Cognitive radio-based Long Term Evolution (LTE) networks enjoy the interesting features of cognitive Radios (CR), such as learning and reconfigurability, which enable them to perform channel switching, hoping to find a channel with higher quality. This process of searching and sensing other channels is a restoration (recovery) process where its objective is to find the best channel in the shortest time. We propose in this paper a history-aware greedy restoration scheme which is triggered periodically, not only when the quality of the current operating channel of the user goes below a threshold. Based on the state of the current channel, our scheme calculates the optimal number of channels to be sensed in this restoration period and this number is dynamically updated based on the sensing results. Intrinsic features of learning and history-awareness of CRs are used to create the list of the channels to be sensed based on the channels' background and historical information in order to improve the restoration mechanism by providing a shorter restoration time or a restored channel with a higher quality. We show that it provides improvements for the CR-based LTE network's throughput, compared to the other restoration schemes which work based on only greediness or history-awareness.