Quality of service (QoS) in internet cache coherence

Sustersic, John; Hurson, Ali R.

doi:10.1504/ijhpcn.2005.009418

Cited by 1 publication

(7 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The caching protocol discussed in Section 2.1.2 assumed a weak consistency for data objects, allowing the propagation of stale data. The problem of stale data can be addressed by a technique offered in [7]. While the offered technique was intended for Internet caching, it is just as appropriate in the domain of pervasive computing, as the service-oriented architecture of pervasive computing is similar to the service-oriented architecture of most websites.…”

Section: Imposing a Quality-of-service Metric On Cached Data Objectsmentioning

confidence: 99%

“…The decision for consistency or response time is not a binary decision, but rather, there is a continuous domain from which a user can select a point between strongest consistency (high response time) and weakest consistency (low response time) for a particular data object. In [7], this domain is known as a quality-of-service (QoS) domain, and it is illustrated in Figure 2. Values in this domain represent the minimum probability, between zero and one, that the data object received will reflect the latest write to this object.…”

Section: Trading Off Consistency For Response Timementioning

confidence: 99%

“…The cache replacement policy outlined in Section 2.1.7 keeps the freshest data objects in cache, and as such, mobile clients become effective servers of cached data to their fellow clients; however, clients in a pervasive environment should not be concerned with serving others, but rather, be concerned with serving themselves. The proposed policy [7] is intended for use in a web cache, where the purpose of the cache is to benefit the global population of clients. A pervasive environment is slightly different, in that the client itself is hosting the cache in order to serve itself first and other clients second.…”

Section: Alternative Cache Replacement Policymentioning

confidence: 99%

“…On the other hand, if a client can tolerate a stale data object, then a cached copy of the object can be received from one of many less-loaded, and possibly physically nearer, clients; thus, reducing the response time. In the technique introduced in [7] the user can decide, for each data object, whether consistency or response time is important.…”

Section: Trading Off Consistency For Response Timementioning

confidence: 99%

“…While the proposed method [7] achieves its goal by allowing users to customize the quality-of-service level of their requested data objects, there are alternative metrics for quantifying quality of service which may be more appropriate in a pervasive setting. The offered QoS metrics should give the pervasive user control and flexibility, such that the user attains high satisfaction and minimal distraction.…”

Section: Alternative Qos Metricsmentioning

confidence: 99%

See 4 more Smart Citations

Balancing Transparency, Efficiency and Security in Pervasive Systems

Wenstrom¹,

Bentivegna²,

Hurson³

2008

Advances in Computers

Self Cite

View full text Add to dashboard Cite

This chapter will survey pervasive computing with a look at how its constraint for transparency affects issues of resource management and security. The goal of pervasive computing is to render computing transparent, such that computing resources are ubiquitously offered to the user and services are proactively performed for a user without his or her intervention. The task of integrating computing infrastructure into everyday life without making it excessively invasive brings about tradeoffs between flexibility and robustness, efficiency and effectiveness, as well as autonomy and reliability. While not the primary goal of pervasive computing, efficiency in resource management should be considered in order to best utilize a limited set of resources (bandwidth, computing, etc.) as to avoid congestion and avoid creating a visible and distracting bottleneck in the eyes of the user. As solutions to efficiently manage the resources in a pervasive computing environment, three techniques will be examined: the distributed caching and sharing of data between mobile hosts, the broadcasting of services by public service providers, and the ability for mobile hosts to adaptively adjust the quality of offered services. Likewise, security is often an afterthought in many computing projects, though it should be of high consideration in a pervasive environment where users share public resources to operate on private data. Specifically, how can a user be authenticated in this environment with minimal or no user intervention? Solutions such as single sign-on via smartcards and biometrics will be examined to carry out authentication in a pervasive environment. As the feasibility of ubiquitous computing and its real potential for mass applications are still a matter of controversy, this chapter will look into the underlying issues of resource management and authentication to discover how these can be handled in a least invasive fashion. The discussion will be closed by an overview of the solutions proposed by current pervasive computing efforts, both in the area of generic platforms and for dedicated applications such as pervasive education and healthcare.

show abstract

Section: Imposing a Quality-of-service Metric On Cached Data Objectsmentioning

confidence: 99%

Section: Trading Off Consistency For Response Timementioning

confidence: 99%

Section: Alternative Cache Replacement Policymentioning

confidence: 99%

Section: Trading Off Consistency For Response Timementioning

confidence: 99%

Section: Alternative Qos Metricsmentioning

confidence: 99%

See 3 more Smart Citations