A prediction-based handover decision for wireless networks

Future wireless must be able to integrate most different heterogeneous IP-based networks and supply the user coordinate service without a broken connection within a diverse and overlay access network environment. One of the most challenges of wireless network integration is handover decision. To select the best network considering various network attributes we present a network evaluation method (NEM as a genetic criterion with user preference and usual estimation standards. With NEM, our handover decision can support not only vertical handover, also horizontal handover meanwhile by comparing cells available in current network. Furthermore, we propose a new prediction algorithm (DGRAD) to predict monitoring values of each factor prior to handover execution phase which addresses more accurate input parameters into NEM functions and provide reasonable final handover suggestions for users.