBPF
Accelerating distributed protocols using eBPF-based optimizations to improve performance by reducing user-kernel crossings and networking stack traversal, demonstrated with significant throughput and latency improvements in Multi-Paxos.
LemonNFV - Consolidating Heterogeneous Network Functions at Line Speed
Innovative NFV framework consolidating heterogeneous network functions without code modification, achieving high performance and interoperability through binary-level loading, intercepted I/O scheduling, and restricted memory allocation.
HALP - Heuristic Aided Learned Preference Eviction Policy for YouTube Content Delivery Network
Explore HALP, a machine learning-based caching algorithm for YouTube's CDN that improves efficiency, reduces byte miss rates, and addresses deployment challenges in large-scale production environments.
Poseidon - Efficient, Robust, and Practical Datacenter CC via Deployable INT
Poseidon: A novel congestion control protocol leveraging in-network telemetry for efficient, robust, and practical datacenter networks. Improves RTT, throughput, fairness, and convergence, resulting in significant application performance gains.
Acoustic Sensing and Communication Using Metasurface
Innovative acoustic system using metasurface and speakers for improved SNR, sensing accuracy, and communication reliability. Offers dynamic steering and outperforms phased array speakers in various scenarios.
Following the Data, Not the Function - Rethinking Function Orchestration in Serverless
Explores data-centric approach to serverless function orchestration, introducing Pheromone platform for efficient, scalable workflows with reduced latencies and simplified complex application implementation.
SPEEDEX - A Scalable, Parallelizable, and Economically Efficient Decentralized Exchange
Explore SPEEDEX, a high-throughput decentralized exchange offering secure asset trading, improved scalability, and prevention of internal arbitrage and front-running attacks within Layer-1 blockchains.
Formal Methods for Network Performance Analysis
Explores using formal methods for network performance analysis, offering a more exhaustive approach than simulation. Demonstrates effectiveness in analyzing packet scheduling algorithms and generating problematic workloads.
CellDAM - User-Space, Rootless Detection and Mitigation for 5G Data Plane
Innovative solution for detecting and mitigating data plane attacks in 5G networks without infrastructure changes. Utilizes state-dependent model checking to verify data signaling correctness at the device level.
LOCA - A Location-Oblivious Cellular Architecture
Innovative cellular architecture design protecting user location privacy while maintaining identity-based services. Explores MVNO-based systems, offering provable privacy guarantees and scalability for real-world deployment.
DChannel - Accelerating Mobile Applications With Parallel High-bandwidth and Low-latency Channels
Explores a novel wireless channel design using parallel high-bandwidth and low-latency channels to improve mobile app performance, demonstrating significant speed gains in web browsing and gaming.
Test Coverage for Network Configurations
NetCov: A tool for revealing network configuration lines tested by network tests. Helps improve test suites, increase reliability, and efficiently infer configuration elements' contributions to data plane elements.