— We consider the sensing of the frequency spectrum based on energy detection for cognitive radios. It has been shown that the reliability of the sensing can be improved by using several cooperating cognitive radios. In such a network, the cognitive nodes exchange their individual sensing information to a coordinator node through specific control channels. The coordinator node then combines the received sensing informations in order to make a decision about the primary network presence. In this paper, we assess the impact of the noise coming from the quantization of the energy measure at each node and from the imperfect communication on the control channels. We also propose two complementary approaches to compensate for the noise: the system designer can firstly reduce the noise impact by using an energy coding technique that smartly shares out the bit error probabilities and secondly compensate for the remaining noise by designing a new soft metric fusion rule at the coordinator n...