No abstract

Technology adoption research has a tradition of using and improving Davis' (1989) ''Technology Acceptance Model'' (TAM) and extended versions of it. This article suggests a break with this tradition by showing that the TAM is limited in its understanding of technology adoption. Two alternative approaches are proposed that focus on the role of knowledge and user-technology match, and on the role of temporary dynamical contexts in the process of adoption decision-making. Together with the TAM, both approaches were empirically tested and compared to the TAM by incorporating them in a questionnaire regarding the adoption intention of HDTV in the Netherlands. Results show that the constructs of both approaches show significant relations with the respondents' adoption intentions of HDTV and, together, offer a good alternative to the TAM. This result can be seen as a basis for more future research that uses technological and contextual factors as a starting point for adoption research. Using this starting point will contribute to a better understanding of future technology adoption processes.Whenever a new media technology is developed, scholars from various social science disciplines attempt to understand the factors that influence its possible success or failure. In the last 10 years, the digitization of television has been subject to studies of standardization (

The Session Initiation Protocol (SIP) is a popular application-level signaling protocol that is used for a wide variety of applications such as session control and mobility handling. In some of these applications, the exchange of SIP messages is time-critical, for instance when SIP is used to handle mobility for voice over IP sessions. SIP may however introduce significant delays when it runs on top of UDP over lossy (wireless) links. These delays are the result of the exponential back-off retransmission scheme that SIP uses to recover from packet loss, which has a default back-off time of half a second.In this paper, we empirically investigate the delay introduced by SIP when it runs on top of UDP over IEEE 802.11b links. We focus on the operation of SIP at the edge of an 802.11b cell (e.g., to update a mobile host's IP address after a handoff) as this is where SIP's retransmissions scheme is most likely to come into play. We experiment with a few 802.11 parameters that influence packet loss on the wireless link, specifically with different link-level retransmission thresholds, signal-to-noise-ratios (SNRs), and amounts of background traffic. We conduct these experiments in a controlled environment that is free from interfering 802.11 sources.Our results indicate that (1) SIP usually introduces little delay except for an SNR range of a few dBs at the very edge of an 802.11 cell in which the delay increases sharply, and (2) that a maximum of four 802.11 retransmissions suffices to limit the delay introduced by SIP retransmissions. The first result is of interest to developers of SIP applications who have to decide at which SNR to initiate a handoff to another network. The second result allows network providers to optimize their 802.11b networks for delay sensitive SIP applications.

Abstract.We consider the distribution of channels of live multimedia content (e.g., radio or TV broadcasts) via multiple content aggregators. In our work, an aggregator receives channels from content sources and redistributes them to a potentially large number of mobile hosts. Each aggregator can offer a channel in various configurations to cater for different wireless links, mobile hosts, and user preferences. As a result, a mobile host can generally choose from different configurations of the same channel offered by multiple alternative aggregators, which may be available through different interfaces (e.g., in a hotspot). A mobile host may need to handoff to another aggregator once it receives a channel. To prevent service disruption, a mobile host may for instance need to handoff to another aggregator when it leaves the subnets that make up its current aggregator's service area (e.g., a hotspot or a cellular network). In this paper, we present the design of a system that enables (multi-homed) mobile hosts to seamlessly handoff from one aggregator to another so that they can continue to receive a channel wherever they go. We concentrate on handoffs between aggregators as a result of a mobile host crossing a subnet boundary. As part of the system, we discuss a lightweight application-level protocol that enables mobile hosts to select the aggregator that provides the 'best' configuration of a channel. The protocol comes into play when a mobile host begins to receive a channel and when it crosses a subnet boundary while receiving the channel. We show how our protocol can be implemented using the standard IETF session control and description protocols SIP and SDP. The implementation combines SIP and SDP's offer-answer model in a novel way.

No abstract