CSE Ph.D. student Andrew Huynh (BS '10, Ph.D. '15 expected) is the Lead Data Scientist for the OASIS project and a member of the Distributed Health Labs in Calit2’s Qualcomm Institute. Data science is critical to OASIS, which launched a six-week campaign on Dec. 12 to raise $50,000 for R&D, production and deployment of a target 1,000 ‘citizen-sensors’ by mid-2015. THe campaign on Indiegogo.com is the first-ever philanthropic campaign by UC San Diego on any crowd-funding website – and therefore a test case for future efforts by researchers and students who may want to solicit gift funds from the public at large for high-profile research projects.
Huynh has been part of DHLabs since October 2012, when the informal group of researchers received a Calit2 Strategic Research Opportunities (CSRO) grant to pursue the concept of a ‘tricorder’ – the sci-fi handheld device popularized by Star Trek for scanning, analyzing and recording health data. The concept of the tricorder is so powerful that the UCSD team will compete in 2014 for the $10-million Qualcomm Tricorder XPRIZE, a global competition to stimulate innovation and integration of precision diagnostic technologies in order to make valid health monitoring available directly to consumers in their homes.
With institutional support from Calit2 and other sources, the team – led by Qualcomm Institute research scientist (and occasional CSE lecturer) Albert Yu-Min Lin and UC San Diego School of Medicine professor Eliah Aronoff-Spencer – has been developing the OASIS project to revolutionize global health and environmental monitoring, especially in remote and undeveloped areas of the planet.
Huynh (pictured at left from the OASIS video on YouTube) is leading the data effort on the OASIS project, including the development of an Open Health Stack using cloud infrastructure, mobile apps, and sensors to collect and analyze data from individual users and their environment. The team has already built prototype optical and electrochemical sensors that are networked through a smartphone, and the system is “almost ready” for integrated testing. The sensor device, called SENSE, is the first layer of the Open Health Stack and it will track vital signs and environmental contaminants such as heavy metals in water (and eventually infectious diseases such as cholera in streams). A MyOasis smartphone app will interact with the sensors and visualize the collected data. The final layer of the stack, called KEEP, is a secure data-storage and analysis platform for detecting large-scale trends such as flu outbreaks with the help of machine-learning algorithms.
If the OASIS campaign on Indiegogo brings in the desired $50,000 from donors, Huynh and his colleagues believe that they will be able to get the price of building each SENSE device down to below $50, a critical threshold. In addition to U.S. users, OASIS will distribute the devices at inaugural research sites in Mongolia (primarily to detect toxic heavy metals in ground water) as well as to health workers in Mozambique and Haiti. Read the full news release. Visit the OASIS project on Indiegogo.com. Learn more about Distributed Health Labs.
Two professors from UC San Diego's Computer Science and Engineering department are among the 50 members of the Association for Computing Machinery (ACM) elected Fellows of the organization for 2013. As announced on Dec. 10, CSE professors Yuanyuan (YY) Zhou and Mihir Bellare were among the elite group of researchers hailing from leading universities, corporations and research labs.
CSE Prof. Mihir Bellare (far right) was cited for “contributions to provable security methods supporting high-quality, cost-effective cryptography.” Bellare works in cryptography and security, particularly practical, proven-secure schemes. He co-designed the HMAC message encryption scheme that is used in many methods for ecrypting data over networks, including SSL, TLS, SSH, and IPSEC. Bellare previously received the ACM Kanellakis Theory and Practice Award, as well as the RSA Conference Award in Mathematics.
CSE Prof. YY Zhou (pictured near right) was recognized for her “contributions to software reliability and quality.” Zhou’s research interests include operating systems, networking, reliability and large-data analysis. She has focused on techniques for analyzing system data to improve software quality, manageability and reliability.
