Tabu Search Optimization of Magnetic Sensor Systems for Magnetocardiography

Abstract

This paper addresses the question of optimal sensor placement for magnetocardiographic field imaging. New magnetic sensor technologies allow less restrictive sensor positioning in this application. We develop a constraint framework for sensor positioning and use tabu search (TS) and particle swarm optimization (PSO) for finding an optimal set of sensors, whereby a new PSO algorithm is designed to fit the needs of our constraint framework. Numerical simulations are carried out with a three compartment boundary element torso model and a multi-dipole heart model. We find an optimal value of about 20 to 30 vectorial sensors and both TS and PSO yield similar sensor distributions. The comparison to sensors on regular grids shows that optimization of vectorial magnetic sensor setups may significantly improve reconstruction quality and that the number of sensors can be reduced

Sensor Scheduling With Time, Energy, and Communication Constraints

“…Establish the best MSE performance γ 0 by (17). (19) or (20) with the ρ provided and the additional linear equality constraints z ij = 1, ∀ (i, j) ∈ K, and z ij = 0, ∀ (i, j) ∈ N , to obtain the current estimate Z via a standard convex optimization solver [34].…”

“…5. If the IRL1 iterative process of (19) or (20) has converged (or 20 iterations have been completed since the last convergence) and the solution is not binary, i.e.,…”

“…First, we show the implicit energy constraint approach, i.e., without any explicit information about the energy profiles of the sensors our goal is to operate the sensor network over T time instances such that we do not activate the same sensors at each time. The 1 / ∞ style optimization problem (19) balances between the estimation accuracy of the sensor network and making sure that the sensing is distributed more evenly between the network's sensors. Results are shown in Fig.…”

“…Convex optimization techniques have also been proven useful for experimental design [7,Chapter 7.5] with 1 [8], [9] and reweighted 1 norm minimization [10] approaches such as [11], [12], [13], [14]. Earlier work used information theoretic approaches like mutual information maximization [15], [16] and cross entropy optimization [17] or other search heuristics like genetic algorithms [18], tabu search [19] and branch-and-bound methods [20] to solve the sensor placement problems. Several recent works have also considered nonlinear sensor networks [13], tracking applications [21], [22], distributed sensing scenarios [23], [24], correlated noise models [25], estimation of continuous variables [26].…”

Spintronic Memristor as Interface Between DNA and Solid State Devices

“…Generally, magnetic sensors can be utilized to detect the changes or disturbances of magnetic filed, i.e., the strength or direction of magnetic flux. For example, magnetic sensors with high sensitivity have been widely used in heart disease monitor by detecting the bio-magnetic signals from the heart (known as mag-netocardiography, or MCG) [2], [3]. The magnetic sensors in bio-medical applications are required to detect the low-field signals that are much weaker than the earth's magnetic field ( ) [4].…”