A least squares path-loss estimation approach to handover algorithms

In this paper, a new class of linear handover algorithms is introduced and analyzed for two conflicting performance measures, namely, the average number of outages and the average number of handovers, for a given trip of the mobile station (MS). The new algorithm is based on the least squares (LS) estimate of path-loss parameters of the various radio links, assuming the distances between the MS and surrounding base stations (BSs) are known. First, a simplified system scenario consisting of only two BSs in a lognormal fading environment is assumed. Performance measures are derived through a theoretical approach, along the trip of a mobile terminal from one BS to the other. It is seen that for an average number of 1.2 handovers (the minimum value is one, given by the ideal reference case), the LS algorithm reduces the average number of outages by more than 30% with respect to the classical averaging algorithm. Furthermore, the above performance measures improve for longer estimation windows: both the estimation of the mobile speed and the window length tradeoff of averaging algorithms do not apply to the LS method. By using an extended version of the basic algorithm, these results were also observed in simulations of a more general system scenario, consisting of several BS's and a mobile terminal moving along a generic path.