One previous paradigm was based on sensor positioning selection to allow key variable observability. Later, the problem was posed as a minimum cost problem subject to sensor network performance (precision and gross error robustness first and accuracy as a replacement later). In a previous paper (Nguyen and Bagajewicz, submitted for publication), we presented a new paradigm in sensor network design, one that maximizes value, defined as economic value of information minus its cost. We presented two methods to solve the problem and we concluded that their performance was acceptable for small and medium size problems, but not for large problems. In this paper we use parallel computing approaches for large problems. (C) 2010 Elsevier Ltd. All rights reserved.