Wireless sensor networks (WSN) are considered as a future main platform in many technologies where they are applied in several fields such as industry, medicine, militaries, environmental monitoring and others. In this paper, we propose a new vision of WSN communication based on a new IDWPT / DWPT transformation architecture. In this context, our efforts are devoted, in the begging, to the development of a reconfigurable functional architecture for WSN in both modes: mono-user and multi-users. The functional architecture is based on inverse discrete wavelet packet transform (IDWPT) in the emitter and discrete wavelet packet transform (DWPT) in the receiver. To improve the performance of the proposed functional model, we have launched many simulations under AWGN noise and with different wavelets. Due to the importance of hardware implementation, in the life of WSN, we have also developed in this work a new hardware architecture of DWPT/ IDWPT. This can provide a throughput, configurable and generic architecture. Where we developed this hardware architecture in VHDL at RTL level and implemented in a NexysVideo board (Artix 7 FPGA). A limited amount of hardware was used to achieve high throughput, achieved by intelligent sharing of hardware resources between the low-pass and high-pass filters in the wavelet packet transformation algorithm. In addition, the proposed model was developed without using memory between or within different depths of DWPT / IDWPT transform.