Hierarchical feedback adaptation for real time sensor-based distributed applications

Cardei, Mihaela; Jha, Rakesh Mohan; Pavan, A.

doi:10.1109/isorc.2000.839527

Cited by 10 publications

(9 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then, the PSM sets again the input sensor period to the maximal current period of all stages, typically lower than maximum period from the contract. We have proved in [1] that the end-to-end latency decreases, and that after a finite number of frames the pipeline enters a region of stability where the end-to-end latency and the output frame rate are within the contracted region.…”

Section: Hierarchical Feedback Adaptation For Parallel Data-flow Applmentioning

confidence: 98%

“…We assume a sensor enters periodically data frames in the pipeline. Each frame is processed by a stage or a composite stage [1] (consisting of parallel strings of elementary stages) and then sent to the next stage. Such a pipeline application is depicted in Figure 7.…”

Section: The Higher-level Service Manager For Integrated Servicesmentioning

confidence: 99%

“…We have implemented an innovative and efficient hierarchical feedback adaptation mechanism for parallel pipeline applications [1]. It performs feedback adaptation at two levels in the SM hierarchy.…”

Section: Hierarchical Feedback Adaptation For Parallel Data-flow Applmentioning

confidence: 99%

“…The ATR pipeline contract requires an acceptable output frame period interval of [1,5] s, and a frame latency of 0.7-13 s. The seven ATR stages run at a variable workload between 0.02s and 1.5s and within the same period interval [1,5] s. We first present timing measurements for the feedback adaptation at the CPU SM and PSM SM level ( Figure 10). We measured the processing overhead of the feedback adaptation code (part 2 in Figure 10) and the time it takes the SM to react from the moment it receives the current QoS from the application until its adaptation command is enforced (part 2 + part 3).…”

Section: Performance Metrics and Evaluationmentioning

confidence: 99%

See 3 more Smart Citations

Hierarchical Architecture for Real-Time Adaptive Resource Management

Cardei¹,

Jha²,

Cardei³

et al. 2000

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. This paper presents the Real Time Adaptive Resource Management system (RTARM 1 ), developed at the Honeywell Technology Center. RTARM supports provision of integrated services for real-time distributed applications and offers management services for end-to-end QoS negotiation, QoS adaptation, real-time monitoring and hierarchical QoS feedback adaptation. In this paper, we focus on the hierarchical architecture of RTARM, its flexibility, internal mechanisms and protocols that enable management of resources for integrated services. The architecture extensibility is emphasized with the description of several service managers, including an object wrapper build around the NetEx real-time network resource management. We use practical experiments with a distributed Automatic Target Recognition application and a synthetic pipeline application to illustrate the impact of RTARM on the application behavior and to evaluate the system performance.

show abstract

Section: Hierarchical Feedback Adaptation For Parallel Data-flow Applmentioning

confidence: 98%

Section: The Higher-level Service Manager For Integrated Servicesmentioning

confidence: 99%

Section: Hierarchical Feedback Adaptation For Parallel Data-flow Applmentioning

confidence: 99%

Section: Performance Metrics and Evaluationmentioning

confidence: 99%

See 2 more Smart Citations

Hierarchical Architecture for Real-Time Adaptive Resource Management

Cardei¹,

Jha²,

Cardei³

et al. 2000

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Also, applications are externally adapted when they are allocated extra resources made available by other applications departing the system (QoS expansion). The second type of adaptation is called Feedback Adaptation [1] and is similar to closed loop control. The RM continuously monitors the application and uses the difference between the measured QoS and the contracted QoS to adjust the application operation parameters without changing the resource allocation or the negotiated QoS contract.…”

Section: Introductionmentioning

confidence: 99%

Resource Management for Ad-Hoc Wireless Networks with Cluster Organization

et al. 2004

View full text Add to dashboard Cite

Abstract. Boosted by technology advancements, government and commercial interest, ad-hoc wireless networks are emerging as a serious platform for distributed mission-critical applications. Guaranteeing QoS in this environment is a hard problem because several applications may share the same resources in the network, and mobile ad-hoc wireless networks (MANETs) typically exhibit high variability in network topology and communication quality. In this paper we introduce DYNAMIQUE, a resource management infrastructure for MANETs. We present a resource model for multi-application admission control that optimizes the application admission utility, defined as a combination of the QoS satisfaction ratio. A method based on external adaptation (shrinking QoS for existing applications and later QoS expansion) is introduced as a way to reduce computation complexity by reducing the search space. We designed an application admission protocol that uses a greedy heuristic to improve application utility. For this, the admission control considers network topology information from the routing layer. Specifically, the admission protocol takes benefit from a cluster network organization, as defined by ad-hoc routing protocols such as CBRP and LANMAR. Information on cluster membership and cluster head elections allows the admission protocol to minimize control signaling and to improve application quality by localizing task mapping.

show abstract

An open platform for reconfigurable control

et al. 2001

View full text Add to dashboard Cite

omplex dynamic systems such as aircraft, power systems, and telecommunication networks present major challenges to control systems designers. Both the military and civilian sectors of our economy are demanding new and highly sophisticated capabilities from these systems that traditional controls technology is not offering. Among these capabilities are the following: • Adaptability/dynamic reconfigurability: Large-scale engineered systems must have the ability to quickly and gracefully react to a changing environment or to changes in their own configuration without compromising their operational integrity. Extreme performance dynamic systems must be able to support online switching of algorithmic components and rapid redirection of the interconnections among them, as well as changing the priorities at which information is flowing. This ability is particularly useful for hybrid control systems. • Plug-and-play extensibility: New advances are continually being made in control algorithm design, communications and sensor technology, and high-performance hardware platforms. To take full advantage of these innovations as they become available, we must be able to insert new technology into the system architecture without redesigning the components already in the system. • Interoperability: Today's control systems operate in distributed, heterogeneous environments; that is, the software components may be running on different processors, using different programming languages, hardware platforms, and network protocols, often over wireless links. Real-time communication must be provided among these distributed components while satisfying stringent constraints on bandwidth, response time, and reliability. • Openness: Reconfigurability and component interchangeability require software architectures that are flexible and that support tools and algorithms from a variety of sources and domains. This requires a shift away from traditional control system implementation, which tends to be practiced with a particular ap-June 2001 IEEE Control Systems Magazine 49

show abstract

Hierarchical feedback adaptation for real time sensor-based distributed applications

Cited by 10 publications

References 12 publications

Hierarchical Architecture for Real-Time Adaptive Resource Management

Hierarchical Architecture for Real-Time Adaptive Resource Management

Resource Management for Ad-Hoc Wireless Networks with Cluster Organization

An open platform for reconfigurable control

Contact Info

Product

Resources

About