Recently, there has been growing interest in transaction processing in mobile computing environments where communication between mobile clients and the server is asymmetric. Such communication renders conventional concurrency control protocols ineffective. Many studies have proposed new protocols to handle this problem. However, they all have an implicit assumption that the server is able to broadcast consistent data to the mobile clients in each broadcast cycle. To broadcast consistent data, a broadcast transaction at the server is needed to read the entire database in a serializable way. Conventional protocols do not work effectively because the "long-lived" broadcast transaction will encounter very high data conflicts with other concurrent update transactions at the server, leading to long delay and unpredictable completion time. This project aims at solving the problem of how to schedule the broadcast transaction such that it can read the most up-to-date database consistently and at the same time meet the timing constraint requirement.