“We are extremely proud of the accomplishments of professors Bellare and Zhou,” said Rajesh Gupta, Chair of the CSE Department. “They have had an outstanding impact on their respective fields, and the ACM Fellowship is a fitting recognition of the high regard in which they are held among their peers in the computer science and engineering community.”
The 2013 inductees continue a trend averaging two UCSD CSE faculty members recognized as ACM Fellows every year since 2010. In that period, Andrew Kahng was recognized in 2012 for his contributions to physical design automation and to design for manufacturability of microelectronic systems. Keith Marzullo, Dean Tullsen and Amin Vahdat became ACM Fellows in 2011, with Pavel Pevzner and Stefan Savage elevated to Fellow status in 2010. Read the full news release.
What is your ‘urban tribe’? Your computer may soon be able to tell. CSE researchers are developing a computer-vision algorithm that uses group pictures to determine to which urban tribe an individual belongs. So far, on average, it’s accurate 48% of the time (compared to 9% if left to chance). But the researchers, including CSE professors Serge Belongie and David Kriegman, would like the algorithm to perform at least as well as a human might. “This is a first step,” said Belongie, a co-author of the study presented at the British Machine Vision Conference in the fall. “We are scratching the surface to figure out what the signals are.”
For purposes of algorithm development, the researchers used eight of the most popular ‘urban tribes’: biker, country, Goth, heavy metal, hip hop, hipster, raver and surfer. While humans can generally recognize urban tribes at a glance, computers cannot. So the algorithm segments each person in six sections—face, head, top of the head (where a hat would be), neck, torso and arms – and the algorithm analyzes each segment for haircuts, hair color, jewelry, tattoos, etc. The researchers analyzed group photos in order to pick up on social cues such as clothing and hairdos that may be shared by members of a particular tribe. A by-product of the research was the development of an extensive dataset of urban tribe images, which they plan to make available to other research groups.
An algorithm able to identify a person’s urban tribe could have a wide range of applications, from generating more relevant search results and ads, to allowing social networks to provide better recommendations and content. There also is a growing interest in analyzing footage from cameras installed in public spaces to identify groups rather than individuals.
In addition to Belongie, Kriegman and CSE Ph.D. student Iljung Sam Kwak, co-authors of the research included Peter Belhumeur of Columbia University, UC Berkeley Ph.D. alumnus Lubomir Bourdev, and Ana C. Murillo from the University of Zaragoza in Spain. Read the full news release.
Center for Networked Systems (CNS) research scientist George Porter has co-authored two papers with CSE students and colleagues to be presented Dec. 11 at the 9th ACM International Conference on Emerging Networking Experiments and Technologies (CoNEXT) in Santa Barbara, Calif.
According to “Bullet Trains: A Study of NIC Burst Behavior at Microsecond Timescales,” a lot is known at a macro level about the behavior of traffic in data center networks. This includes the ‘burstiness’ of TCP, variability based on destination, and overall size of network flows. However, according to lead CSE graduate student Rishi Kapoor (at right), Porter and CSE professors Geoff Voelker and Alex Snoeren, “little information is available on the behavior of data center traffic at packet-level timescales,” that is, at timescales below 100 microseconds. Some 30 years ago, an MIT study compared packets of data with train cars – sent from a source to the same destination back-to-back like train cars pulled by a locomotive. In the context of data centers, however, the UC San Diego researchers came to the conclusion that those trains are more aptly termed “bullet trains” when viewed at microsecond timescales. Porter and his colleagues examined the various sources of traffic bursts and measured the traffic from different sources along the network stack, as well as the burstiness of different data-center workloads and the burst behavior of bandwidth-intensive applications such as data sorting (MapReduce) and distributed file systems (NFS and Hadoop). “Our analysis showed that network traffic exhibits large bursts at sub-100 microsecond timescales,” said Porter. “Regardless of application behavior at the higher layer, packets come out of a 10 Gigabit-per-second server in bursts due to batching.” The larger the burst, he added, the greater the likelihood of packets being dropped.
