We present lessons learned in developing cross platform multi-party team applications. Our apps include a range of communication and collaboration scenarios: document and content sharing in a team space, an agent-based meeting helper, phone number dialer via a voice-over-IP (VoIP) gateway, and multi-party call in peer-to-peer or client-server mode. We use web real-time communication (WebRTC) to enable the audio and video media paths in the apps. We use frameworks such as Chrome Apps and Apache Cordova to create apps that can be accessed from a browser, or installed on a desktop, mobile device, or wearable. The challenges and techniques described in our paper related to audio, video, network, power conservation and security are important to other developers building cross-platform apps involving WebRTC, VoIP and cloud services.

ConnectedSpaces is a novel federated collaboration environment that integrates different collaboration tools and paradigms for enterprise use. Collaborations are organized into containers called spaces. Recently we have extended ConnectedSpaces to enable spaces and their contents to have semantic p

Native multicast deployment is relatively slow and linked with a number of issues. However, there are a number of native multicast regions. Application Layer Multicast (ALM) can be used in areas of the network where there is no native multicast available. The SAM (Scalable Adaptive Multicast) Research group within the IRTF is investigating hybrid approaches to multicast, involving native deployments were available and ALM in other regions. SAM is using a P2P overlay to connect the nodes. Here we describe a protocol and API extensions to RELOAD for constructing Scalable Adaptive Multicast (SAM) sessions using hybrid combinations of ALM, native multicast, and multicast tunnels. The Automatic Multicast Tunneling (AMT) relay and gateway elements are employed for interoperation between native regions and ALM regions.

The development of hybrid multicast simulation models is required for analyzing proposed hybrid multicast architectures such as those from the IRTF Scalable Adaptive Multicast Research Group. However most network layer simulators don't scale to the number of nodes needed for analyzing large overlays, and most overlay simulators don't have multicast routing models needed for analyzing hybrid approaches. In this work we have extended the OverSim simulator and INET framework which run on OMNET++ to include a multi-destination multicast routing protocol (XCAST). This paper describes our implementation experience.

Online collaboration platforms such as Google Wave provide basic tools for collaboration and communication as well as an interface to include external applications. Separately, we have defined a collaboration platform called ConnectedSpaces which extends these basic capabilities towards enterprise use. Using the ConnectedSpaces model, here we categorize, illustrate, and analyze new types of feature interactions for collaboration platforms. Furthermore, we discuss handling feature interactions which are caused by using ConnectedSpaces as a communication endpoint with enterprise communication platforms. The work in this paper is novel as it is the first investigation into feature interactions with collaboration platforms. Our approach uses a runtime feature interaction technique which can cope with features provided by different organizations.

Structured peer-to-peer (P2P) overlays typically offer either multi-hop or one-hop performance. Here we present Chameleon, which offers a variable hop performance trading hop count with bandwidth requirement. Chameleon combines opportunistic and active routing table maintenance techniques into a novel two-tier approach which allows nodes to adapt to variations in their available bandwidth. Chameleon uses opportunistic routing table updates for lower bandwidth nodes and active stabilization for high bandwidth nodes, offering a performance from O(log n)-hop up to O(1)-hops. In doing so its nodes can operate successfully in large scale (mobile) networks where peers need to adapt to continuous changes in the available bandwidth. This paper presents a detailed design of Chameleon and simulation results verify its adaptable performance. We show that significant synergy between the two tiers improves the performance when compared with the standalone algorithms.

The Scalable Adaptive Multicast Research Group (SAM RG) within the Internet Research Task Force (IRTF) is chartered to define a hybrid multicast protocol that opportunistically combines overlay multicast protocols and native multicast protocols to create and maintain hybrid multicast trees. A design for SAM hybrid multicast trees has been proposed that leverages the AMT multicast tunneling protocol. Evaluating hybrid multicast protocols is difficult for two reasons. The simulation environment must combine both a scalable overlay and a detailed network layer that includes routers with native multicast support. Suitable metrics are needed to compare the tree quality with pure overlay multicast trees. This paper presents our approach to simulating and evaluating hybrid multicast protocols.

Location privacy management is an important requirement for location based services. Here we describe the design and implementation of location privacy enforcement mechanisms based on the IETF Geopriv specifications as part a location-based services platform that we previously developed. Our design involves a new and more efficient session-based access validation mechanism. In addition, we evaluate the expressiveness of Geopriv rules for describing a variety of privacy rules. We find a number of limitations and propose some methods for extending the rule notation for addressing these limitations. We conclude with an overview of our prototype implementation.

Temporal event correlation deals with removing temporally redundant events to produce synthesized events, simplifying downstream processing. In situation management applications and location-based services, it is necessary to incorporate spatial constraints into the correlation process. This requirement is difficult to meet because the traditional algorithms for spatial processing are not well-suited to spatial dynamics occurring in these applications. In this paper we formalize spatial-temporal event correlation compared to the functionality of temporal correlation, present an architecture for efficient implementation, analytically analyze the architecture's performance relative to two other designs, and describe a system design for efficiently supporting situation management applications and location-based services.

