No abstract

During the past couple of decades several proposals for image coders using the singular value decomposition (SVD) have been put forward. The results using the SVD in this context have never been spectacular.The main problem with the SVD is that the transform itself must be transmitted as side information. We demonstrate through some simple experiments that for a given image reconstruction quality, more scalar parameters must be transmitted using the SVD, than when using the discrete cosine transform (DCT).Also, using an alternative interpretation of the SVD we show that the SVD representation necessitates quantization of individual factors as compared to quantization of the associated product. This is clearly suboptimal.

The plethora of wireless devices resulting in a humongous data usage has already pushed the current mobile networks to their limit. Therefore the research and development of the next generation mobile network must take place now. In this regard a mobile network operator has a pivotal role in understanding the required performance of the coming fifth generation network, and also influencing its final design. This paper presents some network research topics seen from an operator's point of view. It provides an overview of some recent results within the following areas: network architecture utilizing network function virtualization and software defined networks, performance of deploying self-organized network functions, spectrum sharing, inter-cell interference reduction methods, and backhaul with gigabit radio links.

Singular value decomposition (SVD) is an adaptive transform with excellent energy compaction properties for separable data. Hence, applying SVD on seismic stack sections seems to be a good choice. In order to achieve a good SVD compression system, bit-efficient representations must be found for both the transform (singular vectors) and the transform coefficients (singular values). In this paper we present an SVD coding system which exploits the orthonormality of the singular vectors to find a minimum number of parameter representation for the singular vectors. These parameters and the singular values are quantized using pdf optimized quantizers. The SVD coder is designed to give full control of the mse within each image block. The proposed method has the ability to compress seismic stack sections at a ratio of 30:1 with no loss of visual information. Regarding the novelty of this scheme this must be considered to be a good result, although it is outperformed by a standard JPEG image coder in terms of signal to noise ratio.

No abstract