Prestaciones de las Redes Ad Hoc Inalámbricas: Teoría a través de Capas

Abstract

The current PhD Thesis, carried out in the scope of Wireless Ad Hoc Networks, is an analytic study of the MHWN (Multihop Wireless Network) performance, mainly in terms of the end-to-end BER (Bit Error Rate). This is a crosslayer analysis which takes into account the Physical, Link and Network layers, included in the OSI (Open System Interconnection) model. The end-to-end BER is a fundamental parameter for measuring of the MHWN performance; however, it is still undetermined in an analytical way when the effects from different layers interact and, so, this is the main motivation for this Thesis.

Our study begins with the characterization, in an accurate analytical way, of the end-to-end BER with the only consideration of the Physical layer effects. Next, not taking into account interferences (Link layer) in the Ad Hoc Network yet, we incorporate the effects of the routing protocol (Network layer) through a directivity analytical model.

To complete our multilayer analysis, we propose an analytical model to calculate the multihop interference power in MHWN. Using this multilayer model, which takes into account the interactions among the previously mentioned layers, we deduce expressions for the end-to-end BER in presence of multihop interference, according to the number of hops which make up the route, as well as the Euclidean distance between its extremes. Additionally, to obtain such expressions, we formulate the models which relate both scales (number of hops and distance between extremes) of the MHWN routes.

Once this analysis of performance in ``conventional'' communications is carried out, we study two novel scopes in communication networks with the objective of decreasing the end-to-end BER in MHWN: Cooperative Communication and Network Coding. With reference to Cooperative Communication, we propose two distributed DD (Double Differential) coding schemes to solve the problem of performance decrease with the appearance of CFO (Carrier Frequency Offset) among cooperative nodes and the destination one. Regarding Network Coding, we propose a coding algorithm adequate for dense and multihop WSN (Wireless Sensor Network), which decreases BER through a distributed coding using LDPC (Low Density Parity Check) codes.