Practical Implementation of Hybrid Accuracy-Time Spectrum Sensing for Cognitive Radio Networks
Author
Ivanov, Antoni Stefkov
Term
4. term
Publication year
2016
Pages
72
Abstract
Kognitiv radio (CR) er trådløse enheder, der kan tilpasse deres frekvens og modulation i realtid. Ved at udnytte frekvensspektret mere effektivt kan de øge netværkets kapacitet og samtidig undgå forstyrrelser af licenserede "primære" brugere. En central funktion er spektrumsansning: at opdage, hvornår dele af spektret er ledige (såkaldte spektrums-huller). Hurtig ansning giver flere muligheder for at bruge ledige kanaler, men kan sænke nøjagtigheden; langsom ansning øger nøjagtigheden, men kan koste muligheder. Vi håndterer denne afvejning med en dobbelt tilgang. Først vurderes spektrumoptagelse med en fuzzy-logik-algoritme, der bruger den målte signal-til-støj-ratio (SNR) og modtagne signalniveauer som input. Denne beslutning sammenlignes med resultatet fra en statistisk metode. Dernæst vælger en afledt lukket formel en driftsindstilling, der balancerer ansningstid og nøjagtighed under de aktuelle forhold. Algoritmen er implementeret på en softwaredefineret radioplatform (USRP og GNU Radio) og evalueret i simulationer. Resultaterne viser, at tilgangen er effektiv i forhold til eksisterende metoder, og vi analyserer også ydeevnen af den praktiske implementering.
Cognitive radio (CR) refers to wireless devices that can adapt their operating frequency and modulation in real time. By using the spectrum more efficiently, they can increase network capacity while avoiding interference with licensed "primary" users. A key function is spectrum sensing: detecting when parts of the spectrum are free (so-called spectrum holes). Faster sensing creates more opportunities to use free channels but can reduce accuracy; slower sensing improves accuracy but may miss opportunities. We address this trade-off with a dual approach. First, the CR decides on spectrum occupancy using a fuzzy logic algorithm that takes the measured signal-to-noise ratio (SNR) and received signal levels as inputs. This decision is then compared with the outcome of a statistical method. Second, using a derived closed-form expression, the device selects an operating point that balances sensing time and accuracy for the current radio environment. We implemented the algorithm on a software-defined radio platform (USRP with GNU Radio) and evaluated it in simulations. The results indicate our approach is efficient relative to existing methods, and we also analyze the performance of the practical implementation.
[This abstract was generated with the help of AI]
Keywords
Documents