Monthly Archives: April 2020

MIT researchers have designed a congestion-control scheme for wireless networks that could help reduce lag times and increase quality in video streaming, video chat, mobile gaming, and other web services.

To keep web services running smoothly, congestion-control schemes infer information about a network’s bandwidth capacity and congestion based on feedback from the network routers, which is encoded in data packets. That information determines how fast data packets are sent through the network.

Deciding a good sending rate can be a tough balancing act. Senders don’t want to be overly conservative: If a network’s capacity constantly varies from, say, 2 megabytes per second to 500 kilobytes per second, the sender could always send traffic at the lowest rate. But then your Netflix video, for example, will be unnecessarily low-quality. On the other hand, if the sender constantly maintains a high rate, even when network capacity dips, it could  overwhelm the network, creating a massive queue of data packets waiting to be delivered. Queued packets can increase the network’s delay, causing, say, your Skype call to freeze.

Things get even more complicated in wireless networks, which have “time-varying links,” with rapid, unpredictable capacity shifts. Depending on various factors, such as the number of network users, cell tower locations, and even surrounding buildings, capacities can double or drop to zero within fractions of a second. In a paper at the USENIX Symposium on Networked Systems Design and Implementation, the researchers presented “Accel-Brake Control” (ABC), a simple scheme that achieves about 50 percent higher throughput, and about half the network delays, on time-varying links.

The scheme relies on a novel algorithm that enables the routers to explicitly communicate how many data packets should flow through a network to avoid congestion but fully utilize the network. It provides that detailed information from bottlenecks — such as packets queued between cell towers and senders — by repurposing a single bit already available in internet packets. The researchers are already in talks with mobile network operators to test the scheme.

“In cellular networks, your fraction of data capacity changes rapidly, causing lags in your service. Traditional schemes are too slow to adapt to those shifts,” says first author Prateesh Goyal, a graduate student in CSAIL. “ABC provides detailed feedback about those shifts, whether it’s gone up or down, using a single data bit.”

Joining Goyal on the paper are Anup Agarwal, now a graduate student at Carnegie Melon University; Ravi Netravali, now an assistant professor of computer science at the University of California at Los Angeles; Mohammad Alizadeh, an associate professor in MIT’s Department of Electrical Engineering (EECS) and CSAIL; and Hari Balakrishnan, the Fujitsu Professor in EECS. The authors have all been members of the Networks and Mobile Systems group at CSAIL.

Achieving explicit control

Traditional congestion-control schemes rely on either packet losses or information from a single “congestion” bit in internet packets to infer congestion and slow down. A router, such as a base station, will mark the bit to alert a sender — say, a video server — that its sent data packets are in a long queue, signaling congestion. In response, the sender will then reduce its rate by sending fewer packets. The sender also reduces its rate if it detects a pattern of packets being dropped before reaching the receiver.

In attempts to provide greater information about bottlenecked links on a network path, researchers have proposed “explicit” schemes that include multiple bits in packets that specify current rates. But this approach would mean completely changing the way the internet sends data, and it has proved impossible to deploy. 

“It’s a tall task,” Alizadeh says. “You’d have to make invasive changes to the standard Internet Protocol (IP) for sending data packets. You’d have to convince all Internet parties, mobile network operators, ISPs, and cell towers to change the way they send and receive data packets. That’s not going to happen.”

With ABC, the researchers still use the available single bit in each data packet, but they do so in such a way that the bits, aggregated across multiple data packets, can provide the needed real-time rate information to senders. The scheme tracks each data packet in a round-trip loop, from sender to base station to receiver. The base station marks the bit in each packet with “accelerate” or “brake,” based on the current network bandwidth. When the packet is received, the marked bit tells the sender to increase or decrease the “in-flight” packets — packets sent but not received — that can be in the network.

If it receives an accelerate command, it means the packet made good time and the network has spare capacity. The sender then sends two packets: one to replace the packet that was received and another to utilize the spare capacity. When told to brake, the sender decreases its in-flight packets by one — meaning it doesn’t replace the packet that was received.

