1. Bridging the Information Divide Victor Bahl Senior Researcher Manager, Networking Group Microsoft Research January 2005
2.
3. Technology alone is not the answer… The Real Digital Divide The Economist Mar 10th 2005 Deeper socio-economic issues have to be addressed However technology can help….
4.
5. Thirst for Connection… VSAT for kiosk in Kodia, Madhya Pradesh n-Logue corDECT wireless tower
32. What are people doing with mobile phones today? SMS Text-messaging Purchasing Web Surfing & email Watching & Sharing Video Taking & Sharing Pictures Expression of Identity Tracey Lovejoy
34. SmartPhone Ecosystem Talk to people Access information on the Internet Monitor personal health & diagnose problems Improve social interactions Share experiences via AirBlogs Energy Harvesting MEMS Display WITTY HealthGear SPOT Barcode Reader μ PEN Xnav MSR Hardware Group WLAN, Cellular
35. What could people be doing with mobile phones tomorrow? WiTTY Health Monitoring Social Grouping
36. WITTY W ho I s T alking T o Y ou? Array microphone + Bone vibration sensor = Speaker gating Lower noise Longer battery life Mike Sinclair :: Zhengyou Zhang Reduces effects of ambient noise
37.
38.
39.
40.
41.
42. Nuria Oliver :: Fernando-Flores Mangas :: Mike Sinclair No Apnea Mild-Severe Apnea Severe Apnea Test Results
43.
44.
45.
46.
47.
48. Call To Action Together academia, government, and industry must develop common vision Perform scenario & systems based research tackling hard problems Partner in building and deploying real-world test beds
This talk describes technology options for providing connectivity and services to people around the world who are currently not benefiting from the technology innovations. The goal is to inspire researchers & students to work on problems that have the potential of impacting billions. The research described in this talk represents a few projects researcher in MSR are pursuing in this space. We recognize that the problems are not simple and solutions are not limited to providing Internet technology only. To elevate the standard-of-living, people from govt., industry and academia have to come together and go after a common vision. MSR researchers are investing time and money and making our software available to academia (in source form) to speed up this march towards reaching a vision of ubiquitous access & services. The talk does not cover all projects MSR is engaged in that are relevant in this space -- One notable omission is MSR India’s work on rural computing.
A substantial portion of our world still does not have connectivity. There people see technology as a luxury and its value is not clear. However, there exist plenty of examples where easy information access and communications has made a difference in improving people’s lives.
Project Akshaya – Akshaya addresses three issues in IT Dissemination to masses in rural India: (1) Bringing the benefits of technology to the Households (Access) , (2) providing ample information base in local language relevant to citizens' lives (Contents) and (3) sufficient understanding of the world of Information Technology and how it can touch their lives (Skillsets) in terms of using simple programs, internet browsing for information, emails etc... http://akshaya.net/akshaya/ Project Bhoomi - Bhoomi is a self-sustainable e-governance project for the computerized delivery of 20 million rural land records to 6.7 million farmers through 177 Government owned kiosks in the Indian state of Karnataka which has eliminated red tape and corruption in the issue of land title records, and is fast becoming the backbone for credible IT-enabled Government services for the rural population. http://www.revdept-01.kar.nic.in/Bhoomi/Importance.htm Project RAJiv –The Govt. launched the Rajiv Internet Village programme to bring Govt. services/benefits closer to the people living in villages and rural areas. Rajiv Internet village centres are being set up in 8618 villages across the State. These centres are expected to be functional by October, 2005. http://www.aponline.gov.in/Quick%20Links/events/RIV%20Launch/RaJiv.htm
This graph shows that even in developed nations like the US, a large percentage of the people with low income do not have broadband Internet connectivity. The $40 - $50 /month cost is a bit too much for them to pay. In order to make access to information ubiquitous, the cost has to come down. Wireless community networks, can accelerate deployment by lowering cost and reachablity barrier. One Internet gateway into the neighborhood can feed most houses in the neighborhood.
Kentaro and Karishma have visited many villages in India to come up with this list of services. Browsing on the web, many rural projects that have been started by many government initiatives & companies focus on providing such services.
Cheskin (www.cheskin.com) consulting & strategic Market Research firm was hired by Microsoft to determine what services people most care about. Their study included 3 focus groups / market across 2 markets (Seattle, Austin). The 3 groups were (1) Urban focus group – a mix of single family home and shared home dwellers, acquainted with 4+ neighbors, broadband users, multiple PC owners, some with home networks (2) Suburban focus group – a mix of single family home and condominium dwellers, acquainted with 4+ neighbors, broadband users, multiple PC owners, some with home networks, and (3) Rural focus group - single family home dwellers, dial up or broadband users, multiple PC owners, some with home networks. The lettering in red indicates commonality with Kentaro’s and Karishma’s findings.
