In wireless sensor network (WSN), the localization scheme is vital for efficiently utilizing the acquired sensing data from sensor nodes. The received signal strength (RSS) based localization is among one of the most preferable schemes for various applications, due to its ease of implementation without any additional hardware cost. In RSS-based localization scheme, the anchor nodes need to acquire RSS measures of unknown/target nodes and send them to base station where localization of unknown/target node is performed. As the number of unknown/target nodes increases, the control traffic for localization increases, due to RSS measures for each unknown/target node are gathered and send to the base station (BS), which can cause more energy consumption and reduce life time of sensor networks. In this paper, we propose a localization scheme for multiple unknown/target nodes based on Bayesian Compressive Sensing (BCS). The RSS values received from multiple unknown/target nodes are superposed in each anchor node, and further can be constituted as a form of compressed data. In the proposed scheme, each anchor node sends only a compressed RSS measure from multiple unknown/target nodes to the BS and the BS estimates the location of unknown/target nodes using Bayesian Compressive Sensing. The proposed scheme is evaluated through simulation to authenticate its effectiveness in term of accuracy of localization, and reduction of traffic overhead.