Over the past few years, Vehicular Delay Tolerant Networks (VDTNs) have been exploited in those communications that do not take place through end-to-end connectivity. VDTNs framework is characterised by frequent disconnection, network partitioning, and high delay. A number of routing protocols have already been proposed for VDTN. These protocols make effort to achieve significant delivery potential with low network overhead. Geographic routing is an alternative. The protocols in geographic routing utilise the fact that most vehicles are equipped with Global Positioning System (GPS). The major concern of Geographical Routing Protocols (GRPs) is to enhance the delivery potential and to minimise the delay. This paper provides a comprehensive investigation of recently proposed GRPs. Additionally, open issues and future orientations are included in order to motivate the further research.

Although the currently available digital cameras typically provide 256 levels of brightness data at each pixel, a real world scenes contains a very wide range of brightness variations. High Dynamic Range (HDR) images contain wider range of brightness information than a normal image and are a new trend in photography. In the conventional method multiple photographs of same scene are captured at different exposures and then combined to produce HDR images. Clearly these methods require a still scene and are not able to produce HDR image for moving scene or moving camera. We present a method based on Spatial Varying Exposure (SVE) system. Here we capture only a single photograph of the scene but vary the exposure spatially. By this method it is possible to create the HDR image for dynamic scene by trading-off spatial resolution.Key words-dynamic range, spatial varying exposure, Bayer Filter Array, exposure I.INTRODUCTIONCurrently available digital cameras are able to capture the images of size in megapixels with 24 bit of RGB color information. This information is more than enough to be displayed even on a full High Definition (HD) display device like LCD and LED TV. But the contrast ratio or dynamic range captured by cameras is still in the range of 100:1.Dynamic range or contrast is the term used to describe the ratio between the highest and the lowest value of the image brightness. For natural scene the dynamic range is the ratio between the density of luminous intensity of the brightest sunbeam and the darkest shadow [1]. A real-world scene can produce a dynamic range as high as of the order of 10 6 :1. Clearly the dynamic range captured by a digital camera is much below the reality. The High Dynamic Range imaging provides the solution for this issue.There are several methods available to capture an HDR image. The most common way to achieve high dynamic range is to combine multiple images captured in different exposure settings [2]. Fig. 1 illustrates the HDR image problem. But this kind of methods fails when the object is not a still scene. A method based on Spatially Varying Exposure (SVE) is presented in this paper. By trading-off image size, an HDR image is produced in single capture of scene. As the method takes only one shuttering for capturing a multiple exposure photograph, it is more suitable for the moving objects. Fig.1. (a) An under-exposed image of Dept. of E.E., IISc. (b) An over exposed image (c) Tone mapped HDR image created from the (a) and (b) using [2]. II. RELATED WORKHDR images of still natural scenes can be acquired by capturing multiple images of the same scene with different exposures [2] [3]. Recently some very high sensitivity cameras are also coming into the picture to capture HDR data [4]. Such HDR sensor designs focus on enhancing the dynamic range by adding the maximum amount of light, but are very costly and are not commonly available.The spatially varying exposure in Color Filter Array in single photograph is a good choice for creating HDR image for moving or dynamic scene, at the ...

No abstract

Vehicular delay tolerant networks (VDTNs) suffer due to insufficient connection opportunities in sparse infrastructure. The sparsity of the network results in poor delivery ratio (DR). Deployment of additional small nodes named throwbox increases the contact opportunities, resulting in improved performance in terms of DR. However, randomly deployed throwboxes do not provide optimum results. Therefore, this article proposes the throwbox deployment as an optimization problem. Customized binary particle swarm optimization-based optimization method is employed to solve the problem. With the objective of maximizing the DR, the method optimize the placement of k-throwboxes in the network. The performance of the proposed methods is tested and compared by simulating a VDTN model in opportunistic network environment simulator. The optimization framework was implemented in Matlab. Exhaustive experiments were carried out by varying the number of throwboxes (k). A statistical comparison of the results with the existing throwbox deployment methods validates the effectiveness of the proposed optimization-based approach. The article also presents an analysis for choice of value of k for a desired network performance.