Efficient Technique for Joint Parameter Estimation of Mixed Far Field Targets and Near Field Jammers in Phased Array Radar

Parameter estimation of the impinging signal in the sensor arrays is one of the challenging issues in Phased Array Radar (PAR). Most often, a PAR operates in the presence of jammers deployed in its near field. In order to mitigate the jammers signals, it is important that PAR must know the exact Direction Of Arrival (DOA) and ranges of the jammers along with the DOAs of the actual targets located in the far field. This information is crucial for the PAR to steer the mainbeam and nulls in the desired and undesired (jammers) directions, respectively. In this work, we consider a PAR that simultaneously receives the desired plane waves and the undesired spherical waves from the far field targets and near field jammers respectively. The plane waves are the function of the unknown DOA, whereas the spherical waves depend on the DOAs and ranges of the jammers. A novel approach based on the Gradient Based Optimizer (GBO) algorithm and Marine Predators Algorithm (MPA) is proposed to estimate jointly the unknown parameters related to the far field targets and near field jammers. The GBO uses two operators called Gradient Search Rule (GSR) and Local Escaping Operator (LEO). The GSR enhances the convergence rate and improves the exploration tendency to obtain the better solution in the search space, while the LEO helps the algorithm avoid getting stuck in the local optima. The MPA uses different velocities ratios based on predators and prey movement in the ocean to achieve better exploration and exploitation in the search domain. A cost function is developed and optimized using GBO and MPA, which is based on the Penalty Function (PF) and Mean Square Errors (MSE) of the system. The PF controls the deviation in the MSE from the desired result during the optimization process. The proposed schemes are not only compared with each other but also with the state-of-the-art MUSIC algorithm. Extensive Monte Carlo simulations are enacted for diverse scenarios to justify the legitimacy of the proposed algorithms.