Low connectivity is a challenging problem in wireless communication sensor networks with non-uniform sensor node densities. Reliable information transmission requires the full connectivity of the network because of the multi-hop communication of wireless sensor networks. A network is fully connected if each pair of nodes can communicate with each other, either directly or through intermediate relay nodes. The current study presents a Harmony Search (HS) algorithm to improve the connectivity of the non-uniform density wireless sensor networks. HS is a meta-heuristic algorithm, which mimics the actions of harmonizing musical instruments, including memory-based, random, and pitch-adjusted play, during improvisation. In the present work, HS is used to optimize the phase shift parameters of sensors node inside the cluster to ensure maximum radiation field at the reception point. This process results in a significant increase in the coverage area of cluster nodes, consequently increasing network connectivity.