Robust Joint Channel Estimation and Symbol Detection via Sphere Decoding

J.L. Lagunas Morales and S. Roy (Canada)


Sphere decoder, channel estimation, underdetermined channels, virtual MIMO, joint detetection.


In this paper, we present a method which jointly performs robust channel estimation and signal detection. In most re ceive antenna arrays, a known signal preamble at the begin ning of the data frame is used to estimate the channel ma trix. In this document, a first channel estimate is based on the received signal subspace analysis which uses statistics from the entire frame of the received signal. Subsequent es timations of the channel matrix are obtained jointly at each signal detection step. Detection is achieved via the Sphere Decoder (SD) which is a powerful and low-complexity suboptimal M-estimator. This technique achieves an ex cellent estimation of the channel matrix by using all de tected symbols of the data frame. This is equivalent to having the training preamble occupy the entire frame. The Mean Square Error (MSE) of the channel estimation for this method is very low over a wide range of preamble lengths, making it a bandwidth-efficient technique. The approach is applicable in both full rank (N = M) and rank deficient (N < M) channels with practically the same performance as that of a clairvoyant detector with perfect CSI knowledge. This chracteristic is crucial for the use of Virtual-MIMO algorithms.

Important Links:

Go Back