Data Transmission in Mobile Edge Networks: Whether and Where to Compress?

Future multimedia communication systems aim to support high-rate data transmission and deliver millisecond-scale latency performance. To achieve this goal, this letter revisits the multimedia compression strategy in the framework of mobile edge computing from the perspective of latency minimization. The uniqueness of this letter is that the delay caused by both data compression/decompression and data transmission are considered. According to where to compress and decompress data, three different schemes are investigated and the end-to-end latency of each scheme is analyzed. We first present some comparative theoretical results on the latency performance of different schemes for given compression ratio. Then, we prove that the optimal compression ratio to minimize the end-to-end latency has a binary structure. This initial work reveals that it is optimal to transmit without compression in the case that the communication capacity is sufficiently large.