Anas Youssef Boubas

Joint Routing and Radio Resource Management in Multihop Cellular Networks using Genetic Algorithms

This paper addresses the problem of joint routing and radio resource management in multihop cellular networks with fixed relays. The objective is to maximize the total network utility. This paper focuses on quality of service (QoS-) sensitive applications that are captured by sigmoidal-like utility functions. The non-concavity of the user utility functions, the complex relation between routing decisions and network resource allocation, and the dependence of link capacities on all interfering transmission powers render the problem non-convex, and extremely difficult to solve. However, we present a meta-heuristic based on genetic algorithms that solves the problem successfully, efficiently and globally.

Rapid Interlacing and High Compression Rates For Images Using Singular Value Decomposition

This paper introduces a method for compressing and progressively transmitting an image. The method is based on singular value decomposition (SVD), and is superior to both the GIF and JPEG progressive standards. SVD is done offline to create the image file. Once created, the image file can be transmitted normally, and recomposed online by the receiving party.

All-optical processing using light intensity and wavelength recognition

A simple multi-valued processing logic relying on two basic operations, namely, addition and comparison, is presented. These two basic operations have minimal switching levels, since the addition operation is performed without switches and the comparison operation is performed with only one switch. A simplified illustrative model of a system based on this logic is also presented. The paper includes a limited survey of optical components that can be used in the presented system.

A Multicomponent Image Quality Assessment based on Singular Value Decomposition

Most of the image quality measures rely on the amount of noise added to the image. Determination of both the noise amount and its type is crucial, as some applications may be sensitive to some noise types more than the others. We propose a new multicomponent quality measure that is based on singular value decomposition (SVD), and provide a way to identify some noise types using different components of the measure. The measure is compared to peak signal to noise ratio (PSNR), and is shown to distinguish between noise types the PSNR is not able to distinguish between. Standard images were used in the experiments.