Local status history based interpolation algorithm for networked racing game.
Date of Issue2009
School of Computer Engineering
Centre for Advanced Media Technology
This thesis addresses the problems of un-smooth vehicle movement and inconsistent views of the virtual world in network based racing games caused by network latency and bandwidth limitation. These problems significantly affect game experience of the players. We propose two methods to provide smooth vehicle movement without losing much consistency under most network conditions. The first method is called Local Status History Based Interpolation Algorithm (LSHBIA). To predict and interpolate the future movement of another player’s vehicle over the network, the LSHBIA uses the local status history of that vehicle in addition to the received update information from the network. The method can provide smooth vehicle movement in network based racing game, under good network conditions where the latency is less than 600ms for most of the time. However, under bad network conditions where the latency often exceeds 600ms, LSHBIA cannot maintain the smoothness of vehicle movement well. Another method, AI enhanced LSHBIA, is then proposed. The method uses AI techniques to control the movement of other players’ vehicles when the latency is greater than 600ms. Otherwise it will use the same approach in LSHBIA to handle the movement of vehicles. The method can still provide players with smooth vehicle movement in their network based racing games even when the network condition is bad.