Towards a knowledge-based intelligent handover in heterogeneous wireless networks

This work presents a knowledge-based handover anticipation scheme. We propose to construct a knowledge plane using intelligent agents. Agents are delegated to information collection for nourishing the knowledge plane and also, handover decision and a proper network selection. We use the parameters of RSS, bandwidth, cost, user preferences, and mobility pattern of the mobile user. We estimate the future attachment point for the mobile user and anticipate the handover. Finally, we analyze the performance of our approach for (1) assuring the quality of service for different types of ongoing sessions during the handover, (2) reducing handover latency, (3) reducing the ping-pong effect and (4) signaling overhead.