On Building a Low Latency Network for Future Internet Services

Reducing network latency has always been a challenging problem due to the diversity of its potential causes. Though there have been many successful solutions, they are limited to certain applications or suffering from high resource consumption. In this paper, we propose a hybrid architecture which combines existing methods to meet the low latency requirements of future Internet services. We first make a comprehensive survey to the existing solutions and employ the classical queueing model to analyze the strengths and weaknesses of them. Then we present our hybrid solution, which makes a combination of P2P CDN, data redundancy, deterministic link and TCP optimization algorithms according to their characteristics. The related techniques we adopted work cooperatively in different scenarios to optimize network latency. Simulation results on two representative scenarios show that our architecture is powerful in offering low latency service, thus competent for future Internet applications.