N-bit Parity Implementation based on Binary Neural Networks with O(n/2) Neuron and O(2n) Connection

Ehsan Lotfi

Abstract


In communication systems, n-bit parity problem (NPP) is widely used for error detection and correction. In this paper, an efficient architecture for hardware implementation of NPP is proposed. For this purpose, first we introduce extending single neuron that can be trained by perceptron learning rule to solve 2-bit and 3-bit parity problems. Then, we present novel architecture of cascaded modular neural networks, based on the presented neuron, with O(n/2) nodes and with O(2n) connections to solve NPP.  The main advantages of proposed parity networks are low number of neurons, connecting weights and inputs for each neuron.

Keywords


Perceptron; XOR; Neural Nets; Optimum Architecture

Full Text:

PDF

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.

ACSA©: World Science Publisher United States