Low power and low complexity algorithms for signal processing applications have gained increasing importance with the deployment of portable communication equipment. Most of the current power reduction techniques rely on reducing the power by VLSI implementation. This approach is expensive and limited by technology. Hence algorithm design optimization is a must for low energy consumption. Here, we propose using finite memory detection algorithms as low complexity-low energy near optimal detection algorithm that trades a small amount of detection performance for a reduction in complexity and power consumption. The negligible loss in detection performance is easily accommodated in wireless video and audio transmission applications. In data applications, this small loss can be further reduced with error correcting codes at the expense of a slight loss in communication bandwidth. We present simple algorithms for deriving the near optimum finite memory detectors in the time invariant and time variant case. The same algorithms can be used in tandem configurations in decenteralized detection.
[1]
Ernest G. Baxa,et al.
Adaptive Signal Detection with Finite Memory
,
1971
.
[2]
Saleem A. Kassam,et al.
Optimum Quantization for Signal Detection
,
1977,
IEEE Trans. Commun..
[3]
Maurizio Longo,et al.
Quantization for decentralized hypothesis testing under communication constraints
,
1990,
IEEE Trans. Inf. Theory.
[4]
H. Vincent Poor,et al.
On Optimum and Nearly Optimum Data Quantization for Signal Detection
,
1984,
IEEE Trans. Commun..
[5]
Thomas M. Cover,et al.
Optimal Finite Memory Learning Algorithms for the Finite Sample Problem
,
1976,
Inf. Control..
[6]
T. Cover.
Hypothesis Testing with Finite Statistics
,
1969
.
[7]
Patrick Duvaut,et al.
Optimum quantization for detection
,
1988,
IEEE Trans. Commun..
[8]
Ramanarayanan Viswanathan,et al.
Optimal serial distributed decision fusion
,
1988
.