A Concurrency Control Scheme For Mobile Transactions In Broadcast Disk Environments

In wireless broadcast environments, conventional concurrency control schemes are not suitable for the mobile transactions due to the limited bandwidth in the uplink communication channel and the limited battery power of mobile clients. Existing methods perform poorly in broadcast disk environments where the data access patterns of mobile transactions are skewed. These problems will increase the average response time of the update mobile transactions and waste the uplink communication bandwidth.In order to solve these problems, we extend FBOCC (Forward and Backward Optimistic Concurrency Control) method by GMCCI (Group of Minor Cycle with Control Information) and CRB (Conflict Resolution by Backoff)to improve performance. We call the extension of FBOCC method as MVOCCBD (Multiversion Optimistic Concurrency Control for Broadcast Disks). Using the control information broadcast by the server, GMCCI method improves the effectiveness of the partial backward validation. When there are multiple update transactions competing for the same data to write, the CRB method computes and distributes the different restart times for conflicting transactions along broadcast cycles to decreases the possibility of further conflicts. The OCCBD method is a new optimal concurrency control scheme suitable for broadcast disk environments. The OCCBD method allows mobile transactions to access the more updated data, and reduces the average response times of mobile transactions through early aborts and restarts. The OCCBD method also reduces the amount of uplink communication bandwidth for the final validation of the update mobile transactions.We present an in-depth experimental analysis of our method by comparing it with the FBOCC method, we analyze the performance of the MVOCCBD and FBOCC methods by varying the number of read operations, skewness of data access patterns and the ratio of the number of broadcast data items and the number of mobile transactions. Our performance analysis shows that it significantly decreases the average response time, and the amount of uplink bandwidth over the FBOCC method.