Markov Decision Process-Based Resource and Information Management for Sensor Networks

Akselrod, D.; Lang, Thomas; McDonald, Michael; Kirubarajan, Thiagalingam

doi:10.1007/978-3-642-01341-6_8

Cited by 2 publications

(1 citation statement)

References 65 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other work in Miller et al [2009], , Krishnamurthy and Djonin [2007], and Y. has modeled the SM problem as a Partially Observable MDP (POMDP), while work in Akselrod et al [2007] adopted hierarchal MDP in solving the SM problem. A comprehensive study of the formulation of the SM problem as MDP is offered in Akselrod et al [2010]. The main drawback of such stochastic approaches is that the proposed algorithms may suffer from combinatorial explosion when solving moderate to large-size problems.…”

Section: State-of-the-art Smmentioning

confidence: 99%

A Collaborative Energy-Aware Sensor Management System Using Team Theory

Hilal

Basir

2016

ACM Trans. Embed. Comput. Syst.

View full text Add to dashboard Cite

With limited battery supply, power is a scarce commodity in wireless sensor networks. Thus, to prolong the lifetime of the network, it is imperative that the sensor resources are managed effectively. This task is particularly challenging in heterogeneous sensor networks for which decisions and compromises regarding sensing strategies are to be made under time and resource constraints. In such networks, a sensor has to reason about its current state to take actions that are deemed appropriate with respect to its mission, its energy reserve, and the survivability of the overall network. Sensor Management controls and coordinates the use of the sensory suites in a manner that maximizes the success rate of the system in achieving its missions. This article focuses on formulating and developing an autonomous energy-aware sensor management system that strives to achieve network objectives while maximizing its lifetime. A team-theoretic formulation based on the Belief-Desire-Intention (BDI) model and the Joint Intention theory is proposed as a mechanism for effective and energy-aware collaborative decision-making. The proposed system models the collective behavior of the sensor nodes using the Joint Intention theory to enhance sensors' collaboration and success rate. Moreover, the BDI modeling of the sensor operation and reasoning allows a sensor node to adapt to the environment dynamics, situation-criticality level, and availability of its own resources. The simulation scenario selected in this work is the surveillance of the Waterloo International Airport. Various experiments are conducted to investigate the effect of varying the network size, number of threats, threat agility, environment dynamism, as well as tracking quality and energy consumption, on the performance of the proposed system. The experimental results demonstrate the merits of the proposed approach compared to the state-of-the-art centralized approach adapted from Atia et al. [2011] and the localized approach in Hilal and Basir [2015] in terms of energy consumption, adaptability, and network lifetime. The results show that the proposed approach has 12× less energy consumption than that of the popular centralized approach. ACM Reference Format:Allaa R. Hilal and Otman Basir. 2016. A collaborative energy-aware sensor management system using team theory.

show abstract

Section: State-of-the-art Smmentioning

confidence: 99%