Presentations

Apple released the most popular smartphone in the world, up to this point, the iPhone, in 2007. The iPhone has a variety of capabilities that would be useful in an emergency situation, such as location conveyance and camera usage. Today, 9-1-1 and E9-1-1 calls do not take the capabilities of any smartphone available today to the fullest extent.

In this talk, we discuss and demonstrate a system called V911, which enables callers to transmit relevant context and information to a public safety answering point (PSAP), such as location and video streaming. PSAP operators can pull up related information, such as information about the person and nearby first responder assets. Operators then forward relevant context, information, and streams to first responders that will be going to the scene.

The underlying system behind V911, Rover, enables standards to be easily integrated with other ones. We describe how HTTP Enabled Location Delivery (HELD) (as implemented by Raytheon/BBN Technologies) and Location in SIP/IP Core (LOCSIP) (as implemented by Telcordia Technologies) can interoperate with each other through one or more Rover ecosystems.

Distributed applications have ubiquitous in the computer world. Individual components that perform separate tasks cooperate to build complete systems that span multiple organizations. For example, a purchase from an online retailer may involve remote transactions between the storefront web application, their inventory management system, credit card merchants responsible for charging and crediting accounts, and a shipping carrier. In order to provide high performance and availability through redundancy, service-oriented providers (such as merchants and shipping companies) often deploy multiple computers providing the same service throughout the network.

Load balancing mechanisms exist to prevent the saturation of any single service by distributing incoming requests evenly to copies of the same service. Decisions, however, as to where the requests should be sent are only based on traffic volume. The nature and the content of the traffic are never taken into consideration to make a more intelligible load balancing or redirection decision.

This research investigates a universal mechanism to facilitate the construction of a router that extracts statistical information from client transactions. Based on this information, the router can alert deployment planning systems responsible for maintaining service quality through adjustment of routing configuration. This research exploits the Global XML Web Services Architecture (GXA) standards that expose the contents of transactions to intermediate routers. My contribution is a prototype implementation of an extended GXA router that supports the insertion of data collection modules. Data collection modules can be configured to generate profiles of GXA transactions characteristics. They can also transmit alerts to deployment managers when presented criteria that indicate a potential need for reconfiguration. These alerts can thereby be used to trigger the computation and deployment of a new system configuration in response to current system conditions.