The wireless connection has become more common in recent years this increases the probability of interference between signals and the environment noise, causing the signal at reception to present a considerable difference from the transmitted signal. In this work we develop algorithms that simultaneously apply adaptive filter techniques and network adaptation. The algorithms developed are intended to be applied in a communication system for IoT devices and wireless sensor networks. The purpose is to reconfigure the filters and the connection between the nodes so that the signal error at reception is as small as possible. In an adaptive filter algorithm we have knowledge of the input and output signals, we start the process by applying an initial value vector 0 which in each interaction is varied as a function of the magnitude of the error. Based on this technique we can develop even more efficient variations. We use variations that locate and exclude frequencies that are less likely to be present in the signal besides adjusting the network configuration. The techniques used are variations of the Least Mean Square algorithm (LMS). The techniques used for adapting the networks are Exhaustive Search (ES) and Sparsity-Inspired (SI) to adapt the filter we apply to ACDA. In addition, it is proposed to combine the algorithms of network adaptation and parameters in a single one that generates a more accurate result. The spread spectrum modulation used in LoRa has the potential to be the main network applied in most IoT projects and applications. In the simulations we will present signals and configurations compatible with the LoRa system.