Used across all packets in the network, that one bit of information becomes a powerful feedback tool that tells senders their sending rates with high precision. Within a couple hundred milliseconds, it can vary a sender’s rate between zero and double. “You’d think one bit wouldn’t carry enough information,” Alizadeh says. “But, by aggregating single-bit feedback across a stream of packets, we can get the same effect as that of a multibit signal.”

Staying one step ahead

At the core of ABC is an algorithm that predicts the aggregate rate of the senders one round-trip ahead to better compute the accelerate/brake feedback.

The idea is that an ABC-equipped base station knows how senders will behave — maintaining, increasing, or decreasing their in-flight packets — based on how it marked the packet it sent to a receiver. The moment the base station sends a packet, it knows how many packets it will receive from the sender in exactly one round-trip’s time in the future. It uses that information to mark the packets to more accurately match the sender’s rate to the current network capacity.

In simulations of cellular networks, compared to traditional congestion control schemes, ABC achieves around 30 to 40 percent greater throughput for roughly the same delays. Alternatively, it can reduce delays by around 200 to 400 percent by maintaining the same throughput as traditional schemes. Compared to existing explicit schemes that were not designed for time-varying links, ABC reduces delays by half for the same throughput. “Basically, existing schemes get low throughput and low delays, or high throughput and high delays, whereas ABC achieves high throughput with low delays,” Goyal says.

Next, the researchers are trying to see if apps and web services can use ABC to better control the quality of content. For example, “a video content provider could use ABC’s information about congestion and data rates to pick the resolution of streaming video more intelligently,” Alizadeh says. “If it doesn’t have enough capacity, the video server could lower the resolution temporarily, so the video will continue playing at the highest possible quality without freezing.”

Read More

Imagine you’ve been diagnosed as Covid-19 positive. Health officials begin contact tracing to contain infections, asking you to identify people with whom you’ve been in close contact. The obvious people come to mind — your family, your coworkers. But what about the woman ahead of you in line last week at the pharmacy, or the man bagging your groceries? Or any of the other strangers you may have come close to in the past 14 days?

A team led by MIT researchers and including experts from many institutions is developing a system that augments “manual” contact tracing by public health officials, while preserving the privacy of all individuals. The system relies on short-range Bluetooth signals emitted from people’s smartphones. These signals represent random strings of numbers, likened to “chirps” that other nearby smartphones can remember hearing.

If a person tests positive, they can upload the list of chirps their phone has put out in the past 14 days to a database. Other people can then scan the database to see if any of those chirps match the ones picked up by their phones. If there’s a match, a notification will inform that person that they may have been exposed to the virus, and will include information from public health authorities on next steps to take. Vitally, this entire process is done while maintaining the privacy of those who are Covid-19 positive and those wishing to check if they have been in contact with an infected person.

“I keep track of what I’ve broadcasted, and you keep track of what you’ve heard, and this will allow us to tell if someone was in close proximity to an infected person,” says Ron Rivest, MIT Institute Professor and principal investigator of the project. “But for these broadcasts, we’re using cryptographic techniques to generate random, rotating numbers that are not just anonymous, but pseudonymous, constantly changing their ‘ID,’ and that can’t be traced back to an individual.”

This approach to private, automated contact tracing will be available in a number of ways, including through the privacy-first effort launched at MIT in response to Covid-19 called SafePaths. This broad set of mobile apps is under development by a team led by Ramesh Raskar of the MIT Media Lab. The design of the new Bluetooth-based system has benefited from SafePaths’ early work in this area.

Bluetooth exchanges

Smartphones already have the ability to advertise their presence to other devices via Bluetooth. Apple’s “Find My” feature, for example, uses chirps from a lost iPhone or MacBook to catch the attention of other Apple devices, helping the owner of the lost device to eventually find it. 

