1 Enhanced Mobility Support for Roaming Users: Extending the IEEE Information Service WWIC 2010 Luleå, June 1-3, 2010 Karl Andersson*, Andrea G. Forte**, and Henning Schulzrinne** * Luleå University of Technology, Sweden ** Columbia University, New York, NY, USA
2 Outline Background Motivation The IEEE Standard Location-to-Service Translation Protocol (LoST) Our Proposal Experiments and Results Conclusions and Future Work
3 Background Networks: New radio access networks being deployed Heterogeneous networking environment with multiple radio access networks involved The vision is ”Always best connected” Devices: End-user devices with multiple radio access technologies Commercial software radios soon to come All IP networks Everything over IP IP over everything
Motivation for the work Information of interest to roaming users –List of available networks –Link Layer Information –Higher Layer Services Goal: To automate the user’s decision process –using static information, such as cost –using information from other users on the network in that particular geographic area Information Server 3GPP network WLAN WiMAX
The IEEE Standard for Media-independent Handover Services Media-independent handover (MIH) framework Covers handover initiation and preparation, but not execution of handovers Host-controlled and network- controlled handovers Main services: –Event services (MIES) –Command services (MICS) –Information services (MIIS) Interfaces defined by Service Access Points (SAPs) –MIH_SAP –MIH_LINK_SAP (linked to technology-specific primitives) –MIH_NET_SAP
Media-independent Information services (MIIS) Defines a set of information elements (IEs) –Structure –Representation Defines a query-response-based mechanism for information retrieval Information exchange handled through binary type-length-value (TLV) coded messages Complex queries supported through the Resource Description Framework (RDF) query language SPARQL
Information Elements (IEs) Types General type, indicating –network type, –operator identifier, or –service-provider identifier Access-network specific, providing specific information on –Quality of Service (QoS), –security characteristics, –revisions of current technology standards in use, –cost, and –roaming partners Point-of-Attachment (PoA)-specific information –MAC address of the PoA, –PoA’s geographical location, –data rates offered, and –channel information Vendor-specific
Location-to-Service-Translation Protocol (LoST) Original purpose: Map location information into Uniform Resource Locators (URLs) representing Public Safety Answering Points (PSAPs) for emergency calling Offers great flexibility A generalized Location-to-URL Mapping Architecture and Framework developed Methods for finding LoST servers described Ongoing work defines: –LoST extensions, –labels for common location-based services, and –a policy for defining new service-identifying labels
Location-to-Service-Translation Protocol (LoST) Distributed architecture Relies heavily on caching XML-based protocol, messages carried in HTTP messages LoST architecture consists of –Seekers –Resolvers –Forest guides –Authoritative Mapping Servers the core query type
10 Proposed Architecture Combine –Location-to-Service Translation (LoST) –IEEE Information Service in a three-layer model offering –a regional scope, –a service provider scope, and –an evaluator scope Collect information from users submitting reports containing information –coverage –quality of service parameters
11 Typical Scenario Client sends LoST request to the LoST resolver LoST server responds Client queries chosen IS server(s) <findService xmlns="urn:ietf:params:xml:ns:lost1" xmlns:p2=" serviceBoundary="value" recursive="true"> urn:service:communication.internet.80221
Contributing Information to the IS Server Selected users may submit reports at three levels –Reports on PoA –Reports containing QoS-related information –Error reports Include the location of the user. Allow the server to estimate the location of the access points or base stations <m:COS m:cos_id="1" m:cos_value="4" m:min_pk_tx_delay="100" m:avg_pk_tx_delay="150" m:max_pk_tx_delay="200" m:pk_delay_jitter="15" m:pk_loss_rate="200" />
Experiments and Results Prototype –Implementation of IEEE IS functionality in Python 3.0 –Columbia University LoST reference implementation Experiments –User traversing a few blocks on Manhattan Results –22 ms for multiple-location queries to IS server
Conclusions and Future Work A generic solution –Allows for competition at all levels –Scales well Improved quality and lower costs for end- users Need for scanning minimized Future work –Integration with software defined radio (SDR) -> “cellular on demand”