Compression's effect on end-to-end latency in file upload systems that utilize encryption

Sammanfattning: Encryption is the process of obfuscating data to restrict access to it while allowing it to be returned to its original non-obfuscated form through decryption. This process is increasingly used within web-based systems to secure data. When encryption is introduced in a system, the overall end-to-end latency of the system typically increases, and this increase depends on the size of the input data given to the encryption algorithm. Arguably, the latency introduced by encryption can be reduced if data sizes can be reduced before encryption. Lossless compression is the process of taking some data and reducing its overall data footprint. Introducing such a process within a web-based system that uses encryption could have the potential of reducing overall end-to-end latency within the system, both by reducing encryption time and data transfer time. This thesis evaluates whether the introduction of compression can reduce end-to-end latency in a web-based file upload system that encrypts the received files before storage. A series of experiments have been performed on a created file upload system where compression has been implemented to be used before upload and encryption. The results of these experiments show that compression can reduce end-to-end latency within web-based file upload systems that use encryption by approximately 39% for upload scenarios and approximately 49% for download scenarios when running in a system configuration with network latency.