Accesso e Instradamento in Reti Radio Cognitive

The always-increasing need of more flexible link for broadband connectivity in mobile conditions has led to a high-level radio technology evolution. Radio environment has to be shared and new strategies for improving the usage efficiency need to be developed in order to allow coexistence between different radio systems. In the past, the spectrum has been assigned to particular services in a fixed way, for example for radio mobile or digital television broadcasting systems. This unyielding use of spectrum resources has led to a really low efficiency; in fact, FCC has demonstrated that radio frequencies are used for only the 30% of their possibilities. Basing on these aspects the idea of new technologies allowing coexistence between different radio systems using the same frequency band has risen up; the Cognitive Radio is one of the most important among them. A cognitive radio is a smart device, which independently or thanks to cooperation with others nodes (the so-called secondary users), is able to exploit unused or underused spectrum fractions; these portions are generally assigned to primary systems, but can be used by cognitive systems in order to provide services which traditionally were not allowed in that bands. On the other hand, secondary nodes need to protect primary users from interference coming from their transmissions, while primary receivers do not become aware of cognitive nodes presence. A secondary node can adopt different strategies; first of sensing algorithm, together with geolocation database approach, allows cognitive devices to detect available spectrum portions for secondary transmissions; at the same time it is really important to forecast the effects of these transmissions on the primary receivers signal quality. The target of this work was the design of an interference prevision model, which can allow to establish the interference produced by a secondary system against a primary DVB-T system. The last one has many different distinctive features: first of all, the DVB-T receviers are passive and it is not possible to establish their position in the scenario. Moreover the primary traffic model is not link ‘burst model’, therefore there are not any temporal intervals in which the spectrum can be considered as free. Many different steps have been developed in order to reach the final model: in the first phase the effect of a single secondary transmission on a primary system with only one transmitter and many different receivers has been evaluated and a closed-form analytical expression for primary outage probability has been provided. Then the scenario has been extended to a secondary network instead of a single cognitive node and we obtained a closed-form analytical expression for the signal-tointerference ratio CDF; it was not possible to determined a closed form expression for the outage probability in this case, because of the too high complexity of the mathematical approach. For this reason the scenario has been modeled thanks to MatLAB simulations, which allowed also to simulate more complex scenarios with more than one primary transmitter and a secondary network. We also introduced a simple routing rule, in order to demonstrate the effect of our prevision model on the path selection strategy. Finally a real test-bed and the experimentation on WiFi-based system, operating in UHF band has been described. This activity was conducted in collaboration with Politecnico di Torino and CSP and has marked some advantages coming from the usage of this system for providing broadband connectivity in Digital Divided zones, instead of using traditional 5 GHz, HyperLAN strategies.