We address the problem of collecting data in a wireless sensor network, where network coding is used for data transmission. The temporal correlation between the measure- ments is exploited to recover the data at the receiver. For this purpose, network encoding operations are considered in conjunction with lossy source compression, performed by an LDGM code set generated during transmission. The decoding is carried out using the iterative doping algorithm based on the sum product algorithm, on a graph which represents the LDGM encoding equations. Consequently, we first study the design tradeoffs of LDGM-based lossy source coding for significant parameters, such as packet size reduction and distorsion. In addition, we aim to show the prospective impact of the network coding on a wireless sensors network. This involves identifying how linear codes improve key performance criteria for networks such as rate, delay, and convergence time. Our work was applied to a real case study to highlight consistency and efficiency of our proposed system