“Find My inspired this system. If my phone is lost, it can start broadcasting a Bluetooth signal that’s just a random number; it’s like being in the middle of the ocean and waving a light. If someone walks by with Bluetooth enabled, their phone doesn’t know anything about me; it will just tell Apple, ‘Hey, I saw this light,’” says Marc Zissman, the associate head of MIT Lincoln Laboratory’s Cyber Security and Information Science Division and co-principal investigator of the project.

With their system, the team is essentially asking a phone to send out this kind of random signal all the time and to keep a log of these signals. At the same time, the phone detects chirps it has picked up from other phones, and only logs chirps that would be medically significant for contact tracing — those emitted from within an approximate 6-foot radius and picked up for a certain duration of time, say 10 minutes.

Phone owners would get involved by downloading an app that enables this system. After a positive diagnosis, a person would receive a QR code from a health official. By scanning the code through that app, that person can upload their log to the cloud. Anyone with the app could then initiate their phones to scan these logs. A notification, if there’s a match, could tell a user how long they were near an infected person and the approximate distance.  

Privacy-preserving technology

Some countries most successful at containing the spread of Covid-19 have been using smartphone-based approaches to conduct contact tracing, yet the researchers note these approaches have not always protected individual’s privacy. South Korea, for example, has implemented apps that notify officials if a diagnosed person has left their home, and can tap into people’s GPS data to pinpoint exactly where they’ve been.

“We’re not tracking location, not using GPS, not attaching your personal ID or phone number to any of these random numbers your phone is emitting,” says Daniel Weitzner, a principal research scientist in the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and co-principal investigator of this effort. “What we want is to enable everyone to participate in a shared process of seeing if you might have been in contact, without revealing, or forcing anyone to reveal, anything.”

Choice is key. Weitzner sees the system as a virtual knock on the door that preserves people’s right to not answer it. The hope, though, is that everyone who can opt in would do so to help contain the spread of Covid-19. “We need a large percentage of the population to opt in for this system to really work. We care about every single Bluetooth device out there; it’s really critical to make this a whole ecosystem,” he says.

Public health impact

Throughout the development process, the researchers have worked closely with a medical advisory team to ensure that this system would contribute effectively to contact tracing efforts. This team is led by Louise Ivers, who is an infectious disease expert, associate professor at Harvard Medical School, and executive director of the Massachusetts General Hospital Center for Global Health.

“In order for the U.S. to really contain this epidemic, we need to have a much more proactive approach that allows us to trace more widely contacts for confirmed cases. This automated and privacy-protecting approach could really transform our ability to get the epidemic under control here and could be adapted to have use in other global settings,” Ivers says. “What’s also great is that the technology can be flexible to how public health officials want to manage contacts with exposed cases in their specific region, which may change over time.”

For example, the system could notify someone that they should self-isolate, or it could request that they check in through the app to connect with specialists regarding daily symptoms and well-being. In other circumstances, public health officials could request that this person get tested if they were noticing a cluster of cases.

The ability to conduct contact tracing quickly and at a large scale can be effective not only in flattening the curve of the outbreak, but also for enabling people to safely enter public life once a community is on the downward side of the curve. “We want to be able to let people carefully get back to normal life while also having this ability to carefully quarantine and identify certain vectors of an outbreak,” Rivest says.

Toward implementation

Lincoln Laboratory engineers have led the prototyping of the system. One of the hardest technical challenges has been achieving interoperability, that is, making it possible for a chirp from an iPhone to be picked up by an Android device and vice versa. A test at the laboratory late last week proved that they achieved this capability, and that chirps could be picked up by other phones of various makes and models.

A vital next step toward implementation is engaging with the smartphone manufacturers and software developers — Apple, Google, and Microsoft. “They have a critical role here. The aim of the prototype is to prove to these developers that this is feasible for them to implement,” Rivest says. As those collaborations are forming, the team is also demonstrating its prototype system to state and federal government agencies.

