DISTRIBUTED AUTONOMOUS WIRELESS CHANNEL ASSIGNMENT ALGORITHM WITH POWER-CONTROL

Local autonomous dynamic channel allocation (LADCA) including power co ntrol is essential to accommodating the anticipated explosion of deman d for wireless. We simulate call performance for users accessing chann els in a regular cellular array with a base located at the center of e ach hexagon, The computer model includes stochastic channel demand and a propagation environment characterized by attenuation with distance as well as shadow fading, Our study of LADCA shows that distributed po wer control and channel access can be combined in an access management policy that achieves satisfactory system capacity and provides desire d call performance, We report: LADCA/power control is observed to be s table alleviating a major concern about users unaware of the signal to interference problems their presence on a channel might cause to othe rs, We learn that local autonomy in seeking access does not trigger ca tastrophic event cycles, but, there are important deleterious effects (next bullet) of a manageable kind. There can be substantial inadverte nt dropping of calls in progress caused by originating calls, Despite great deference to ongoing calls it can turn out that nearly all unsuc cessful calls are unintentionally dropped calls not blocked calls, Als o, during time of high demand, oncoming calls probing for a usable cha nnel can be quite disruptive of channel quality even for those calls t hat are not dropped, An adequate number of channels appropriately limi ts these effects, We suggest that in future studies performance criter ia include a limit on call dropping along with one on call blocking. M odeling user time dynamics is essential, To neglect the dynamics of us ing, probing and switching channels and to instead represent call acce ptance or rejection as instantaneous, is much too optimistic, One miss es inadvertent call dropping and obtains unreliable estimates of the p robability of an unsuccessful call under a specified load, In a 36 cha nnel example, when the system is loaded so that an instantaneous model gives that 1% of the calls are unsuccessful, if we include time dynam ics, we get that 7% are unsuccessful. LADCA contrasts very favorably w ith fixed channel allocation (FCA) in a comparative example, In the ex ample, at a traffic load at which LADCA has an unsuccessful call proba bility of 1%, FCA has an unacceptable unsuccessful call probability of 15%.