SIQuA: server-aware image quality adaptation for optimizing server latency and capacity in wireless image data services [mobile radio]

This paper investigates the problem of providing guaranteed server-side latency while performing dynamic image adaptation. We introduce a server-aware image quality adaptation (SIQuA) technique, used to optimize the runtime image adaptation process so as to provide guaranteed server-side latency with a maximized quality of image service level. SIQuA utilizes dynamic server conditions (server traffic loads and server resource availability), and optimizes both the quality of image serviced and the server-side latency required for runtime image adaptation, simultaneously, through judicious tuning of image compression parameters. We demonstrate that SIQuA can significantly reduce the server-side latency with a minimal impact on the image quality, thereby enabling efficient wireless image services with greatly reduced overall service costs.