Video Transport Over Ad Hoc Networks Note The MESH Topology of the Ad Hoc network provides a possibility to establish multiple paths between a pair of sources and destination nodes. In fact, multiple transmission paths provide additional degrees of freedom for design fault-tolerant video codes and transport mechanisms. In this article, we tried to combine multi-path transmission to illustrate that in addition to the traditional error control technology violates, the difference is effective for suppressing an effective solution for the Ad hoc endload error. . The three mechanisms we have to investigate are: 1) Feedback-based reference image selection; 2) Layered encoding and (ARQ); 3) a plurality of descriptions of action compensation coding. All three technologies are based on the operation compensation prediction technique of modern video coding standards. We study the performance of these three mechanisms by using the Markov channel model and OPNet Modeler to establish an expanded simulation environment. In order to verify its feasibility and performance advantages, we use the notebook and the 802.11b network card to implement a Testbed of an Ad hoc multipath video stream. The results indicate that good video transmission quality can be obtained under standard mechanisms that limit additional overhead. In the different requirements of whether the feedback channel, end-to-end delay constraint, and error characteristics, our three video coding / transmission technologies can be adapted.
I. The characteristics of the primer Ad hoc network are huge challenges brought by transferred video information. The network caused by the mobility of the terminal cannot have a fixed topology. This makes the connection between the nodes in a non-stop-connected-end-end cycle. The fading of the channel and the user who are communicating will make the reliability of the connection quality changes. In addition, the path to the end of the end includes a wide range of wireless links. Therefore, it is more likely to occur in the transmission loss in the AD HOC. The famous Mac and CSMA / CD mechanisms are designed to try their best to transfer data. They have no strict guarantees for the bandwidth and delay of the session. Although bandwidth reservations can be implemented on the Mac through TDMA and CDMA, it is cumbersome to implement the synchronization or code assignment problem that considers the mobility of the terminal. Video transmission requires strict bandwidth and delay protection. However, there are enough bandwidth and a very stable end-to-end route between the two nodes. It is very difficult. In addition, the compressed video information is easily error. For example: When Variable Length Coding, VLC is used, a bit of the fault is easily caused to lose synchronization. In addition, the Motion Compensated Prediction, MCP technology is widely used in modern video coding standards. In the MCP, a frame is predicted from a front-encoded frame (this frame is generated in a reference image), and then the error in the prediction is also encoded and transmitted. Although MCP utilizes temporary relationship between adjacent frames, it makes it necessary to successfully recombine a frame must be established based on its reference image correctly. If there is no effective error protection and hiding technology, the loss of one of the packets in one frame will not only result in the error of this frame, but also make the subsequent frames error. Considering the Easy error characteristics of the AD HOC network path and compressed image in transmission, there must be a set of effective error control technology. For conventional technologies, the forward error correction (FEC) and ARQ must be adapted to control delay constraints and error propagation. In the Ad hoc network, the wireless link breaks through the concept of traditional topologies, which no longer receives the limitations of the physical cable. Although the mobility of the user terminal makes the link more fragile, it provides a variety of topologies. On the one hand, when the two nodes move in the opposite direction, the link will be interrupted; on the other hand, it may discover new topologies for new routes. In addition, Ad hoc's Mesh topology implies a plurality of links in two nodes. It is assumed that there is a plurality of paths, and a video stream can be divided into multiple subflows and each subflow is transmitted on a different path. If these paths are not integrated, the chances of errors in each strand are also independent of each other. Therefore, appropriate flow allocation and error control for video subflow can give us better fault tolerance. In this article, we proposed three video transmission technologies based on MCP (action compensation forecasts) for the AD HOC network. These mechanisms use the characteristics of multiple transmission paths between end-to-ends to achieve good performance. These techniques proposed here more coding efficiency is more compatible with H.26X, MPEG series standards compared to the coding. The technologies we have to investigate include: 1) Based on the feedback-based image selection mechanism (RPS); 2) Layered encoding (LC) and selection retransmission ARQ mechanism (LC with ARQ); 3) Multi-described action compensation encoding mechanism (MDMC). We use a top-down approach to study the performance of these mechanisms. First, we use a very famous Marcov chain model with the underlying detail only in a burst error generator.
This simple model allows us to examine the performance of the system's packet loss rate and loss of connections in a large range. Second, we use opnet to simulate, we include the underlying details, including mobility, multi-path routing, and MAC layer protocols, etc., which can examine these factors on system performance from the perspective of actual situation. In addition, we are willing to be happy with an Ad hoc network-based video stream test bed. This test bed can help us further verify the feasibility and performance advantages of each mechanism. The results of our trial show that video transmission is feasible in the AD HOC network that is carefully closely designed. Muttistream Coding combines Muttistream Coding with MPT (MultiPath Transportation) with a conventional method of using a single path increases the quality of video transmission. The three techniques here are suitable for three specific use environments: whether the feedback channel, end-to-end delay constraints, and the error characteristics of the link. The remainder of the article is organized as follows: In the second quarter, we propose a multi-channel multi-channel video coding general structure. At the same time, we also discuss some questions related to the preceding work involved in this section. Second, discuss multi-channel multi-channel encoding mechanisms in Section III. Section IV and Section 5 discuss the performance indicators of the three mechanisms obtained using the Markov model and OPNet Modeler. Section 6 Discuss the conclusions obtained on the test bed we have established AD HOC network video flow test bed. Section 7 draws conclusions.