Data is from the FCC, December 22, 2004 report under “Local Telephone Competition and Broadband Deployment”, Report title is: “Federal Communications Commission releases data on High-Speed Internet Access Services ”. Note: FCC defines High-speed lines as 200 Kbps in at least one-direction and Advanced Services Lines where speeds > 200 Kbps in both directions. High-speed lines : 32.5 million (30.1 million serve residential & small business subscribers); General break-up ADSL: 11.4 million, Coaxial: 18.6 million, Wireless/satellite: 2.5 million Advanced Services Lines : 23.5 million (of the 32.5 million) – of these 21.2 million serve residential & small businesses. From CIA FactBook US Population (July 2005 est): 295 million Internet users: 159 million (as of 2002)
..has to be inexpensive, easy to use and maintain.
Back of the envelope calculations – See the accompanying Excel spread sheet. From Tren, $1000 / home to do an overlay (term used by wireline broadband providers). Assuming $110 million households (293 million population), the average amount to do a complete wired solution is close to 110 billion.
802.11 chip cost have plummeted, The BoM for 802.11 < $10 We can keep the same Physical layer, i.e. OFDM, but the MAC (carrier Sense Multiple Access with Collision Avoidance CSMA/CA) may change – by doing this we can reuse the chipset, and hence benefit from the lower cost.
There are two options for last mile wireless connectivity (1) an infra-structure based system [the classic hub & spoke Network] (2) a Mesh network [where every node is equal]. Example of (1) is the WiFi networks deployed in companies, airports, hotels public places etc. We prefer the mesh network, because theoretically it does not require ISPs. In a mesh network, anyone can become an ISP. For example an entrepreneur kid in the neighbourhood can get a T-1 link and start providing Internet Service to the neighbourhood. Similarly a local school could provide Internet connectivity to the neighbouring houses. A mesh network is relatively inexpensive to deploy because the “customer” owns his or her equipment. People collaborate and give resources to the network and in-turn cooperatively use the network.
People buy a Wireless Access Point for their home network. These AP are equipped with multiple radios and mesh software. The radios is end devices (laptops, desktops, tablets, phones etc.) to connect. The APs also discover one another and connect to form a wireless mesh backbone network. The network ends at a Internet Gateways (there can be a number of Internet gateways). Internet feed could be in the form of T1, Satellite, WiMAx, etc. Technologies for community networking are directly applicable to: Connecting villages in rural areas City-wide blanket coverage Enterprise-wide wireless networks (i.e. networking in old buildings without cables) What about WiMax ? WiMAx as a long-haul wireless backbone technology can works nicely with wireless meshes. In low density areas WiMax requires high power towers or lots of towers; not very cost effective NLOS reduces range & bandwidth significantly, no longer as attractive End-node Equipment expected to be fairly expensive in the next 3 to 5 years, compared to WiFi Requires infra-structure roll out and ISP services
Within mesh networking there are two different architectural options. (1) infra-structure based, this is like the Metricom Richchet network. Radio transmitters are placed on pole-tops and end-nodes connect to them wirelessly – this is similar to a corporate WLAN network. The difference is that the poletop receivers form a mesh between themselves. (2) every node is a a router. (Note companies like Motorola have products of both kind) Why did Ricochet fail? Required physical infrastructure build-out of wireless access points every ¼ mile or so; more of a last 100 ft. solution than last mile. Went bankrupt as a result and now revived in Denver and San Diego. Back-haul cost was very high
There are others as well, but most are non-academic – for example various community networks that are sprouting up everywhere (1) Seattle Wireless (2) Urbana Champaign wireless (3) Free-Net in Erlangen, Germany (4) Free-Net in Finland etc. Check out: http://www.scn.org/commnet/ (or simply search for “community Network” in one of the search engines. Most of these are grass root efforts. The ones listed above are academic / research efforts.
Video shows three ideas MSR Redmond has developed (1) Multi-radio meshes for capacity & scale; (2) dual-frequency meshes for range (3) layer 2.5 source routing with a multi-radio link metric. In addition to this MSR Redmond has also made significant progress in (4) Fault detection & diagnosis (5) Security . We have also been pioneers in pushing for spectrum regulation reforms. We made one of the first (probably the first) technical proposal on spectrum etiquettes and lobbied the FCC to open up additional bandwidth in the 5 GHz range and more recently yhe 700 MHz band for unlicensed use (in November 2003, FCC opened up additional 255 MHz for unlicensed use.
Announce the Academic Resource Kit. The kit contains complete source code for our mesh connectivity layer and Virtual WiFi technologies. We are also giving away source code for our performance measurement and debugging tools along with DDK, WinXP, WinCE, videos, papers, and presentations.