The researchers focused primarily on the network interface controller (NIC) layer, because the controller is directly implicated in the burst behavior that most affects computer networking speeds. While it would be ideal if packets transmitted within a single flow would be uniformly paced, real life turns out to be more complex. This is primarily because packets are batched differently across the network stack in order to achieve link rates of 10Gbps or higher. For their paper, Porter and his co-authors studied the burst behavior of traffic emanating from a 10Gbps end-host across a variety of data center applications. “We found that at 10- to 100-microsecond timescales, the traffic exhibits large bursts, tens of packets in length,” said Porter. “We also found that this level of burstiness was largely outside of application control, and independent of the high-level behavior of applications.”
In the second study to be presented at CoNEXT, FasTrak: Enabling Express Lanes in Multi-Tenant Data Centers, lead graduate student Radhika Niranjan Mysore (at left), Porter, and their co-author CSE Prof. Amin Vahdat (on leave at Google) explore an issue specifically facing operators of cloud services, such as Amazon EC2, Microsoft Azure and Google Compute Engine. These so-called multi-tenant data centers may host tens of thousands of customers. No customer wants their data or service to leak into those of other customers in the cloud, and typically, cloud operators rely on virtual machines (VMs) as well as network-level rules and policies that hypervisors enforce on every packet going in and out of the host in order to ensure network isolation. As a result, however, VMs carry innate costs in the form of latency (delays) and the increased cost of processing packets in the hypervisor, which affect both the provider and the tenant. The researchers came up with a solution called FasTrak, which keeps the functionality but curbs the cost of rule processing by offloading some of the virtualization functionality from the hypervisor software to the network switch hardware through so-called “express lanes.” There is limited space on a switch – not enough to take care of all the rules required by a server – so for FasTrak, the researchers determined the subset of data flows that could benefit most from offloading via express lanes to hardware. The result: an approximate doubling in latency improvement (i.e., 50 percent less delay or time to finish), combined with a 21 percent drop in the server load. According to the study’s conclusion, FasTrak’s actual benefits are workload dependent, but “services that should benefit the most are those with substantial communication requirements and some communication locality.”
Recent CSE alumna Meg Walraed-Sullivan (Ph.D. ’12) is now a postdoctoral researcher in the Distributed Systems group at Microsoft Research in Redmond, Washington. The group investigates the scalability, security, fault tolerance, manageability, and performance of distributed systems. While in CSE, Walraed-Sullivan (at right) worked on data-center communications challenges such as (a) enabling scalable communication via strategic label assignment, and (b) exploring the relationship between fault tolerance and scalability properties of hierarchical topologies. Walraed-Sullivan is the first author on a paper at the 9th ACM International Conference on Emerging Networking Experiments and Technologies (CoNEXT) in Santa Barbara, Dec. 9-12. On Dec. 10, Walraed-Sullivan will present the paper “Aspen Trees: Balancing Data Center Fault Tolerance, Scalability and Cost,” co-authored with her Ph.D. advisors, Prof. Amin Vahdat (on leave at Google), and Prof. Keith Marzullo (recently on leave at NSF).
The paper flows from Walraed-Sullivan’s dissertation, which introduced a new class of network topologies called ‘Aspen trees,’ named after Aspen trees in nature, which share a common root system (at left). Large-scale data center infrastructures typically use a multi-rooted, 'fat tree' topology, which provides diverse yet short paths between end hosts. A drawback of this type of topology is that a single link failure can disconnect a portion of the network’s hosts for a substantial period of time (while updated routing information propagates to every switch in the tree). According to an advance copy of the CoNEXT paper, this shortcoming makes the fat tree less suited for use in data centers that require the highest levels of availability. Alternatively, Aspen tree topologies can provide the high throughput and path multiplicity of current data center network topologies, while also allowing a network operator to select a particular point on the spectrum of scalability, network size, and fault tolerance – affording data center operators the ability to react to failures locally.
