联合检测估计及其性能分析

二元假设检验问题可视作对一个取值0、1的离散随机变量的估计,于是原先分离的信号检测和参数估计可以纳入统一的框架进行处理,这就是联合检测估计。介绍了该方法的一般形式,并就不同的情况讨论它的性能及其与传统顺序方法的比较。发现在线性模型下,联合方法和顺序方法具有内在的等价性;但对更一般的情况,则要略逊色于顺序方法。联合方法的优势在于只需计算一个广义的MLE而不是多重积分,因而在应用中更为可行。特别在高信噪比条件下,两种方法的差别微乎其微,联合方法是一种更为可取的准最佳方法。

Joint detection and estimation and the performance analysis

The problem of binary hypothesis testing can be addressed in terms of estimation of a discrete parameter which takes value between 0 and 1. Therefore the problem of signal detection and parameter estimation, which were separated originally, can be considered in a general framework named as Joint detection and estimation. In this paper, the basic mechanism of joint scheme is introduced. An analysis of its performance and a comparison with the tradi-tional serial scheme are also made. It can be found that, for the classic linear model with Gaussian assumption, the joint detection and estimation scheme is identical in nature with its counterpart. While in other cases such as nonlinear model, it does not perform as well as the serial one. The advantage of joint scheme is that only a computation of gen-eral MLE instead of integration is needed, which makes it more favorable in applications especially in a high SNR en-vironment.