A neural network approach to blind source separation

Mejuto, C.; Castedo, Luis

doi:10.1109/nnsp.1997.622430

Cited by 3 publications

(4 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…h i is the i-th element of the vector h(z[n]), and ′ denotes the first derivative. The maximum of this cost function can be obtained using a gradient algorithm [2], or a relative gradient algorithm [1,22]. Both approaches use the gradient of Equation (20) given by where adj(·) is the adjunct of a matrix and…”

Section: Unsupervised Approachmentioning

confidence: 99%

“…where we have used that z[n] = W H [n]y[n]. The expression in Equation (22) admits an interesting interpretation by means of the use of the nonlinear function g(z) = z * (1 − |z| 2 ). In this case, Castedo and Macchi [7] have shown that the Bell and Sejnowski rule is equivalent to the Constant Modulus Algorithm (CMA) proposed by Godard [12].…”

Section: Unsupervised Approachmentioning

confidence: 99%

See 1 more Smart Citation

Channel estimation techniques for linear precoded systems: Supervised, unsupervised, and hybrid approaches

et al. 2011

View full text Add to dashboard Cite

Linear precoding is an attractive technique to combat interference in multiple-input multiple-output systems because it reduces costs and power consumption at the receiver end. Most of the frequency division duplex systems with linear precoding acquire the channel state information at the receiver side by using supervised algorithms. Such methods make use of pilot symbols periodically provided by the transmitter. In a later step, this channel state information is sent to the transmitter side through a low-cost feedback channel. Thus, the available channel information allows the transmitter to adapt signals to the channel conditions. Given that pilot symbols do not convey user data, the inclusion of such symbols penalizes the throughput, the spectral efficiency, and the transmission energy consumption of the system. In this work, we propose to mitigate the above-mentioned limitations by combining both supervised and unsupervised algorithms to acquire the channel state information needed by the transmitter. The key idea consists of introducing a simple criterion to determine whether the channel has suffered a significant variation which could require the transmission of pilot symbols. Otherwise, when small fluctuations happen, an unsupervised method is used to track these channel variations instead. This criterion will be evaluated by considering two types of strategies for the design of the linear precoders: Zero-Forcing (ZF) and Wiener criteria. * Corresponding author. Phone: +34 98167000 ext. 1366, FAX: +34 981167160

show abstract

Section: Unsupervised Approachmentioning

confidence: 99%

Section: Unsupervised Approachmentioning

confidence: 99%

Channel estimation techniques for linear precoded systems: Supervised, unsupervised, and hybrid approaches

et al. 2011

View full text Add to dashboard Cite

show abstract

“…where h i is the i-th element of the vector h(z[n]) and ′ denotes the first derivative. The maximum of this cost function can be obtained by means of using a gradient algorithm [3] or a relative gradient algorithm [10], [11]. Both approaches use the gradient of Equation ( 16) which is obtained as follows…”

Section: B Unsupervised Approachmentioning

confidence: 99%

“…This means that the sources are recovered in the same order as were transmitted. Taking into account, Equation (11) implies that optimum separation matrix produces a diagonal matrix Γ [n] and, therefore, the mismatch of Γ [n] with respect to a diagonal matrix allows us to measure the variations in the channel.…”

Section: A Decision Rulementioning

confidence: 99%

A Novel Hybrid Approach to Improve Performance of Frequency Division Duplex Systems with Linear Precoding

Castro

García-Naya

Iglesia

et al. 2010

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Linear precoding (LP) is an attractive technique to combat interference in Multiple-Input/Multiple-Output (MIMO) communication systems because it reduces cost and power consumption in the receive equipment. In most Frequency Division Duplex systems with LP, the Channel State Information (CSI) is acquired at the receiver by using supervised algorithms which work with pilot symbols periodically sent by the transmitter. Subsequently, the CSI is sent to the transmit side through a low cost feedback channel. In order to reduce the overhead inherent to the periodical transmission of training data, we propose to acquire the CSI by combining supervised and unsupervised algorithms. The simulation results show that the performance achieved with the proposed scheme is clearly better than that with standard algorithms.

show abstract

Combination of supervised and unsupervised algorithms for communication systems with linear precoding

Dapena

Castro

Labrador

2010

The 2010 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

A neural network approach to blind source separation

Cited by 3 publications

References 13 publications

Channel estimation techniques for linear precoded systems: Supervised, unsupervised, and hybrid approaches

Channel estimation techniques for linear precoded systems: Supervised, unsupervised, and hybrid approaches

A Novel Hybrid Approach to Improve Performance of Frequency Division Duplex Systems with Linear Precoding

Combination of supervised and unsupervised algorithms for communication systems with linear precoding

Contact Info

Product

Resources

About