Lead Investigators

Jeff Burke, Deborah Estrin, Fabian Wagmister

Overview

In this project, we are going to design and develop the architecture for a data-centric network fabric to support urban participatory sensing. The architecture will enable embedding network attested location and time context in sensor readings, and will provide context resolution control, “selective data sharing” / policy-based privacy, and related security mechanisms.

Approach

We have started working on the network architecture research. Our initial focus is on network-attested location and time context. As a first step, we are experimenting with the localization features provided by the CISCO hardware, and have designed the network layout as discussed in the next section.

The next step will be to focus on context operations and resolution control. We will develop applications for data collection and tagging, on mobile phones and server platforms. We will also work on privacy principals for selective sharing of collected data, and develop applications following these principals. The next phase will be deployment in end-user testbeds and finally a refinement of the architecture.

Systems/Experiments and Accomplishments

We will set up a Wi-Fi network using CISCO’s products and general purpose hardware at four locations: two on the UCLA campus, one at the Los Angeles State Historic Park, in its Interim Public Use (IPU) phase and one portable Wi-Fi system for temporary installations at other places. The Wi-Fi network will provide the infrastructure/testbed for designing and developing the proposed architecture.

Figure 1

We have designed the general layout for the network (Figure 2). We have been assigned four class C networks for the project and working on configuring the routers provided by CISCO for DHCP and VPN between the sites to hand out addresses. We will use five access points to cover each of the UCLA campus sites, six access points to cover the LA SHIP and four will be used for the portable system. This is more access points then necessary to cover each of these locations, but we have chosen to do this in order to get better localization information from the system, which is performed using RF Fingerprinting in the CISCO hardware. RF Fingerprinting uses a database of fingerprints of each co-ordinate on the floor-plan of the area being covered. A fingerprint is calculated using signal strength and taking into consideration factors including reflection, attenuation, and multi-paths. A client’s location is found by querying the fingerprint database with the signal strength reported by all the access-points that can see the client. 

At each of our locations only one access point will be connected to the router, this will be the root access point the others will be connected to the root access point in an ad-hoc mesh network. Each access point has two radios (802.11a and 802.11b). The 802.11a radio will be used as back-haul for the mesh network, and clients will connect to the 802.11b radio.

We are working with LA DWP to install the access points on street lampposts around the state historic park and UCLA facility management to install the access points on various buildings on campus to cover the locations.

Figure 2

We have done some experiments with the CISCO AP’s and localization hardware in a small test set up on the roof of our building. The figure below is a heat map from one of our tests. We have also done some QOS experiments.

Figure 3

We are also working on a poster for the IEEE Workshop on Local and Metropolitan Area Networks, which will describe the network and its uses in urban participatory sensing as well as using the network for community participation and interaction, for the communities in the neighborhood of the LA SHIP.

We are also setting up a WFS server for location information. We will also design a way to publish location information from the wireless network following the OpenGIS standards. This will enable setting up meshups with location data easy.

People

Faculty:

• Jeff Burke, UCLA, REMAP/CENS
• Deborah Estrin, UCLA,CENS
• Fabian Wagmister, UCLA, REMAP

Graduate Students:

• Vidyut Samanta, UCLA, REMAP/CENS

Undergraduates:

• Jason Ryder, UCLA

External Research Partnerships

Cisco (current), Nokia (current), CA State Parks (current)