Cube-Based Multitarget 3D Localization Using Bayesian Learning-Based Turbo Decoding in Wireless Sensor Networks

Penalized Maximum-Likelihood-Based Localization for Unknown Number of Targets Using WSNs: Terrestrial and Underwater Environments

“…Consequently, the optimal estimator can be constructed as ( 21) on page 10. It is worth noting that analytical solution for θ may not be feasible because the partial derivatives of the objective function in (21) are not solvable, and thus an optimization tool is required to find θ that maximizes L CD,opt (θ). Among many possible optimization tools candidates, the Genetic algorithm (GA) is adopted in this paper.…”

“…By comparing ( 28) with ( 21), it can be realized that the complexity of DTE location estimator is relatively low as the exponential terms have been totally avoided and the inner double-summation in (21) has been omitted. This simplification will be at the expense of the localization accuracy, however, the performance of both estimators is expected to converge at high SNR.…”

“…for which l ∈ {1, 2, • • • , L} and l ̸ = ĺ. However, for the sake of brevity, we provide the detailed analysis for the FIM of the optimal estimator with CD provided in (21) with Ξ i,k = h i,k . Similar steps can be followed to derive the FIM of the other estimators.…”

“…Similar steps can be followed to derive the FIM of the other estimators. Consequently, by referring to L CD,opt (θ) provided in (21), it can be realized that the first and second partial derivatives of L CD,opt (θ) with respect to χ l ∈ {x l , y l , z l } can be respectively given by…”