Walraed-Sullivan and her co-authors also outline a corresponding failure-notification protocol, ANP, whose "notifications require less processing time, travel shorter distances, and are sent to fewer switches, significantly reducing re-convergence time and control overhead in the wake of a link failure or recovery.“ The paper concludes that “Aspen trees provide decreased convergence times to improve a data center’s availability, at the expense of scalability (e.g., reduced host count) or financial cost (e.g., increased network size).” The paper to be presented at CoNEXT details a thorough exploration of the tradeoffs among fault tolerance, scalability and network cost for data centers using an Aspen tree topology.
On Dec. 5, CSE Prof. Victor Vianu (at right) was in Switzerland and gave a talk at the École Polytechnique Fédérale de Lausanne (EPFL). The topic: automatic verification of the increasingly common workflows centered around data. Vianu drew on joint work with fellow CSE Prof. Alin Deutsch, Microsoft program manager and CSE alumnus Elio Damaggio (MS ’08, Ph.D. ’11), former CSE visiting scholar Fabio Patrizi (now a professor at the University of Rome-La Sapienza), and Richard Hull of IBM Research. Tools have been developed for high-level specification of such workflows and other data-driven applications. “Such specification tools not only allow fast prototyping and improved programmer productivity but, as a side effect, provide convenient targets for automatic verification,” said Vianu in the abstract for his talk, pointing to a notable example: IBM’s business artifact framework, which has been successfully deployed in practice.
Vianu presented a formal model of data-centric workflows based on business artifacts, and results on automatic verification of such processes. “Artifacts are tuples of relevant values, equipped with local state relations and accessing an underlying database,” he said. “They evolve under the action of services specified by pre- and post-conditions, that correspond to workflow tasks. The verification problem consists in statically checking whether all runs of an artifact system satisfy desirable properties, expressed in an extension of linear-time temporal logic.” In his talk at EPFL, Vianu exhibited several classes of specifications and properties that could be automatically verified. Determining that the results thus far have been “quite encouraging,” he said those results suggest that, unlike with arbitrary software systems, significant classes of data-centric workflows may be amenable to fully automatic verification. To do so, Vianu concluded, requires a “novel marriage of techniques” from the database and computer-aided verification areas.
This week teams of students taking CSE 118, Applications in Ubiquitous Computing, will be presenting their finished apps to lecturer and research scientist Nadir Weibel (at left) and his teaching assistant Eric Seidel. Each group did a quarter-long Microsoft Kinect project, and a week-long Google Glass project. Then next week onDec. 10 from 3-5pm, the students' 10 Kinect and Google Glass applications will be on display in the main lobby of the CSE Building. Passers-by will be invited to test the apps, which were designed to reflect a theme running through CSE 118 this quarter: Augmented Reality. The department will offer snacks and refreshments during the demonstrations just outside of CSE 1202.
Weibel is a research scientist and lecturer in CSE and a research health science specialist at the VA San Diego Health System. He works on human-centered computing at the intersection of computer science, cognitive science, communication, health and social sciences. While he has been at UC San Diego since 2009, Weibel joined the CSE faculty as a lecturer and research scientist in 2013.
The 50 students taking CSE 118 were broken into 10 groups, and each team of five students was given access to a Kinect for the whole quarter. "We went over the SDK and how to use the different features," says Weibel, referring to the Software Development Kit. "There was no guideline in terms of the kind of apps to be developed, but there was continuous coaching and discussion during the quarter on the appropriateness of the proposed idea." Health apps prevailed, such as "Ubicook", a Kinect app (at right) for the "chef inside of all of us, enabling people everywhere to make sense of the supplies they have on hand. Ubicook solves the manual input process of locating recipes by scanning and identifying food items and automatically querying recipe sites to save time and money."
