Solving Engineering Applications with LAMGAC over MPI-2

Traditionally, the Local Area Network (LAN) has been used for parallel programming. However, with recent advances in wireless communications, some authors think that the future of Wireless Local Area Networks (WLAN) depends on the demonstration of its application to real problems. In this paper we prove that wireless communications are suitable to perform parallel computing by presenting interesting experimental results for two engineering applications that follows the Master-Slave style programming with data dependencies among iterations. We consider a dynamic variation of slave processes as new portable nodes are added to the WLAN or detached from it.