Cloud storage services, because of their simplicity, have become more popular than other services for personal cloud usage. Consequently, there have been concentrated efforts to enhance their use cases, functionalities, reliability, performance, and speed in collaborative operations. These services are capable of handling lots of file-related operations among numerous users.

Criticizing traditional web application programming interface (API) methods, the authors propose a platform for effective automatic collaboration, file sharing, and updating across heterogeneous clouds. They discuss cloud proxies that facilitate intercloud file storage and transmission in an effective and reliable manner, with an acceptable response time for file updates; for example, for efficient access, leased virtual machines are physically near to the cloud server node. Moreover, a unified inter-proxy advanced protocol supervises the data sync operation and coordination. The platform is equipped with data flow scheduling algorithms that balance the related system overhead and timeliness of the data synchronization techniques, that is, deduplication, compression, and bundling, particularly in bursty workloads. Control mechanisms are devised to support data integrity and reduce redundant updates during file operations.

A detailed description of the main components of the platform introduces the CoCloud client, the control server, and the related cloud proxy. A proxy-per-cloud network architecture is built for each cloud server, while latency and bandwidth are approximated for cross-cloud file operations. The main components of the defined inter-proxy advanced transfer protocol are presented: adaptively chunked deduplication, which manages the transmission of just the modified parts of a file; wisely adjusted compression; and multi-level bundling, which affords multipath asynchronous file transmission between peer proxies. Other fruitful technical descriptions of the platform include the cross-cloud data transfer optimization algorithm, the inter-proxy dataflow scheduling algorithm, and file collaboration control mechanisms. Finally, the comparative results of a performance evaluation of the prototype version are discussed. It definitely demonstrates one step in the progress of cloud computing utilization.

This is a useful paper with a coherent tutorial theme. Thanks to the authors for choosing a subject that is not very complicated but important. They have developed the matter well, both technically and theoretically.