Data compression is essential to large-scale data centers to save both storage and network bandwidth. Current software based method suffers from high computational cost with limited performance. In this project, we are migrating the fundamental workload of the computer system to FPGA accelerator, aiming high throughput performance and high energy efficiency, as well as freeing some CPU resources.
Software-defined radios (SDR) have a potential to bring major innovation in wireless networking design. However, their impact so far has been limited due to complex programming tools. Ziria addresses this problem. It consists of a novel programming language and an optimizing compiler. It is able to synthesize a very efficient SDR code from a high-level PHY description written in Ziria language.
Tetris is a cluster scheduler that packs, i.e., matches multi-resource task requirements with resource availabilities of machines. It allows cluster operators to trade-off cluster efficiency (makespan) for job completion time as well as performance for fairness.
This project targets on using automatic techniques to reduce MTTR of large-scale online service systems.
MODIST is a practical software model checker for unmodified concurrent, distributed and cloud systems. MODIST explores different execution paths systematically as well as simulating a variety of environment faults to discover subtle corner-case defects. We have applied MODIST in Oracle Berkely DB, MPS(Paxos implementation), SQL Azure, Windows Azure Storage and other real systems, and found many new bugs.
This is the website of the rack-scale computing research project at MSRC
This project re-imagines and re-engineers wide area networks, to more than double their efficiency and allow flexible sharing of resources.
Face In The Crowd examines the social impact of crowdsourcing platforms—cloud-based computational systems that allow the outsourcing of work through open requests—and how they might shape the future of work.
Labs: New England
Scalable and Practical App Digging Engine
Connected devices – sensors and actuators – have a growing impact on our society, environment and health. For example, there have been significant advances in gaining visibility into buildings' daily operations. The next step is to enable people to do more with the increasingly ubiquity of connected devices. To this end, Human-Building Analytics (HBA) data platform explores, for a wide spectrum of users, (1) more natural programmability for connected devices and (2) more personalized analytics.
-- Making it easy for app developers of all levels to test their apps under real-world contexts on the cloud or real devices --
Waypoint project is up and running in Building 99.
As PHY layer data rates increase, CSMA MAC overheads dominate. The 9 us slot width at 1Gbps data rate can result in MAC efficiency of under 10%. WiFi-Nano proposes a novel speculative transmission based technique that leverages self-interference cancelation and allows for using 800ns slots -- reducing CSMA overheads by an order of magnitude.
The quest for higher data rates in WiFi is leading to the development of standards that make use of wide channels (e.g., 40MHz in 802.11n and 80MHz in 802.11ac). We argue against this trend of using wider channels, and instead advocate that radios should communicate concurrently over multiple narrow channels for efficient and fair spectrum utilization. We propose WiFi-NC, a novel PHY-MAC design that allows radios to use WiFi over multiple narrow channels simultaneously.
Dhwani enables information theoretically secure Near Field Communication (NFC) on existing mobile phones without requiring any special hardware or PKI infrastructure. It uses existing microphones and speakers on phones to perform acoustic NFC.
The LKW project is aimed at designing low-power algorithms and systems for admission control to speech systems: i.e., detecting foreground speech, recognizing leading keywords and verifying speakers on a continuously-on wearable device. Our goal is to consume under 10 mW average on generic embedded hardware available today and under 100uW on custom hardware.
In data centers, the IO path to storage is long and complex. It comprises many layers or “stages” with opaque interfaces between them. This makes it hard to enforce end-to-end policies that dictate a storage IO flow’s performance (e.g., guarantee a tenant’s IO bandwidth) and routing (e.g., route an untrusted VM’s traffic through a sanitization middlebox). We are researching architectures that decouple control from data flow to enable such policies.
The Scalable Hyperlink Store is a specialized "database" for the web graph. SHS maintains the web graph in main memory, distributed over many machines. The system is available as C# source code as well as precompiled binaries.
A framework to reason about weaker forms of consistency and isolation in a replicated database.
Energy drain in mobile devices is well recognized to be a serious problem. One solution is to provide tools and guidelines to enable application writers build more energy efficient programs. This project explores an alternative that mitigates the ill-effects of an energy hungry application. Our system, E-Loupe, offers a finer-grained approach to ensure predictable energy drain in mobile devices.
Vision is the ultimate source of sensory input that we humans consume. We believe that the next generation of computers will provide the ability to continuously capture and analyze visual information in real-time, thus greatly enhancing the overall experience and efficiency of their users.
Column store technology can provide very substantial performance improvements on data warehousing workloads. This project investigated how to integrate columnar storage into SQL Server. The solution adopted was to add a new index type, columnstore index, that stores data column wise instead of row wise. Columnstore indexes first shipped in SQL Server 2012 and significant enhancements will be included in the next release.
This research project in MSR SVC aims to answer the following question: Can we allow programmers to write cloud applications as though they are accessing centralized, strongly consistent data while at the same time allowing them to specify their consistency/availability/performance (CAP) requirements in terms of service-level agreements (SLAs) that are enforced by the cloud storage system at runtime?
Optimus is a framework for dynamically rewriting an execution plan graph in distributed data-parallel computing at runtime. It enables optimizations that require knowledge of the semantics of the computation, such as language customizations for domain-specific computations including matrix algebra. We address several problems arising in distributed execution including data skew, dynamic data re-partitioning, unbounded iterative computations, and fault tolerance.
Real-time information about businesses such as, the current occupancy and music levels, as well as the type or exact song playing now, can be important factors in the local search decision process. In this work, we propose to automatically crowdsource such rich, real time business metadata through user check-in events.