Several trends in SIP application server deployments in large scale telephony environments exacerbate the classic problem of feature interaction : use of distributed feature servers, mixing of legacy and green-field feature servers, and the co-existence of multiple third-party feature implementations provisioned in the same environment. Next-generation SIP application servers will include an application router (AR) to provide more application control over feature sequencing. As we discuss here, the AR can be augmented to incorporate feature interaction detection and resolution logic. We describe a novel design for run-time feature interaction detection and resolution in an environment of distributed feature servers using a SIP application server with application routing function, such as that defined in JSR 289. The approach is based on the algorithm of the Kolberg-Magill (K-M) method for feature interaction detection. Here we extend the notation of the algorithm to cover advanced call control services, enable the algorithm to work in topologies involving B2BUAs (Back-to-Back User Agents) and SBCs (session border controllers), and test the approach with a substantial feature set of 32 features found in large enterprise solutions.

SQL injection attacks continue to be a major problem for Web applications. We investigate design considerations for an application layer honeypot to attract and learn about SQL injection attacks. The honeypot responds with indications of vulnerability leading attackers ultimately to disinformation that could be useful to track them. The honeypot restricts attackers from escalating the attack to the operating system or launching attacks on other systems. The honeypot could emulate the appearance of common defenses against SQL injection in order to seem more genuine. Finally, we describe considerations to implement an experimental honeypot with honeyd.

Current P2P communication applications offer only a limited set of call processing services. However, this is likely to change due to services being integrated into the overlay and third party service provisioning. The particular focus of this paper is the problem of feature interactions. We illustrate some examples and show the applicability of previous methods of telephony feature interaction analysis to the peer-to-peer model.

A method is provided for parallelizing overlay operations in an overlay network. The method includes: identifying an overlay operation having a parallel messaging scheme; determining a destination address for each parallel message in the messaging scheme; encoding each destination address into a data packet; and transmitting the data packet over the overlay network using a multi-destination, multicast routing protocol.

We propose the integration of peer-to-peer network overlays with underlay networking in which multidestination multicast routing is available. Network overlay operations are parallelized by using multidestination multicast messages in the underlying network in place of same-source unicast messages. This mechanism is generally applicable to structured overlays including one-hop, multi-hop, and variablehop, and unstructured overlays. The main result is significant message reduction, which varies according to the overlay algorithm. 1.

One-hop overlays offer significant latency reduction compared to multi-hop overlays, but at a cost of increased maintenance traffic and routing table size. Recently the EDRA maintenance algorithm has been proposed and shown analytically to give low maintenance traffic compared to other active maintenance one-hop overlays. We identify issues with EDRA which lead to incorrect event detection and propagation. We define EDRA* fixing these issues and show through simulation EDRA* is stable, and compare the performance of EDRA* with the opportunistic maintenance mechanism of EpiChord.

Peer-to-peer overlays can be used for service discovery over a global network fabric. We describe and evaluate a new service indexing mechanism for in-band streaming services such as application relays, mixers, and media transcoders. For this type of service, the location of the service in the network and service admission status are key attributes. We describe and analyze a service indexing mechanism which uses network position-based advertisement. We show that this mechanism gives good service selection, provides a close to uniform advertisement distribution in the overlay, reduces message overhead, and exhibits acceptable stability and setup delay.

Structured peer to peer (P2P) overlay networks are becoming increasingly popular. Multi-hop systems achieve a successful lookup in O(log N) hops, whereas one-hop systems approach O(1) hops. Both approaches, but especially one-hop overlays suffer from a high number of identical messages being sent to a number of nodes on the overlay. Previous work showed that P2P networks benefit from the integration of the overlay network with the underlay network in which multi-destination multicast routing is available. This allows combining identical messages from the same source into joint multi-destination multicast messages to significantly reduce the number of messages. Our experimentation has centered around the one-hop EpiChord overlay. Here the problem is described using a Markov Model for more advanced analysis. The Markov Model is believed to be novel in two aspects: it is the first to investigate one-hop overlays and it is the first to study the performance of multi-destination multicast including the consideration of retransmissions of requests.

Peer-to-peer networking technologies are gaining increasing popularity for file sharing and media streaming applications. To support these set of applications, structured peer-to-peer (P2P) overlay networks may be employed to provide an overlay substrate. In structured P2P overlay networks, peer nodes, content objects and messages are often identified using a set of well defined identifications (Ids.) Objects are stored on peer nodes based on a set of predefined rules. Messages are routed towards the destination node, for example, the root node of the message Id, after an average of h routing hops. In the absence of faults, a high probability of successful message forwarding, i.e., a message is delivered correctly to the destination node, can be achieved even when a large fraction of the peer nodes crash. However, most current overlay networks are not secure. When one or more malicious nodes are presented in the overlay, they can prevent correct message delivery throughout the overlay. This paper studies attacks aimed at preventing correct message delivery in structured peer-to-peer overlays. Several different routing algorithms are analyzed to understand means to improve messaging security without significant additional cost.

