چكيده به لاتين
Abstract:
Nowadays, growing interest in civilian and military applications of Wireless Sensor Networks (WSNs) has resulted in focusing of many researchers for improving performance of these networks via annexing with some other technologies. For an example, Cognitive WSN (C-WSN) as the next generation of WSNs opportunistically utilizes unused spectrum based on Cognitive radio (CR) method. Therefore, applicable high performance spectrum sensing is an important open problem in C-WSNs. Many references proposed methods based on simple low performance Energy Detector (ED) as the suitable method for spectrum sensing in C-WSNs, whereas we believe such low performance method wastes the precious resources of sensor nodes. Therefore, we re-described the exact problem of spectrum sensing in C-WSNs and proposed united spectrum sensing model against the conventional cooperative spectrum sensing in CR networks for achieving high performance spectrum sensing in cognitive WSNs. Since, there are no appropriate criteria to compare proposed algorithms or investigate whether or not an algorithm is suitable for a constraint resource network such as C-WSNs, we proposed an appropriate set of criteria, related boundaries and a comparative method for evaluating algorithms in C-WSNs. Finally, we proposed and evaluated United Multi Taper Method (UMTM) as a high performance spectrum sensing in C-WSN. The method is applicable on 30 – 80% of conventional WSN nodes against ED with 20 – 70% applicability, if we dedicate 1 – 10% of node's total memory for spectrum sensing algorithm. Also, UMTM saves SNR about 14 dB for PD > 90%, versus ED spectrum sensing method.
