Wireless Sensor Network (WSN) is a set of inexpensive densely deployed wireless sensor nodes with limited functionalities and scarcity in energies, whose observations are forwarded or relayed by intermediate nodes to the Base Station (BS). In the networks with densely deployed nodes, the observations are likely to be highly correlated in the space domain. This type of correlation is referred as spatial correlation, which produces unfavorable redundant readings causing energy wasting. In this thesis, the main task is to reduce these nodes that have redundant readings by using a clustering algorithm called Event Based Clustering (EDC) algorithm. The clustering algorithm is based on exploiting the spatial correlation that is used to cluster the sensor nodes. Exploiting spatial correlation is proposed by using Vector Quantization (VQ) with respect to the distortion constraints. Furthermore, this algorithm is applied for energy harvesting sensor nodes. Also, the inessential sensor nodes that have correlated readings are reduced for improving the Energy-Efficacy (EE) with acceptable level of event signal reconstruction distortion at the sink node. After applying the EDC algorithm, the communication model is changed from single-hop model to two-hop (clustered-network) model. Hence, a theoretical framework of distortion function, i.e., accuracy level, for both single-hop and two-hop communication models is derived. Then, single-hop and two-hop communication models are compared in terms of achieved distortion level, number of alive nodes, and energy consumption for different sizes of event area. Finally, the effects of various harvested energy level on the clustered-network is studied with respect to the same terms.