Adaptive Resource Management in Application to Surveillance Networks using Stochastic Methodologies

Vol.5 No.1

Year: 2010

Issue: July - September

Title: Adaptive Resource Management in Application to Surveillance Networks using Stochastic Methodologies   

Author Name: Charles C. Castello, Jeffrey Fan   

Synopsis:   

A wide range of applications have been developed in recent years pertaining to surveillance networks, which include defense, environmental protection, manufacturing, weather forecasting, and structural monitoring. The following research aims to present a novel method of resource allocation in surveillance networks (e.g. Wireless Sensor Networks) by using stochastic modeling techniques and reconfigurable System-on-a-Chip (SoC) systems. The basic idea behind the proposed framework is that a set amount of system resources (e.g. processing power, transmission bandwidth, system memory, etc.) can be dynamically allocated to different nodes within the system depending on the application or needs at any given time. The allocation of these resources is based on a stochastic approach which models resource demands by utilizing known and unknown random distributions. These distributions are analyzed using their associated polynomial expansions for known distributions and importance sampling for unknown distributions. An example of this would be using the Hermite Polynomial Chaos (PC) representation of random processes for Gaussian and log-normal distributions. The proposed framework results in intelligent surveillance networks with the ability to allocate resources in real-time.