Performance evaluation of Kalman filter sensor fusion based algorithm and gain fusion based algorithm

The aim of this paper is to analyze the different schemes of data fusion techniques. The data will be generated for different sensors and fused together to generate or deduce an output. The different schemes for data fusion technologies are mentioned in the text. For our analysis we will consider KFA, XXX YYY techniques. The output will be shown with the help of graphs. These schemes are applicable for multi sensor data fusion where the data is generated or collected via different sensors. DF (Data fusion) or multisensory data fusion (MSDF) is the process of combining or integrating measured or pre-processed data or information originating from different active or passive sensors or sources to produce a more specific, comprehensive, and unified dataset or world model about an entity or event of interest that has been observed. Sensor data fusion has wide number of applications from telecommunications, radar, satellite communication to data networks. The tracking of moving objects includes targets, mobile robots, and other vehicles which uses measurements from sensors is of considerable interest in many military and civil applications that use radar, sonar systems, and electro-optical tracking systems (EOTs) for tracking flight testing of aircrafts, such as missiles, unmanned aerial vehicles, micro- or mini-air vehicles, and rotorcrafts. It is also useful in nonmilitary applications such as robotics, air traffi c control and management, air surveillance, and ground vehicle tracking. In practice, scenarios for target tracking could include manoeuvring, crossing, and splitting (meeting and separating) targets. Various algorithms are available to achieve target tracking for such scenarios. The selection of the algorithms is generally application dependent and is also based on the merits of the algorithm, complexity of the problem (data corrupted by ground clutter, noise processes, and so on), and computational burden. Target tracking comprises estimation of the current state of a target, usually based on noisy measurements. The problem is complex even for single target tracking because of uncertainties in the target’s mathematical model, especially for manoeuvring targets (which need more than one model and one transition model, and so on), and process/state and measurement noises. The complexity of the tracking problem increases for multiple-targets using measurements from multiple sensors.