Rivest emphasizes that collaboration has made this project possible. These collaborators include the Massachusetts General Hospital Center for Global Health, CSAIL, MIT Lincoln Laboratory, Boston University, Brown University, MIT Media Lab, The Weizmann Institute of Science, and SRI International.

The team also aims to play a central, coordinating role with other efforts around the country and in Europe to develop similar, privacy-preserving contact-tracing systems.

“This project is being done in true academic style. It’s not a contest; it’s a collective effort on the part of many, many people to get a system working,” Rivest says.  

Read More

For decades researchers have envisioned a world where digital user interfaces are seamlessly integrated with the physical environment, until the two are virtually indistinguishable from one another. 

This vision, though, is held up by a few boundaries. First, it’s difficult to integrate sensors and display elements into our tangible world due to various design constraints. Second, most methods to do so are limited to smaller scales, bound by the size of the fabricating device. 

Recently, a group of researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) came up with SprayableTech, a system that lets users create room-sized interactive surfaces with sensors and displays. The system, which uses airbrushing of functional inks, enables various displays, like interactive sofas with embedded sensors to control your television, and sensors for adjusting lighting and temperature through your walls.

SprayableTech lets users channel their inner Picassos: After designing your interactive artwork in the 3D editor, it automatically generates stencils for airbrushing the layout onto a surface. Once they’ve created the stencils from cardboard, a user can then add sensors to the desired surface, whether it’s a sofa, a wall, or even a building, to control various appliances like your lamp or television. (An alternate option to stenciling is projecting them digitally.)

“Since SprayableTech is so flexible in its application, you can imagine using this type of system beyond walls and surfaces to power larger-scale entities like interactive smart cities and interactive architecture in public places,” says Michael Wessely, postdoc in CSAIL and lead author on a new paper about SprayableTech. “We view this as a tool that will allow humans to interact with and use their environment in newfound ways.”

The race for the smartest home has now been in the works for some time, with a large interest in sensor technology. It’s a big advance from the enormous glass wall displays with quick-shifting images and screens we’ve seen in countless dystopian films. 

The MIT researchers’ approach is focusing on scale, and creative expression. By using the airbrush technology, they’re no longer limited to the size of the printer, the area of the screen-printing net, or the size of the hydrographic bath — and there’s thousands of possible design options. 

Let’s say a user wanted to design a tree symbol on their wall to control the ambient light in the room. To start the process, they would use a toolkit in a 3D editor to design their digital object, and customize for things like proximity sensors, touch buttons, sliders, and electroluminescent displays. 

Then, the toolkit would output the choice of stencils: fabricated stencils cut from cardboard, which are great for high-precision spraying on simple, flat, surfaces, or projected stencils, which are less precise, but better for doubly-curved surfaces. 

Designers can then spray on the functional ink, which is ink with electrically functional elements, using an airbrush. As a final step to get the system going, a microcontroller is attached that connects the interface to the board that runs the code for sensing and visual output.

The team tested the system on a variety of items, including:

• a musical interface on a concrete pillar;
• an interactive sofa that’s connected to a television;
• a wall display for controlling light; and
• a street post with a touchable display that provides audible information on subway stations and local attractions.

Since the stencils need to be created in advance via the digital editor, it reduces the opportunity for spontaneous exploration. Looking forward, the team wants to explore so-called “modular” stencils that create touch buttons of different sizes, as well as shape-changing stencils that adjust themselves based on a desired user interface shape. 

“In the future, we aim to collaborate with graffiti artists and architects to explore the future potential for large-scale user interfaces in enabling the internet of things for smart cities and interactive homes,” says Wessely.

Wessely wrote the paper alongside MIT PhD student Ticha Sethapakdi, MIT undergraduate students Carlos Castillo and Jackson C. Snowden, MIT postdoc Isabel P.S. Qamar, MIT Professor Stefanie Mueller, University of Bristol PhD student Ollie Hanton, University of Bristol Professor Mike Fraser, and University of Bristol Associate Professor Anne Roudaut. 

Read More