Channel Equalization Using Multilayer Perceptron Networks
📝 Abstract
In most digital communication systems, bandwidth limited channel along with multipath propagation causes ISI (Inter Symbol Interference) to occur. This phenomenon causes distortion of the given transmitted symbol due to other transmitted symbols. With the help of equalization ISI can be reduced. This paper presents a solution to the ISI problem by performing blind equalization using ANN (Artificial Neural Networks). The simulated network is a multilayer feedforward Perceptron ANN, which has been trained by utilizing the error back-propagation algorithm. The weights of the network are updated in accordance with training of the network. This paper presents a very effective method for blind channel equalization, being more efficient than the pre-existing algorithms. The obtained results show a visible reduction in the noise content.
💡 Analysis
In most digital communication systems, bandwidth limited channel along with multipath propagation causes ISI (Inter Symbol Interference) to occur. This phenomenon causes distortion of the given transmitted symbol due to other transmitted symbols. With the help of equalization ISI can be reduced. This paper presents a solution to the ISI problem by performing blind equalization using ANN (Artificial Neural Networks). The simulated network is a multilayer feedforward Perceptron ANN, which has been trained by utilizing the error back-propagation algorithm. The weights of the network are updated in accordance with training of the network. This paper presents a very effective method for blind channel equalization, being more efficient than the pre-existing algorithms. The obtained results show a visible reduction in the noise content.
📄 Content
Channel Equalization Using Multilayer Perceptron Networks SABA BALOCH*, JAVED ALI BALOCH**, AND MUKHTIAR ALI UNAR*** RECEIVED ON 23.12.2011 ACCEPTED ON 21.06.2012 ABSTRACT In most digital communication systems, bandwidth limited channel along with multipath propagation causes ISI (Inter Symbol Interference) to occur. This phenomenon causes distortion of the given transmitted symbol due to other transmitted symbols. With the help of equalization ISI can be reduced. This paper presents a solution to the ISI problem by performing blind equalization using ANN (Artificial Neural Networks). The simulated network is a multilayer feedforward Perceptron ANN, which has been trained by utilizing the error back-propagation algorithm. The weights of the network are updated in accordance with training of the network. This paper presents a very effective method for blind channel equalization, being more efficient than the pre-existing algorithms. The obtained results show a visible reduction in the noise content. Key Words: Blind Channel Equalization, Neural Networks, Noisy Signal, Multi Layer Perceptron, Error-Back Propagation. * Assistant Professor, Department of Electronic Engineering, Mehran University of Engineering & Technology, Jamshoro. ** Assistant Professor, Department of Computer Systems Engineering, Mehran University of Engineering & Technology, Jamshoro.
Meritorious Professor, Department of Computer Systems Engineering, Mehran University of Engineering & Technology, Jamshoro 1. INTRODUCTION W ith the passage of time, digital communication has almost prevailed analog communication. Prominent factors behind the current situation are the escalating demand and falling prices of digital equipment. Digital communication basically includes transferring of certain digital information for instance voice, images or data from the transmitting end to the receiving end, but the data transferred should be received in the actual form [1-2]. Practically this cannot be achieved. ISI is one of the most influential problems faced practically in digital communication. This causes distortion to some of the transmitted symbols due to other transmitted symbols. Performing equalization on the channel can minimize the ISI. The two major reasons of ISI in a channel are as follows: (1) As the channel used for communication has a limited bandwidth, it causes the pulse waveform passing through it, to disperse or spread. If we consider a channel with a much larger bandwidth in comparison to the pulse bandwidth, the spread or dispersing of the pulse should be minimal. On the other hand when the bandwidth of the channel is almost same as the signal bandwidth, the spreading will exceed the symbol duration and cause the signal pulses to overlap [2-4]. This overlapping of symbols is called interference between symbols. (2) Multipath is a signal propagation phenomenon due to which signals may reach the receiving antenna by two or more paths. This causes the transmitted signal to be dispersed in time, which Mehran University Research Journal of Engineering & Technology, Volume 31, No. 3, July, 2012 [ISSN 0254-7821] 469 Mehran University Research Journal of Engineering & Technology, Volume 31, No. 3, July, 2012 [ISSN 0254-7821] 470 Channel Equalization Using Multilayer Perceptron Networks results in overlapping of different transmitted symbols. This is also known as ISI, which can cause high error rates, if not compensated [2,4]. The ISI problem can be solved by devising a means to offset or minimize the ISI at the receiving end before detection. An equalizer can be used as a compensator for the ISI. Many equalization techniques have been proposed and implemented. In some techniques, there is a need to transmit a training sequence prior to signal transmission and some perform equalization without using a training sequence [5-6]. Studying the previous techniques showed the presence of noise even after the equalization process. This motivated us to propose a method which would reduce the noise to a minimal level. This can be achieved using ANNs, which has the advantage of accuracy and provides us with faster response. In the following section we have discussed channel equalization and its types. 2. CHANNEL EQUALIZATION AND BLIND EQUALIZATION One of the most prominent functions for the receivers in many data communication systems is channel equalization. The requirement for data communication is that a specific analog medium be used to transmit the digital signals from the source to the receiver. Practical restraints in analog channels make them imperfect and may cause undesired distortions to be introduced [7-8]. In linearly distorted channels, the distortion can be effectively removed and compensated with the help of channel equalization. In other words inter symbol interference can be mitigated by performing equalization. Here the equalizer coefficients are initially adjusted by transmitting a known trail sequence to the receiver. However, i
This content is AI-processed based on ArXiv data.