Conventional point object model based automotive radar tracking system assumes at most one detection received from the target object at a time. However, in real applications for an extended object, such as a passenger car, located within a close range to a high-resolution radar or LIDAR, the system usually receives multiple reflections from different parts of the object. This can introduce large bias into a velocity estimation performed by a point object model based tracking system. Doppler-azimuth profile based approach accounts for the cluster of detections and could give a very accurate velocity vector of the extended object. However, depending on the position and orientation of the object, the linear equation set could be ill-conditioned, in which case the estimated velocity will suffer from substantial error. In this paper, we first propose a new approach to estimate the heading of a moving object using principle component analysis based on the detection cluster trajectory. We then propose an approach to fuse the three above-mentioned velocity estimators as each estimator faces challenges in different situations. Road data from a 77GHz radar is used for performance illustration.

Many problems involve both decision and estimation where the performance of decision and estimation affects each other. They are usually solved by a two-stage strategy: decision-then-estimation or estimation-then-decision, which suffers from several serious drawbacks. A more integrated solution is preferred. Such an approach was proposed in [14]. It is based on a new Bayes risk as a generalization of those for decision and estimation, respectively. It is Bayes optimal and can be applied to a wide spectrum of joint decision and estimation (JDE) problems. In this paper, we apply that approach to the important problem of joint tracking and classification of targets, which has received a great deal of attention in recent years. A simple yet representative example is given and the performance of the JDE solution is compared with the traditional methods. Issues with design of parameters needed for the new approach are addressed.

This paper presents a novel approach to predict the Internet end-to-end delay using multiple-model (MM) methods. The basic idea of the MM method is to assume the system dynamics can be described by a set of models rather than a single one; by running a bank of filters (each corresponds to a certain model in the set) in parallel at the same time, the MM output is given by a combination of the estimates from these filters. Based on collected end-to-end delay data and preliminary data analysis, we propose an off-line model set design procedure using vector quantization (VQ) and short-term time series analysis so that MM methods can be applied to predict on-line measurement data. Numerical results show that the proposed MM predictor outperforms two widely used adaptive filters in terms of prediction accuracy and robustness.

This paper presents a novel approach to predict the Internet end-to-end delay using multiple-model (MM) methods. The basic idea of the MM method is to assume the system dynamics can be described by a set of models rather than a single one; by running a bank of filters (each corresponds to a certain model in the set) in parallel at the same time, the MM output is given by a combination of the estimates from these filters. Based on collected end-to-end delay data and preliminary data analysis, we propose an off-line model set design procedure using vector quantization (VQ) and short-term time series analysis so that MM methods can be applied to predict on-line measurement data. Numerical results show that the proposed MM predictor outperforms two widely used adaptive filters in terms of prediction accuracy and robustness.