Adaptive handoffs through the estimation of fading parameters

Vijayan and Holtzman (1993) and Zhang and Holtzman (1994) outlined a model for analyzing handoff algorithms based on signal strength measurements. Delay in handoff and number of unnecessary handoffs were identified as performance measures. It was further shown that the signal averaging distance and hysteresis level could be chosen so that both performance measures were minimized. Vijayan and Holtzman (1993) pointed out that the optimal parameters were particularly sensitive to the shadow fading standard deviation. It was suggested that using more averaging and less hysteresis would provide some robustness. However, in order to detect jumps in signal strength (street corner effects), less averaging and more hysteresis is required. Thus, there is a conflict of requirements for robustness to shadow fading and jump detection. The present authors suggest estimating the shadow fading standard deviation as a solution to resolve the conflict. A simple scheme is outlined. It is based on using squared deviations of the received signal strength. Several sensitivity studies were performed and the results are included. The use of these estimates in conjunction with maximum Doppler frequency estimates to estimate speed in an adaptive handoff algorithm is specified.<
>