Distributed processing communication scheme for real-time applications considering admissible delay

This paper proposes a distributed processing communication scheme for a real-time network application that provides interactive services for multiple users. In the proposed scheme, the application is processed on a data processing function in the distributed servers. The distributed servers are selected to maximize the number of in-service users as the first priority and to minimize the delay time at in-service users as the second priority. In this scenario such as interactive services, the application must work under the condition that an event occurrence order at the actual time is guaranteed. This leads to delay time on users' communication at wide-area networks. In the proposed scheme, an event occurrence order at the actual time is reproduced in the virtual time defined at each server. We formulate an optimization problem to select a suitable server having the data processing function as an integer linear programming (ILP) problem. The ILP minimizes the number of excluded users because of violation of admissible delay as the first priority and minimizes the delay time for determining the virtual time as the second priority. The numerical results show that the delay time of the proposed scheme is reduced up to 50 percent than that of the conventional scheme.