Harnessing quantum mechanics properties, quantum computers have the potential to outperform classical computers in many applications and are envisioned to affect various aspects of our society. Different approaches are being explored for building such computers. One of such potential approaches is Measurement based quantum computation (MBQC), introduced by Raussendorf and Briegel in 2001. In MBQC a large number of qubits are prepared in a highly entangled clusters, called cluster states. The required quantum computation is then performed by a sequence of measurements. Cluster states are being physically realized using continuous variables (CV) and discrete variables (DV) approaches. CV-based approaches can be further categorized as Frequency domain multiplexing (FDM), Time domain multiplexing (TDM), Spatial domain multiplexing (SDM) and hybrid. We discuss and compare these approaches in detail. We also discuss cluster states generation in DV and report some recent results where photons and superconducting qubits are used.