Recently Wei and Montemayor [1] have demonstrated a Java Card™ smart card application for authenticating playback for DRM-protected media. In this scheme the Java Card stores the encrypted master key and communicates during media playback with a signed content handler to securely provide the segment keys to the content handler for decryption and rendering of the media. In this paper we summarize the Java Card-based design and analyze its vulnerabilities.

We have developed a technique for providing an instant messaging and presence (IMP) middleware that allows IMP protocol implementations to be dynamically loaded as a client application roams among various service provider networks with different IMP services. This technique addresses the problem of maintaining IMP sessions while roaming between domains with different IMP protocols. We describe the mechanism, relate it to network handoff mechanisms, and analyze a prototype implementation.

When participating in online communities and ad hoc interactions, peer trust is based on societal relationships where identity alone is typically insufficient. We propose a new method for automatic establishment of peer trust that is suitable for peer-topeer and ad-hoc networking applications which overcomes the shortcomings of existing approaches. This method includes a new type of property-based certificate which we call a Social Certificate, and a new approach for trust negotiation in which privacy is assured by isolating the negotiation using a Secure Trust Negotiation Agent. Each Secure agent acts as the negotiator for its peer, upholding the peer's trust negotiation policies while keeping shared information secret. We describe these new methods and the security technology underpinning the Secure Trust Negotiation Agent.

We describe the use of property-based peer trust to provide security in Sleeper, a new energy-efficient service discovery protocol which uses proxied advertisement and discovery to dynamically offload service discovery workload from power-limited devices. Property-based digital certificates have previously been used in trust-negotiation methods. In earlier work we developed a new method for secure and private peer-to-peer trust negotiation using property-based digital certificates and a secure trust negotiation agent. Here we describe how property-based peer trust can be used to secure the Sleeper service discovery protocol

We extend the BDI (belief desire, intention) agent model by enabling agent beliefs to be based on real-time situations that are generated by a situation management (SM) system. This has several advantages for multi-agent systems using BDI agents. First, because of the use of event correlation and data fusion techniques in situation management, agent platforms can support highly reactive distributed applications. Second, the situation manager provides a semantically rich representation of the world and can dynamically adapt its representation for situations over time. From the system architecture perspective, we discuss several alternatives for how existing BDI-capable agent platforms can incorporate this extension. These alternatives range from complete SM functionality in each agent to having shared SM functionality among multiple agents. We also consider environments where different agent platforms use our situation-based BDI (SBBDI) agent method and must interoperate. We include an example of an SBBDI agent system for homeland security threat assessment

Peer-to-peer technologies have emerged as an alternative to traditional client-server networking which can enable highly scalable end-to-end applications. The coupling of agent platforms with peer-to-peer overlay networks enables the integration of many agent platforms and enriches the semantic functionality of peer-to-peer systems. Here we present two important results based on combining recent work in FIPA agents and peer-to-peer service oriented middleware. First we describe the coupling between peer-to-peer overlay and the FIPA agent platform, including how the overlay can be used by APs to discover each other. Second we describe the relationship between agent services and peer services, including service discovery and mapping mechanisms. Finally we discuss the benefits of this integration.

The design of service discovery protocols which limit power consumption of devices has received little consideration but is an important problem because of the growing use of power-limited mobile networked devices. We describe Sleeper, a new discovery protocol which uses proxied advertisement and discovery to dynamically offload service discovery workload from power-limited devices. The advertisement structure supports several modes of service discovery including conventional service advertisements, meta discovery, taxonomic-based discovery, location-based discovery and federated discovery. In addition, Sleeper supports both push and pull modes of advertisement, but unlike other protocols uses service popularity to select push or pull modality. We describe an implementation of the new protocol and provide an analytical model for comparing response time to other push-based service discovery methods.

We analyze the delivery of an object-oriented multimedia content model, namely MHFG-5 (Multimedia Hypermedia Expert Group), for interactive multimedia applications in a DAVIC-compliant ADSL access network using low-cost memory-constrained set top units.We present detailed latency budgets for MPEG-2 DSM-CC-based transactions including STU configuration, engine download, and application and scene activation.We use simulation and analysis to assess the tradeoffs in memory management and application response time.We discuss our implementation of a subset of DSM-CC and MHEGJ, and latency measurements for presenting MHEG objects.The results are formulated as graphs which can be used by an application designer when encoding an MHEG application.