Trilateration-based localization schemes typically extract the needed information, such as distances between transmitters and receivers, from noisy received signal strength (RSS) measurements specific to the employed technologies. Such applications employ non-stochastic methods without appropriately targeting noise. This paper proposes employing a recent optimal stochastic Newton-Raphson (NR) algorithm with measurement noise rejection capability for a class of localization applications. In order to show the effectiveness of this algorithm, we numerically consider an anchor-free localization problem with random initial guess. Numerical results show that the proposed recursive algorithm provides significant improvement over the traditional NR method.