EL DISEÑO DE PATRONES DE PILOTOS PARA LA ESTIMACIÓN DE CANAL PARA RADIOS COGNITIVOS

Abstract

Contemporary wireless communications systems that furnish services to multiple usersatthesame timearecomplex. Theenvironmentwherethey operatebecomesdifficulttoadministerwiththeaggregatednoise,anduser mobility while balancing the system’s limited resources, i.e., the electromagnetic spectrum space. Asnewmanifoldcommunicationstandardsarisetocopewiththeuser’s and industry’s needs such as high mobility, long battery duration, multirateapplications,andmulti-standardmanagementtheneedtomakeuseof more intelligent systems is inevitable. Hence, developing highly flexible communication algorithms across the whole transceiver design is desired to fulfill those needs. The Cognitive Radios (CR’s) can help to alleviate these problems with its high capacity to adapt and handle access to licensed and unlicensed (secondary) users. To be efficient in the adaptation to different transmission scenarios, sensing the spectrum together with channel estimation are the cornerstones to build up valuable communicationslinks. Furthermore,theCR’sheavilydependsontheobservation of the channel state quality to decide, learn and act. Both, sensing and channel estimation are some of the most challenging areas and some of the most important to be investigated. Traditional approaches to performing spectrum sensing and channel estimation on quasi-static scenarios have been developed throughout digital communications history ending-up on good results to improve the recovery accuracy of the original signal at the receiver end, over several scenario conditions. Interestingly, in the case of CR’s radios, they have to be aware of their location and supervise their electromagnetic environment to be able to adapt the CR’s transmission techniques optimally. In thisfashion,thechannelstateinformation(CSI)feedbackhasasubstantial impact on the complete communications system. With the next step in radio evolution on the horizon, CR’s will create devices to help to make the highest efficient use of the radio spectrum, therefore designing and implementing the best in class in channel estimation algorithms that contribute to the cognitive radio’s optimal operative state will be needed. This thesis presents a proposal on the design of the pilot allocation for the discrete channel estimation over transmission blocks on a multipleuser and frequency selective channel. The adopted the OFDM modulation scheme allows to cope with signal errors due to the high mobility, inter-symbol interference (ISI) and inter-carrier interference (ICI) effects. Furthermore,theproposedmethoddesignsthepilotpatternsequencewith the use of wavelets as a technique to decompose the spectrum and identify most affected bands (by noise and interference) and place more pilots tonesinthosesubcarrierstoimproveaccuracyontheCSI.Therefore,minimizingtheerrorontheestimatedchannel. Havingtheimprovedpilottone sequence information at the receiver a less compute demanding method can be devised. It is demonstrated throughout statistical simulations the effectiveness of the proposed algorithm, such that the SNR loss on BER is minimum when compared with a system that has complete knowledge of the CSI.