Average lifetime improvement of a critical application sensor network is always a promising research area where a lot of scope for applying a technique like cross-layer approach, energy-aware routing, load balancing etc. Due to the heterogeneous nature of sensor network and dynamic situation static solutions always fail to give an optimal solution. The performance of network like an average lifetime, throughput, packet delivery ratio, overhead due to routing etc can be improved through an optimized load balanced network. Optimisation technique like simulated annealing, genetic optimization and particle swarm optimization yield a time-varying performance due to its operation of principle. The various network parameters can be used to represent chromosome in the genetic algorithm (GA) and similarly this parameter can be represented in particle swarm optimization (PSO) optimization. The performance of GA and PSO is measured in the dynamic environment but testing is done in a standard environment.