Interesting components to point out GPS (location services) WiFi (local area networking); Bluetooth (personal area networking); GSM / GPRS (cellular networking) Strong Arm processor,
Play the Audio file
Current ways to Diagnose Sleep Apnea “ Classical” Diagnoses of Sleep Apnea Nocturnal Polysomnography (PSG): Patient stays in the sleep center for 1-2 nights System continuously monitors 8 physiological signals via simultaneous multi-channel measurements Very expensive, cumbersome, time consuming, just one sample (i.e. one night at the sleep enter) and subject to manual scoring and human error Pulse Oximetry: Useful as screening and diagnostic tool One simple, light-weight sensor on finger, toe or earlobe
21 volunteers: 80% male, ages 25-65 30% healthy, 70% with sleep apnea or suspected Wore HealthGear for one full night in their own homes
Slam is an application for the smartphone being developed by the social computing research group. This application allows users to create and communicate with groups of people from their mobile device. Take the dinner party scenario: you are at a dinner party and you just want the other folks there to have the pictures you are taking from your camera phone, but options for sharing, such as via mms or posting to a website are so much effort they spoil the spontaneity and fun of it. Slam supports really simple group formation from the mobile device just for this reason. Once you quickly create a slam group for the dinner party, picture sharing is built in to the picture taking process: pictures taken from within slam are distributed to all other group member’s mobiles with a single message. Based on previous research done in the group as well as ethnographic studies of mobile device use, the more common usage scenario for slam is group-based communication for social coordination. At a conference and want to coordinate a lunch? Again, quickly make a slam group for on-the-fly coordination around when and where to meet and who is going to the lunch. And of course these groups can be more persistent groups like friends and family. From a social computing perspective, we’re really trying to enable people to be not just always on, but always in touch with the people they care about.
The mobile device is just critical to providing always in touch social interactions because it’s the computer you have with you when are out in the physical world being social. This allows you to supporting naturally existing social interactions: Humans are very evolved for physical social interaction, with well established practices for self-presentation, the reading of social contextual cues, and so on. Given that the device is with the user in her physical world social environment, we can use technology to augment these interactions rather than creating exclusively digital social interactions like many of the social networking sites. This also allows you to support hyperawareness: because the device is always with the person, she can always be aware of people and events she cares about. Ethnographic studies of early adopters of these mobile social technologies highlight the role of in-time, in-place updates and changes of plans, or “hypercoordination”. For example, if you look at the social coordination practices of teens in Japan, rather than specifying precise times and places to meet, they start with a neighborhood and general time, such as Shibuya in the early afternoon, and through mobile messaging they coordinate on the fly to converge on a time and place.
A couple of screen shots here point out the primary functionality of slam. The large image in the upper right shows the application home screen. Each of the 4 favorite groups –or slams- are represented by the most recently shared photo in order to give a sense of social presence right at the top level. At the bottom of the main screen is a list of upcoming events – or “jams”. Because this is a social system with a focus on supporting our everyday social lives, events are very simple in comparison to the more formal or structured events like scheduled meetings at work or eVites, which are geared toward larger events. Events are simply messages sent to a group with a date and time associated with them that other group members can respond to. So, here we’re trying to support the common and often very fluid social events we all have, such as the Japanese teens meeting on a Saturday afternoon or a group of friends deciding what movie to go to on a Friday night. The screen on the left shows a click into the page for an individual group. Again, all messages are distributed to the entire group, creating a group-wide conversation thread. The conversation brings together text message and photos, as well as group events and membership. Making a new group is as simple as giving it a name, selecting some people to invite – either from other slams or by phone number, and clicking ‘create’. And that really is the core of slam: helping people communicate, coordinate, and share their lives through simple group creation and conversation on mobile devices.
A couple of screen shots here point out the primary functionality of slam. The large image in the upper right shows the application home screen. Each of the 4 favorite groups –or slams- are represented by the most recently shared photo in order to give a sense of social presence right at the top level. At the bottom of the main screen is a list of upcoming events – or “jams”. Because this is a social system with a focus on supporting our everyday social lives, events are very simple in comparison to the more formal or structured events like scheduled meetings at work or eVites, which are geared toward larger events. Events are simply messages sent to a group with a date and time associated with them that other group members can respond to. So, here we’re trying to support the common and often very fluid social events we all have, such as the Japanese teens meeting on a Saturday afternoon or a group of friends deciding what movie to go to on a Friday night. The screen on the left shows a click into the page for an individual group. Again, all messages are distributed to the entire group, creating a group-wide conversation thread. The conversation brings together text message and photos, as well as group events and membership. Making a new group is as simple as giving it a name, selecting some people to invite – either from other slams or by phone number, and clicking ‘create’. And that really is the core of slam: helping people communicate, coordinate, and share their lives through simple group creation and conversation on mobile devices.