For the Google Glass project, there was only funding for two of the systems, so students rotated, and every week, one of the groups would stop working on the Kinect for a "week-long Glass hackathon," says Weibel. For the Glass project, teams were encouraged to focus on an healthcare application. Click here to watch some of the Google Glass project videos.
According to Weibel, "ubiquitous computing is an interdisciplinary field that includes technologies that bridge the digital and physical worlds, systems and applications that incorporate such technologies, infrastructures that support them, human activities and experiences the technologies facilitate, and conceptual overviews that help us understand - or challenge our understanding of - the impact of these technologies." Read more about the CSE 118 course here.
CSE Prof. Larry Smarr will moderate a discussion on "The Quantified Self Movement," the title of an upcoming MIT Enterprise Forum organized by the Forum's San Diego chapter. The January 15 event will take place from 5pm to 8:30pm in the UCSD Medical Education and Telemedicine Building on the UC San Diego campus. Smarr - the founding director of Calit2 and "defacto evangelist" of the quantified-self movement - is the "poster man for the medical strategy of the future," according to a 2012 article in the MIT Technology Review. As promoted by the MIT Enterprise Forum, the session and "lively panel discussion" will attempt to answer certain questions: What is the Quantified Self movement and why are fitness buffs, techno geeks and patients with chronic conditions obsessively monitoring their various personal metrics? Is this trend the beginning of a major shift in how we look at our health? What is the distinction between digital health and self-monitoring? Why are global sports clothing companies investing in this technology? According to organizers, attendees will also learn about the "growing availability of inexpensive monitoring devices, personal genetic profiles, and the increasing sophistication of software apps and social networks which promise to fuel the self-tracking revolution around personal health and fitness." Click here to register and attend the Forum.
Is it possible to teach classes with creative, open-ended projects on a massive scale? Difficult, but not impossible, according to CSE Prof. Scott Klemmer. Massive online courses benefit from the 'wisdom of the crowd,' he says, which can actually enhance a crucial element of creative coursework: peer assessment. The challenge, according to Klemmer, speaking at a lecture organized by the Qualcomm Institute's Technology Enhanced Learning (TEL) Initiative, is that students must know what 'good' means, which can be difficult when they are sitting behind computer screens scattered all over the world. But the path is smoother now that Klemmer and colleagues have developed a set of best practices enshrined in what he calls the Seven Habits for Highly Effective Peer Assessment, which have been used in more than 100 massive open online courses (MOOCs). Klemmer's research is informed by his own collaboration with Coursera in 2012 to launch the first massive-scale class -- a design course - with self and peer assessment. Read news release about peer assessment.
CSE Prof. Andrew Kahng (pictured at left) and former CSE postdoctoral scholar Hailong Yao (right) are named on one of only two patents awarded by the U.S. Patent Office in the first half of 2013 to researchers in the Jacobs School of Engineering. That’s according to the UC San Diego Technology Transfer Office. The patent filed through the campus protects the intellectual property developed by Kahng and Yao for “layout decomposition for double patterning lithography” (U.S. Patent # 8,402,396). The invention, first submitted in 2010, provides systems and methods for layout decomposition to produce exposure layouts that can be used to perform double patterning lithography (DPL). Yao spent two years in Kahng’s lab. Then in 2009, he returned to China to become an assistant professor in the Department of Computer Science and Technology at Beijing’s Tsinghua University, where he had earned his Ph.D. in 2007. “In my postdoctoral research, I focused on the areas of design for manufacturing, delay and leakage optimization, etc.,” said Yao. DPL layout composition was also the topic of a joint 2010 paper Kahng, Yao and other colleagues published in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems. The paper focused on process nodes of 45 nanometers or below. Yao’s current research group in Beijing focuses on the area of VLSI physical design, including topics of floor-planning, placement, routing, clock tree synthesis and routing, timing analysis and optimization.
Have technology you want to protect? Click here for the latest TTO newsletter, including contact information. http://bit.ly/18dQJfg