Optimistic transaction processing algorithms in pure-push and adaptive broadcast environments

Considering the properties of mobile computing environments, push-based data dissemination systems have lately attracted considerable interests. However, skewed access pattern of mobile clients makes average wait time worse and they may want to request the data object to the server explicitly through backchannel. We call the broadcast model supporting backchannel as adaptive broadcast. In this paper, we devise new algorithms for adaptive broadcast based on our previous works; that is, we divide data objects which the server maintains into push data and pull data. Clients have to explicitly request data objects in pull data. Maintaining transactional consistency in both pure-push and adaptive broadcast environments is our main concerns. We also evaluate the performance behavior through simulation study.