Towards User-centric Network Optimization Engine

The increasing diversity of applications as well as the introduction of new wireless last-mile techniques imposes additional challenges on both terminals and entire networks. A user- and application-centric approach to networking is expected result in better user satisfaction and allow for new market opportunities for network infrastructure providers. However, this architectural view leads to increased network complexity and, therefore, requires new autonomic network management and optimization techniques. In this article we focus on one aspect of this domain: network protocol stack optimization for wireless access networks. We provide a design of an autonomic engine that aims to select and adjust parameters of the communication stack in order to achieve high user satisfaction using minimal reconfiguration time. It relies on two basic constructs: utility-based performance assessment, and metaheuristic algorithms for fast optimization. We demonstrate that such an engine allows for efficient autonomous switching between different wireless access methods and consequently significantly improves the performance of the user device using cross-layer parameter optimization.