AAU Student Projects - visit Aalborg University's student projects portal
A master's thesis from Aalborg University
Book cover


Cross-Layer Optimization of Message Broadcast in MANETs

Authors

;

Term

10. term

Publication year

2007

Pages

186

Abstract

Mobile ad hoc-netværk (MANETs) skal kunne udsende beskeder pålideligt og hurtigt trods skiftende trådløse forhold. Dette projekt udvikler en modelbaseret cross-layer-optimeringsmetode, der justerer parametre på tværs af fysisk lag og MAC-lag for en distribueret bildata-tjeneste med stramme krav til beskedfejl-sandsynlighed (Pmf) og ende-til-ende-forsinkelse (De2e). Arbejdet kombinerer modeller af 802.11b’s fysiske kanal (1 og 2 Mbit/s med hensyntagen til path-loss og multipath-fading), medieadgang (ikke-persistent CSMA med skjulte noder og Bianchis backoff-model) samt flooding-broadcast, og bruger dem til at vælge sendeeffekt, PHY-mode og FEC-koderate. Tilgangen evalueres i ns-2 (v2.29) med et realistisk udvidet PHY-lag. I en høj-belastningssituation peger modellen på en optimal transmissionsafstand på 210 m, mens simulationer indikerer optima ved 160 m og 200 m. Resultaterne viser, hvordan cross-layer-parametervalg kan afveje pålidelighed og forsinkelse for at forbedre resiliens i MANET-broadcast ved brug af COTS-teknologier.

Mobile ad hoc networks (MANETs) must broadcast messages reliably and quickly despite variable wireless conditions. This thesis develops a model-based cross-layer optimization method that tunes parameters across the physical and MAC layers for a distributed car-data service with strict targets on message failure probability (Pmf) and end-to-end delay (De2e). The work combines models of the 802.11b physical channel (1 and 2 Mbit/s accounting for path-loss and multipath fading), medium access (non-persistent CSMA with hidden nodes and the Bianchi backoff model), and flooding-based broadcast, and uses them to select transmission power, PHY mode, and FEC code rate. The approach is evaluated in ns-2 (v2.29) with an enhanced, realistic PHY layer. In a high-contention scenario, the model suggests an optimal transmission distance of 210 m, while simulations indicate optima at 160 m and 200 m. The results illustrate how cross-layer parameter selection can trade off reliability and delay to improve resilience in MANET broadcasting using commercial off-the-shelf technologies.

[This summary has been generated with the help of AI directly from the project (PDF)]