02.11.2012 Views

Handover mechanisms in next generation heterogeneous wireless ...

Handover mechanisms in next generation heterogeneous wireless ...

Handover mechanisms in next generation heterogeneous wireless ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

DYNAMIC NEIGHBOUR TRUST INFORMATION RETRIEVAL FOR GLOBAL ROAMING<br />

where R n represents the maximum residence time of the mobile user <strong>in</strong> the nPOA<br />

without handover. The location update cost and the power save mode registration cost<br />

largely depend on the signall<strong>in</strong>g process tak<strong>in</strong>g place <strong>in</strong> networks. The NTC update<br />

<strong>in</strong>terval may change over different access networks.<br />

Tak<strong>in</strong>g different approaches, the PSMN and AOMN methods are <strong>in</strong>tended to<br />

complement each other <strong>in</strong> facilitat<strong>in</strong>g network trust <strong>in</strong>formation distribution among<br />

neighbour<strong>in</strong>g networks. The AOMN that relies on actual handover event can produce<br />

the most accurate <strong>in</strong>formation about neighbour<strong>in</strong>g relationship. In the AOMN, the<br />

mobile users sacrifice their QoS and have the attached network obta<strong>in</strong> its neighbour<br />

network trust pattern <strong>in</strong> return. In contrast, the PSMN derives neighbour<strong>in</strong>g relationship<br />

from analys<strong>in</strong>g location update records. The mobile users selected for participat<strong>in</strong>g <strong>in</strong><br />

the PSMN sacrifice their power consumption <strong>in</strong>stead.<br />

4.5 Performance Evaluation<br />

A random walk model [78] as shown <strong>in</strong> Figure 4.7 is <strong>in</strong>troduced to simulate a multioperator<br />

environment. It is assumed that the radio coverage of every network (denoted<br />

as a cell <strong>in</strong> the model) is hexagonal shaped, and has six neighbour<strong>in</strong>g cells. A mobile<br />

user has equal probability of mov<strong>in</strong>g to any of the six neighbour<strong>in</strong>g networks. The cell<br />

residence time follows a Gamma distribution. The r<strong>in</strong>gs as marked <strong>in</strong> Figure 4.7 are<br />

used to group cells. Each r<strong>in</strong>g of cells is surrounded by the neighbour<strong>in</strong>g outer r<strong>in</strong>g of<br />

cells, and is also adjacent to an <strong>in</strong>ner r<strong>in</strong>g of cells. The outmost r<strong>in</strong>g of cells is referred<br />

to as boundary r<strong>in</strong>g or R<strong>in</strong>g n, and assumed to be <strong>in</strong> an absorb<strong>in</strong>g state. A mobile user<br />

enter<strong>in</strong>g the boundary r<strong>in</strong>g will rema<strong>in</strong> <strong>in</strong> that r<strong>in</strong>g at all times. This can be expla<strong>in</strong>ed by<br />

the nomadic behaviour of a mobile user. Moreover, it is assumed that each network<br />

belongs to an <strong>in</strong>dependent operator to make the handover task more challeng<strong>in</strong>g. Thus,<br />

whenever a mobile user leaves a cell, a NTC exchange between two adjacent cells along<br />

with a handover can be expected. The objective of the proposed NTC process is to have<br />

every network to obta<strong>in</strong> its neighbour network trust pattern. Accord<strong>in</strong>g to the procedure<br />

of Figure 4.4, this requires that a network has to exchange the NTC data with all of its 6<br />

neighbour<strong>in</strong>g networks at least once.<br />

- 69 -

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!