A Reduced-State Viterbi Algorithm for Blind Sequence Estimation of DPSK Sources
نویسندگان
چکیده
The Viterbi algorithm is the optimum decoding algorithm for convolutional codes and has often served as a standard technique in digital communication systems for maximum likelihood sequence estimation. With the Viterbi Algorithm, the computational complexity increases exponentially with the constraint length of the convolutional code. Reducing the constraint length (hence the number of states) would permit major sim-pliication in the implementation of the Viterbi algorithm. In this paper, a reduced-state Viterbi algorithm for blind sequence estimation of DPSK sources is presented. It can reduce number of states in the Viterbi algorithm by at least half. The reduced state DPSK Viterbi decoder can be made much faster without any performance loss. The Viterbi algorithm, also known as maximum-likelihood decoding algorithm, is the optimum decoding algorithm for convolutional codes. It has often served as a standard technique in digital communication systems for maximum likelihood sequence estimation. With the Viterbi Algorithm, the number of states, or equivalently the storage and computational complexity increases exponentially with the constraint length of the convolu-tional code. Obviously, reducing the constraint length (hence the number of states) would permit major sim-pliication in the implementation of the Viterbi algorithm. In 6], Tong proposed a blind sequence estimation scheme by exploiting the second-order statistics of the source. The correlation of the deterministic source was estimated from the observation and then the Viterbi algorithm was applied to reconstruct the input symbols. In this paper, it is shown that number of states in the trellis can be reduced by at least half for DPSK sources while obtaining the same performance beneets of the Viterbi algorithm.
منابع مشابه
Simultaneous Source Decoding and Blind Equalization of Multipath Fading Channels
In this talk, an introduction will be given to some mathematical tools that have recently been developed for blind equalization of frequency selective fading channels. Standard approaches to the problem are implemented in two steps: First, a channel estimate is obtained, and second, the estimated channel is used to recover the transmitted sequence by means of a zero-forcing or MMSE filter, or u...
متن کاملBlind Joint Maximum Likelihood Channel Estimation and Data Detection for SIMO Systems
A blind adaptive scheme is proposed for joint maximum likelihood (ML) channel estimation and data detection of singleinput multiple-output (SIMO) systems. The joint ML optimisation over channel and data is decomposed into an iterative optimisation loop. An efficient global optimisation algorithm called the repeated weighted boosting search is employed at the upper level to optimally identify th...
متن کاملImproving Phoneme Sequence Recognition using Phoneme Duration Information in DNN-HSMM
Improving phoneme recognition has attracted the attention of many researchers due to its applications in various fields of speech processing. Recent research achievements show that using deep neural network (DNN) in speech recognition systems significantly improves the performance of these systems. There are two phases in DNN-based phoneme recognition systems including training and testing. Mos...
متن کاملMatched Filter Bounds for Reduced-Order Multichannel Models
We propose two Matched Filter Bounds (MFBs) to characterize the performance of receivers using reduced-order channel models. The rst one (WMFB) uses the channel model to perform the spatio-temporal matched ltering that yields data reduction from multichannel to single-channel form. The rest of the processing remains optimal. The second one (ICMFB) on the other hand bounds the performance of the...
متن کاملBlind Estimation of Linear and Nonlinear Sparse Channels
This paper presents a Clustering Based Blind Channel Estimator for a special case of sparse channels – the zero pad channels. The proposed algorithm uses an unsupervised clustering technique for the estimation of data clusters. Clusters labelling is performed by a Hidden Markov Model of the observation sequence appropriately modified to exploit channel sparsity. The algorithm achieves a substan...
متن کامل