Đang chuẩn bị liên kết để tải về tài liệu:
Lecture Notes in Computer Science- P61

Lecture Notes in Computer Science- P61:This year, we received about 170 submissions to ICWL 2008. There were a total of 52 full papers, representing an acceptance rate of about 30%, plus one invited paper accepted for inclusion in this LNCS proceedings. The authors of these accepted papers | A Virtual Laboratory Platform Based on Integration of Java and Matlab 289 To realize these functions of the JMPS we utilize Java Socket and multi-thread technology to solve these problems which are involved in listening to remote clients requests and multi-users access. A TCP IP Socket connection is implemented between Bean component and JMPS. By using the connection the results and the remote clients requests can be transmitted in stream format. Once one request is accepted successfully a new thread will be started to maintain the connection between the server and the client. Bean component holds a client socket and JMPS holds a server socket correspondingly. The definition of JMPS is shown in Fig.3. public class JMPS extends Thread public void run get the data stream after detecting request from user streams String getstreams CallMatlah cm new CallMatlab build object parse I streams get sub string and value array result cm.CallFunctionI sub value call Matlab and get result printstreams output the data to clientJ s Bean closeConnection cm.closeEngine close Matlab engine Fig. 3. Definition of JMPS Native methods make it possible to use Matlab s computational engine inside the user s own Java applications applets and servlets as JMatLink connects Java and Matlab. JMatLink uses Java JNI technique to locate and call non-java methods. First of all it compiles C C source files and then changes these files into Dynamic Link Library DLL and imports Matlab library functions so that these functions can be easily called. As a result Matlab engine functions can be used from a Java program by implementing engine function calls in C language native methods. Detailed process of communication between JMPS and Matlab is shown in Fig. 4. JMPS start a Fig. 4. Communication between JMPS and Matlab 290 Y. Sheng et al. thread when detecting a request from a component. Then the thread analyses the data stream from the component to obtain data to be processed and component information .