12.07.2015 Views

Wireless Ad Hoc and Sensor Networks

Wireless Ad Hoc and Sensor Networks

Wireless Ad Hoc and Sensor Networks

SHOW MORE
SHOW LESS
  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Distributed Power Control of <strong>Wireless</strong> Cellular <strong>and</strong> Peer-to-Peer <strong>Networks</strong> 215for every i = 1, 2, 3, …, n. The lower threshold value for all links can be takenequal to γ for convenience, reflecting a certain QoS the link has to maintainto operate properly. An upper SIR limit is also set, to decrease the interferencedue to its transmitter power at other receiver nodes. In the literature, severalDPC schemes have been proposed. The most recent DPC work includesBambos et al. (2000), CSOPC by Jantti <strong>and</strong> Kim (2000), SSCD <strong>and</strong> optimalby Jagannathan et al. (2002), Dontula <strong>and</strong> Jagannathan (2004), respectively,<strong>and</strong> they are discussed in the previous section. This algorithm is given next.5.4.2.1 Error Dynamics in SIRIn the previous DPC schemes presented in Section 5.2 <strong>and</strong> Section 5.3,only path loss uncertainty is considered. As a result, during fading conditions,high value of outage probability, as illustrated using simulations,is observed. The work, presented in this section, is aimed at demonstratingthe performance in the presence of several channel uncertainties.In the time domain, however, the channel is time varying when channeluncertainties are considered <strong>and</strong> therefore gij( t)is not a constant. In Lee<strong>and</strong> Park (2002), a new DPC algorithm is presented where gii( t)is treatedas a time-varying function due to Rayleigh fading by assuming that theinterference Ii( t)is held constant. Because this is a strong assumption, inthis paper, a novel DPC scheme is presented (Jagannathan et al. 2006)where gii( t)<strong>and</strong> the interference Ii( t)are time varying, <strong>and</strong> channel uncertaintiesare considered for all the mobile users. This relaxes the assumptionof other works where both gij( t)<strong>and</strong> Pt j( ) are held at a fixed value.Considering SIR from Equation 5.47 where the power attenuationgij( t)is taken to follow the time-varying nature of the channel <strong>and</strong> differentiatingEquation 5.47 to get( gii( t) Pi ( t)) ′ I( t) − ( gii( t) Pi( t)) I(t)′Ri()t ′ =I () ti 2(5.49)where Ri( t)′ is the derivative of Ri( t), <strong>and</strong> Ii( t)′ is the derivative of Ii( t).To transform the differential equation into the discrete timexl ( + 1) −xl(domain, x′() t is expressed using Euler’s formula as ) , where T is theTsampling interval. Equation 5.49 can be expressed in discrete time as( gii( l) Pi ( l)) ′ I( l) − ( gii( l) Pi( l)) I(l)′Ri()l ′ =2I () li⎡1⎛⎞′⎤⎢=() ( g ′iiI l( l ) Pi ( l )) I ( l ) + ( gii( l )⎥P2 ⎢′i ()) l Ii() l − ( gii() l Pi()) l ∑ gij() l Pj() l + η i () t⎥i ⎢⎝⎜j≠i⎠⎟⎥⎣⎦(5.50)

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

Saved successfully!

Ooh no, something went wrong!